In short
Amit RG (RichPanel) argues SaaS is a “sinking ship” because AI makes code and support cheaper, faster, and more commoditized; humans become the bottleneck. He predicts rapid diffusion (12–24 months) and major job displacement (especially call-center/ticket work). RichPanel’s strategy is to “play fast” in an agent economy by delivering CX automation and infrastructure before it becomes commodity.
Guest backgrounds
Amit RG is CEO of RichPanel (agentic customer service/CX automation). Mike Beckham hosts Operators Podcast. Sean (co-host) discusses agent tools and OpenClaw security concerns. Other referenced figures: Dario (Tario interviews), Intercom founder (Finn), and Sequoia (Amit mentions raising early).
Key claims
90%→100% of code writing by AI; software engineering shifts from coding to directing/reviewing; customer service can be fully agent-driven quickly; pricing shifts from seats/tickets to “tokens plus plus” (work-based + platform margin).
Notable examples
Cloud Code and Cursor ARR growth; Gainsight-style CSM integrations auto-generating QBR decks; Intercom’s Finn; customer service outsourcing (Mexico→Philippines) replaced by internet-based token costs; OpenClaw/agent sandboxing.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Chapters
Tap a time to open that second in VOThe Sinking Ship of SaaS
0:00 to 1:25
Explore the bold claim that SaaS is becoming a dying business.
“You look at like cloud code going from like 0 to 2.5 billion in ARR, cursor going to like 2 billion ARR in just like the last few months.”
The Rise of AI in Software Development
2:37 to 5:10
Discuss the impact of AI on software engineering and the evolving role of human engineers.
“Have you installed this, uh, open claw thing?”
Disruption and Future of SaaS
5:10 to 7:20
Amit shares insights on the future of SaaS and the need for adaptation amidst disruption.
“I think, you know, so what we are seeing is this happened like a year ago, that 90 % of the code was being written by AI.”
Navigating Change in the Software Industry
7:20 to 12:16
Amit discusses strategies for dealing with rapid changes in the software landscape.
“I don't think it will continue to exist unless there is like friction points, you know?”
Innovations and Industry Transformations
13:42 to 14:00
Discussion on how companies are evolving their products to stay relevant against emerging technologies.
“Amit, there was a article posted on X from the founder of Intercom.”
The Shift in SaaS: Embracing Change
14:00 to 15:00
Explore how companies in the SaaS space must adapt or risk obsolescence.
“So they looked at their company and they said, this is not a good direction.”
Margin Compression and Competition
15:00 to 16:00
Understand the inevitability of margin compression in SaaS and its historical context.
“But every industry in all history of capitalism, lots of them started out with obscene profit levels.”
The Future of Software Development
16:00 to 17:00
Discuss the impact of AI on code writing and software development processes.
“That's like in any SaaS company you look at, it's very common to have like 50 % of your revenue going as salaries.”
Integrations and Domain Expertise
17:00 to 18:00
Learn about the importance of integrations and domain expertise in SaaS platforms.
“Or would you just like lend it from somebody else?”
The Rise of AI Agents
18:00 to 19:00
Explore how AI agents will transform business operations and customer interactions.
Show all 31 chapters
Building Trust in AI Solutions
19:00 to 20:00
Discuss the evolving trust dynamics as businesses integrate AI into their processes.
“and this is the next product i want to launch what all will have to happen so you're like okay we'll have to get this, get this much supply.”
The Evolution of Customer Service
20:00 to 21:00
Analyze how AI is set to replace traditional customer service roles.
Cost Efficiency of AI in Customer Support
21:00 to 22:00
Examine the cost implications of AI in customer support versus traditional methods.
“One is like a URL fetch tool, like get me the bunch of URLs that is registered on rich.com.”
Understanding Tokens in AI Work
22:00 to 23:00
Discover how the concept of tokens will redefine work and payment structures in AI.
“And then I think I still have the US manager because that has to roll up to somebody.”
The Competitive Landscape of Customer Support
23:00 to 24:00
Explore the competitive dynamics in the customer support landscape due to AI.
“Like you have to train the thing for like six weeks before it becomes useful, right?”
Future of AI Integrations
24:00 to 25:00
Predict how future integrations will change the face of customer interactions.
“And like, we can't put enough tokens onto that problem to have them do it.”
Versatile AI Work Profiles
28:00 to 29:30
Explore how AI's capabilities can diversify the work profile and redefine customer interactions.
“and they were like, they were rightfully telling me that, hey, there's going to be like objection from the end customers because they want unit of work.”
Shifting Pricing Expectations in SaaS
30:14 to 31:58
Discuss the evolution of pricing models in SaaS and the implications of token-based billing.
“all become so used to thinking and talking in tokens that our pricing expectations for software could actually just wind up being like, now I understand what it costs.”
Convergence of Consumer SaaS Tools
31:59 to 33:58
Examine the potential for convergence among various consumer SaaS tools and their functionalities.
“Amit, do you, do you, something you said is like, so, you know, you could have your agent do all kinds of tasks when it comes to the customer.”
Evolving Roles in Customer Experience
33:59 to 36:16
Explore how customer experience roles will change with AI, focusing on efficiency and management.
“So how, I guess, what are you expecting on the brand side to change?”
Job Displacement and AI in SaaS
36:17 to 39:26
Analyze the impact of AI on job displacement in SaaS and the broader economy.
“And do you see like a head of customer service then becoming more technical or do you think that AI gets so good at just removing the technical needs that they're actually just like a, a, a great finger?”
Future Outlook for SaaS and Consumer Brands
40:17 to 42:00
Discuss the contrasting futures for SaaS and consumer brands in the evolving market.
“Like, is it more true that companies are simply going to go after cost savings?”
The Current Landscape for Consumer Brands
42:00 to 43:19
Discussion on the challenges and opportunities for consumer brands in a competitive market.
“Dude, I think it's the best time ever to be a consumer brand.”
The Future of SaaS and Revenue Quality
43:20 to 44:48
Exploration of revenue quality issues and the competitive landscape in SaaS.
“And that's very much like the other side of this, which is maybe there is actually like a lot more demand for more software because we're just going to need a load more of it.”
Defending RichPanel in a Competitive Market
44:49 to 46:38
Amit discusses strategies for maintaining a competitive edge with RichPanel.
“cohorts look worse than the cohorts from a supplement brand.”
The Vision for Customizable Software Solutions
46:39 to 48:28
Amit shares insights on developing software that adapts to individual business needs.
“The directive I've given to my entire team is our right to win is how fast we play in this new economy and how much value we deliver.”
Preparing E-commerce Brands for the Future
52:05 to 56:00
Discussion on what e-commerce brands should do to adapt and thrive amid industry changes.
“Yeah, I just think it's a great time to be like the challenger SaaS company.”
Understanding AI's Impact on E-commerce
56:00 to 58:04
Learn about the necessity for e-commerce brands to adapt to AI technology and its implications for employment.
“And I like it rolling up to Emily right now.”
Harnessing AI Tools for Competitive Advantage
58:04 to 1:02:10
Discover how businesses can leverage AI tools like Manus and Claude for operational efficiency.
“So who on our team is actually going to dig in and be curious and build with AI?”
Growth Strategies for Rich Panel
1:03:46 to 1:09:10
Explore how Rich Panel plans to grow through innovative solutions like a new returns portal.
“How are you then thinking about, like, that's the defense side.”
Innovative Pricing and Market Needs
1:09:10 to 1:10:00
Understand the challenges of pricing in SaaS and how to address customer needs effectively.
“I think it's smart that you guys are doing that.”
Transcript
Automatic transcript. May contain errors.0:00Sean Frank:You described SaaS as a sinking ship. You look at like cloud code going from like 0 to 2.5 billion in ARR, cursor going to like 2 billion ARR in just like the last few months. I've just never heard a software vendor honestly say it's a dying business. Everyone else is talking about how AI is just going to make their software better and that there's no competition. I think it's irresponsible for any founder to not be going as deep into all of this as possible. You need to get your hand dirty. Like you really need to understand how all this s**t works at a pretty intimate level. 90 % of the code was being written by AI.
0:33Sean Frank:Today it's up to 100%. Customer service is the same thing. It used to be in your warehouse and then it was outsourced to Mexico. That's outsourced to the Philippines. And just because it's the cheapest place to get those tokens, at a certain point, the cheapest place to get those tokens will be on the internet. Half the software engineers, half the lawyer, half the financial analysts will get displaced in the next 12 to 24 months. Half. 50%. You're never going to hire another customer service rep by the end of the year. That is a shocking world to be in. The directive I've given to my entire team is our right to win is how fast we play in this new economy and how much value we deliver.
1:08And, you know, help our customers get that vantage point by getting there quicker. Like 12 months before somebody else does, 6 months before somebody else does.
1:16Sean Frank:Imagine if you're in the 1990s or whatever and you're like, hey, this thing with the internet. They're like, ah, it's boring. I don't care about the internet. Like, dude, you're not going to survive, man. Welcome to the Operators Podcast. My name is Mike Beckham and we are proudly brought to you by Fulfill, Aftersell, Rich Panel, North Theme, Sarah's Analytics, and Postscript. We are a community for entrepreneurs that are building things. And if you want to be a part of this community, you can listen to the podcast, but you can also go sign up for our newsletter. A ton of awesome information in there.
1:46Sean Frank:We also partner with eCommerce Fuel, a forum for you to connect with other entrepreneurs that are building businesses where we can learn from one another. So, without further ado, on to the pod.
2:04Dealing with big box retailers means EDI connections, and that's often a trigger for needing an ERP system. We've been using EDI connections to Costco forever, and the only way that we've really solved that problem to make it seamless is through Fulfill. EDI adds complexity to everything you do and Fulfill solves that complexity with their connections to their systems. You need Fulfill to move from being just a DTC brand to being a true multi-channel brand because big-box retailers are going to require you to connect to their systems using EDI. Let me tell you, it's way easier if you do it with Fulfill.
2:36Um, I got my first thing, Sean, I got a question for you. Have you installed this, uh, open claw thing? Have you got a computer going where you have like your own little, your little dude that can do stuff for you?
2:48Sean Frank:No. So I haven't installed OpenClaw. I've been using a lot of agents. So there's, you know, Claw has one, OpenAI has one. I'm using the Manus one the most, which is owned by Meta. So Meta, if you just ask them, they'll give you a bunch of credits because it gets really expensive fast to have agents do stuff. But I've done a lot of agentic work and I really, really like it. The reason why I haven't done OpenClaw yet is just the security issues. But I've been listening to Jason Calacanis as a podcast this week in startups. They've pivoted just talking about OpenClaw every single day. So I'm definitely diving into what's possible.
3:30Yeah, I've, Amit, I don't know how much you've played with it. I've installed it. I have a MacBook Pro that was not doing anything. So I decided to use that as like a little sort of sandbox machine. And I got this thing sitting on my desk at home and I'm just finding it a riot to like give it tasks and have it do basic things. Like, hey, can you go in and check my email and tell me what the latest three emails are while I'm on my phone, just walking around. I haven't, and this is where I'm curious and I'd like to start because it's pretty remarkable at how quickly that thing grew and how much people are using it.
4:06And I think it's like, it's a good signal for what we're entering into this year. I haven't taken it to the level where like I'm giving it permission to do a lot of editing, writing, like doing anything itself yet, mostly for what Sean said. I'm terrified in some dimension, you know, like for the amount I'm doing with Cloud Code, this thing for some reason freaks me out. So, like, just being honest. So Amit, maybe walk us through, like, you run a software company. The narrative around all this right now can't make you feel good on some dimension, right? And on another, because like you and I, I've walked through what Rich Panel's doing with Agenda Customer Service.
4:51On another, you guys are like charging full speed ahead and you must be optimistic. So give us the thesis. Like where do you think things are now? Where do you think they're going? Just give us Leiland from your perspective because Sean and I are dumb e-commerce operators. We need to get smarter. Absolutely. Absolutely. I think, you know, so what we are seeing is this happened like a year ago, that 90 % of the code was being written by AI. Today, it's up to 100%. And all the software engineering work is limited to now or what the scope of software engineering is, is actually guiding these agents, like having the taste, having the direction and reviewing the work that the agents are finally doing.
5:34So the humans have become the bottleneck. And if you look at Tario's recent interviews, he said that phase one is 90 % SWE work or like code writing. Phase two is 100 % of code writing. I think we are on phase two. Then he said phase three is going to be 90 % of SWE work, which is everything that software engineers do, which is not just limited to writing code, but deciding what to build, talking to customers, talking to users. So he says that will also be done like 90 % and 100 % over the next 12 months. And I certainly see it happening too. So, and I don't see this lightly, but I've said this to the entire team that the core business that we have is sort of like a dying business and we have our feet in both.
6:18so you know one of our foot is inside that sinking ship because i don't think software the way it is today will continue to exist and then we have another feed in this rocket ship where we are also playing part in this you know agent economy and building agents for our customers and even our customers agents so that's how i see happening the sas as we know it is you know going to seize the way exist completely. Yeah. I mean, there's always, there's just, there's going to be more software. Sean, what, actually, Sean, you've got a thought on this, but I got a question for you too, because I think this is interesting.
6:58Sean Frank:I've just never heard a software vendor so honestly say it's a dying business. That's the crazy, that's the crazy thing about the podcast, dude. It's like, you know, everyone else is talking about how AI is just going to make their software better and that there's no competition, but you described SaaS as a sinking ship. So that's probably just worth unpacking. Like, what do you think happens to your business in the next 12 months? I don't think it will continue to exist unless there is like friction points, you know? Like, so all of these things that we discussed that people are slow to adopt and there is this enterprise implementation friction.
7:35So the diffusion is going to be much quicker than anything that we have ever seen. So gone are the days where Salesforce, Of course, the US now have like 20-year modes after they've gotten into Fortune 500 because the pace at which that this technology is like diffusing in organizations, you know, the revenues that you have now is not going to be durable. I would like to share an anecdote with all of you guys. So, do you know Gainsight by any chance, the CSM tool, customer success tool? So, they're both like billions of dollars. What they do is they are for software companies like us and they help us do QBRs and MBRs, like monthly business reviews with our customers.
8:14So you're able to, step one, you actually integrate Gainsight with your product data, okay? So it's a CSM tool, customer success tool. So if you're going to like Rich Panel and talking to your CSM, you're going to like North Beam talking to your CSM or KVO, they all have these CSMs, right? They all have this like CSM tool. What that basically does is it plugs in your product usage data and it tells you like which accounts are at risk, when the renewals are coming and they have these like cadences that they sort of like store there think of it as like crm but for your customer successity right and they have these like risk profiles like who's about to churn why are they about to churn did the usage go down all of that signals right typically it would take you after you buy this tool it would take you like couple of weeks if not more to integrate this with your entire product stack over the last weekend my head of engineer engineering he built this tool and it's it's forget integration it's like completely built out he's a he's a god engineer right but he built it to that level where sean if you go on a qbr with a rich panel uh csm that deck is also automatically generated so like you know there is a dashboard where you see like all the accounts how much each account is paying how many seats do they have when are they renewing all that stuff and the risk profile but you can click like one button and generates like a gamma deck for the review so csm's don't have to spend like three hours preparing for the call they can do it in like three minutes so it's mind-blowing like how amazing not only is it like an alternative to gain side it's it's plus plus at least for our use case so you can clearly like you know draw draw a parallel from here so this guy was able to build it over one weekend to 80 percent level then he took another weekend to perfect it so four days for a very great engineer to build it.
10:0812 months from now, he'll probably take six hours and another 12 months or like six months from then, it will be a non-technical user taking six hours to like build that out. I clearly see it. I don't see any way that's not happening. I could have like smoked or laughed. How do you navigate, Amit? So you're running this company, right rich panel how are you thinking of it from your go-to-market over over the next 12 to 24 months like how does rich panel change yeah so so so the couple of things that has to happen for like all of these companies is like you have to you have to be like not emotionally attached to whatever you've built like i'm 37 years old it took me like all my life to build this thing and you know it's going to like get completely disrupted over like 12 months or 24 months is my dream so i think the mental model has to yeah it's a video game like life is a video game so the levels are changed the whole thing and then you can die again in the next stage so that's the mental model and i say this nonchalantly but i've gone through that entire stages of uh you know what the is happening no this is not going to happen and so all of that it's like the stages of grief yeah all of those five stages of grief i've been through it so that's why i'm seeing it calmly that we have this time we have this opportunity uh and if there are other companies like if you look at like cloud code going from like zero to 2.5 billion in arr cursor going to like 2 billion arr in just like the last few months they went from like one to two in like four months so my thought is like hey why can't we be playing in that economy because those guys are new.
11:51Those are also like software engineers. They're pretty talented and so are we. So why can't we play in that economy? And which is exactly the direction we took, you know, 18 months ago. And I think we have very well listened to play in that new space. Yeah.
12:05Sean Frank:Yeah. And this is where company structure, I think is so important. Like you guys haven't raised venture capital, so you can sit here and be like, yeah, so the business model we're currently in is going to die in the next 12 months. We have to launch a whole new business model. And here's the reason why it's going to work. But, you know, if you had$10 million in venture capital and you just saw software multiples compress, I think there's way more doomerism. It seems like you're very optimistic that your business model might be destroyed. So are you telling me to short software stocks? Not investment advice, Sean.
12:41Revenue has already shorted them. There are like these software companies I used to look up to are trading at like 2x the revenue multiple. Like it's, I think people have already done it. So I don't know what role they will pay going forward.
13:16to end, unify results with your MTA and your MMM. This is a lot of letters, but if you know, you know, and you can start to cut what doesn't work and you can scale what works and you can do this all with confidence. This is why this is such an incredible ad to Northbeam. Northbeam's incrementality measures what results marketing is actually generating, not just what they're claiming credit for. As a CEO, that's like music to my ears. Sign up now and you can lock in 50 % off unlimited tests for the year. Amit, there was a article posted on X from the founder of Intercom. I don't know if you saw this one.
13:50I'll send it to you if you haven't. But it's a very cool example of how they did the same thing or are doing the same thing you're talking about, which is they built the thing that ate their old thing, right? So they looked at their company and they said, this is not a good direction. we need to stand up the thing that kills ourselves right and then that that product i think it's called finn yeah is that like a hundred million in arr already and uh it's basically eating the old thing now like they birth the new child the old child which yeah i mean i think the i guess i got a couple thoughts on this i think that that feels like that's exactly what technology has always done, right?
14:37Usually it's just somebody else coming for your lunch. I also, I don't know if you guys have thought about this, but like, it's sort of, to me, seems strange that we expected SaaS to never commoditize. So like every industry in history has had margin compression. And the only argument I've ever gotten from SaaS people is that our margins are perfect. They can't compress. I'm like, are you sure about that? But every industry in all history of capitalism, lots of them started out with obscene profit levels. And eventually competition arrives and we commoditize. That's just what happens. So is this that moment?
15:23Is it because? And then Amit, is this because AI can write code? So like we no longer have a supply side problem of code, which is what I always understood is the big issue. Yeah. So I think writing code has not been a challenge for the longest time because there are like too many capable engineers that are capable of writing code. And the SAS margin compression has happened like years ago. That's my belief. It's just that we do modeling because, you know, like if you have what's our biggest PNL item, it's salaries, right? And within that also it's developer salaries. That's like in any SaaS company you look at, it's very common to have like 50 % of your revenue going as salaries.
16:08But the way that it is itemized in the P &L, it doesn't show up because a lot of that shows like, oh, this is R &D budget. Like we can decide to not spend it and take that money home. And while that might have been true for Salesforce in 2000, it's no longer true. You have to build X number of integrations. You have to build X number of features just to stay relevant. So I don't think that is R &D budget. That's more like COG's budget. But it's just been wrongly categorized in all SaaS P &Ls for ages. So I know it's the thing that's quite hilarious. Yeah.
16:43Sean Frank:So what do you think the future of something like a Clavio is? You know, Clavio has a lot of integrations. Clavio is like it's it's it's consumption based right they charge me percent if you break it down that way so like are they going to be eaten as well or is this just coming to like the legacy bloated players I'd love to hear your thoughts I think we'll have to like do some simulation in our heads to see like how this is headed right so even if the cost of writing software goes to zero So, you know, let's say, would you still like write it yourself? Or would you just like lend it from somebody else?
17:25Like would you rent it from somebody else? So the arguments can be made for both. If you are renting it from somebody else, you get like best practices on all these 55 integrations already built in that you will probably use. You will have like security baited because, you know, these are like a very talented software engineers. they know like you know you don't like white code security vulnerability and you make sure so the trust factor is going to be huge in the new economy where you get all of this SOC 2 HIPAA compliance from these vendors and the agents are running within that trusted environment.
17:57The second is going to be like this domain expertise so even if you look at like emails the writing code is okay but then all the code is like you know saving all the engagement rates what happened to each email and then doing analytics on top of it so that domain expertise will always remain and the third piece is is the continuous evolvement so even if you guys decide to like replace the 60 sass apps you'll have to like have an it team of like a few engineers just to maintain them right so you may decide like hey it's not worth doing it i will still pay for that sass but it better be like a huge upgrade to the traditional ui where my agent is capable of using your software agent to do its task so it'll be like you will have your own agent that will that you will give objectives to and this agent is then going to rent other agents so instead of people buying software built by other people it'll be like your agent will interact with our agent to get a particular task done so your your agent will be like yeah i run an e-commerce business and this is the next product i want to launch what all will have to happen so you're like okay we'll have to get this, get this much supply.
19:08We'll have to do this testing. We'll have to run these metads. We'll have to like ramp up our customer service team. So they will go to like Rich Panels, AI agent and say like, yeah, I want this much capacity. That's, that's what I see is going to happen. Amit, are we going to see more and more versions of all these companies? So like, let's use, let's just stay on the Klaviyo example. They kind of own e-commerce and email, right there's been one challenger in the last several years um yeah is is because the cost of writing code is effectively zero or going to zero do we see more competition in the space like there's been certain certain ones that just haven't had any competition like does that stop now do we get more you see the uh code is not the only issue right like to be if i'm being like completely honest it's not like rich man is successful because we wrote some code that others couldn't do it's a combination of other things those other things are like hey there has to be like a distribution mode uh there has to be like brand recognition so those things don't go away even in that new economy uh one thing that is also going to change is probably how this trust is going to be built so like traditionally human beings were trusting based on like what they see on x or Twitter or like what their peers are talking about you'll also have to earn the agent's trust next time onwards so like I just integrated like three tools yesterday which my agent recommended that it wants so these tools were like web search tool like I can go and like tell my agent go set up a demo for Pella case or you know Ridge and it needs to like go and like scrape your website to read your FAQs, read your policies, and then set up the demo based on that, right?
20:58So I had to buy like two tools. One is like a URL fetch tool, like get me the bunch of URLs that is registered on rich.com. And the second is web scraping tool. Both of them are like other APIs. So you'll have to become discoverable for your agents because your agents will also be doing a lot of buying for you.
21:16Sean Frank:Yeah, so the agent marketplace is going to be a definite killer service that somebody launches. It's the app store for agents, right? And it's going to have its own reviews and everything else. You know, it is interesting to be at this, you know, crossroads in time. I think maybe you said this on a call we were on. You said, you're never going to hire another customer service rep by the end of the year, right? And that is a shocking world to be in, right? So, you know, we have two full-time people in the U.S. who manage our customer service program. And then there's a bunch of overseas contractors who actually answer the tickets or whatever.
21:53Sean Frank:And you totally see how an AI agent will deliver better service, faster, with way more detail than an overseas contractor. And we'll do it infinitely. We'll do it 24 hours a day or whatever. And then I think I still have the US manager because that has to roll up to somebody. but the people actually answering tickets, I think that will be the first massive layoffs of the AI era. So what are your thoughts on that? Is it just inevitable that agents are going to answer every ticket? Yeah, that's going to happen very quickly. The only reason that you wouldn't have access to that technology is because you didn't act fast or you had other priorities as a founder or a CEO.
22:35But if you really make that a priority, like, hey, I want to get this done, that's going to be possible in the next six months if it's not already possible today. The only thing that is stopping from today is like the amount of setup it takes to like train the agents. You know, like you took Matt, the example of Intercom or there are like other companies. I believe that the way that the current setup is done is pretty manual. Like you have to train the thing for like six weeks before it becomes useful, right? You use your, you know, cloud code. It takes a lot of setup before it becomes useful.
23:11but that's where the SaaS vendors have to like step in and say like hey you just plug it in give it your website give it your pass tickets and it will do all the setup come to you only for reviews that will also become like an agent what's the uh have you guys boiled this down to a cost per ticket I know Sean I think we've talked about this in the past like a couple years ago but like uh using overseas labor there's like a certain cost per ticket we can get to And then ultimately, if AI is doing it, we're talking about tokens. Compute. Right? Got it. Got it.
23:44Sean Frank:Right. So, oh, Matt, I have a very interesting thought on this. Yeah, go ahead. So I think all human work is tokens. This podcast could be translated to tokens. So tokens is just a currency, right? And, you know, a plumber coming to work on your house is infinite tokens because you don't have robots yet. You know what I mean? And like, we can't put enough tokens onto that problem to have them do it. But this has been true for 20 or 50 years or whatever, right? So IT used to be in your building, in your office in Manhattan, and then it moved to an offsite location and then it moved to India, right?
24:22Sean Frank:And it's because the cheapest way to get those tokens done. So like the tokens is just the work in this case. The cheapest way to get that done was to put it in India, right? Customer service is the same thing. It used to be in your warehouse and then it was outsourced to Mexico that's outsourced to the Philippines. And it's just because it's the cheapest place to get those tokens. And at a certain point, the cheapest place to get those tokens will be on the internet. But the problem is just like if you're trying to use Claude 4.6 Opus to do your customer service tickets, the cost per ticket will be a dollar, right?
24:52Sean Frank:But if you're trying to use Kimi, the open source model from China, the cost per ticket is going to be about four cents. So it's really just how fast you can do the actual integration, what Amit was talking about, actually set up these agents and train them. And when do you actually want to go all in on doing it, right? Right now, you can get cheaper cost per tokens if you go through open source models to do your customer service. But if you try to do it off, like, you know, Sienna is a competitor. Their cost per ticket is over a dollar. Yeah, it's too much. Yeah. If you're doing it in the Philippines right now, your cost per ticket is 25 cents.
25:28Sean Frank:So there's at a certain point where humans are competitive in the labor market, but we just know that the way things are going, it could be six months, could be 18 months. Eventually, open source will match whatever cost per ticket you're doing right now. So what am I wrong about there, Amit? 100 % right, Sean. I agree. That's how we also view things. The unit of work is like tokens now. That's the new world that we live in. So 100 % agreed on that point. Based on the initial analysis that we have done, even if you're doing like customer service in Philippines, the cost per ticket comes down to like, for most brands comes down to like$2 or$3.
26:06That's all blended in the labor costs, the number of tickets they do.
26:11Sean Frank:Those people should sign up for Rich Panel because my cost per ticket is like a quarter, bro. It varies from like brand to brand. and you mentioned like you know these other players are charging like a dollar per ticket to resolve with AI there are a couple of problems even for brands that are at three dollars they would not they're not adopting to that AI as quickly it's because saving two dollars per ticket might mean like saving five thousand dollars or ten thousand dollars of Portland line per month but you may have like higher other leverage tasks to focus on instead of spending like six weeks training this agent and even after you train this agent today's uh players are not able to like completely replace the cxt you still need like some cx members here and there they automate some the escalations are still going they're doing like the tier one ticket so when it like makes when these two things hit like when it the six weeks time to set up becomes maybe 30 minutes or one hour and when it is not replacing half the team but the entire team people are going to like flock towards it because it will just become like a no brainer so that's the next piece I see happening in the next 6 months, that's pertaining to customer service and I was talking to Sequoia by the way Sean we have raised venture from Sequoia like in the very early days And we said no for the remaining stuff for the subsequent rounds because we became profitable.
Read the full transcript
27:43So I was just talking to them and we had like a huge debate over like how to charge for this. So my inclination was that we charge based on tokens plus plus. So if the cost per token for one ticket was like 25 cents, we charge like one third on top of it for providing the platform on which it runs. That was my inclination. and they were like, they were rightfully telling me that, hey, there's going to be like objection from the end customers because they want unit of work. They want like cost per ticket. And the reason I want to go per token is because the work profile is going to become very versatile.
28:19So you will be able to tell your customer support agent that at the end of each conversation, assess the sentiment and if the sentiment is good, then upsell this product. It's not a ticket anymore. It's an upsell conversation. You will tell your AI agent like, hey at the end of each conversation i wanted to go to memory.md and tell me what was interesting about it and then you know every week or every day in the morning send me a slack message telling me like what was interesting about it now this is not cost per ticket this is plus plus work right so and the thing is going to become so versatile that you do like ask it to do like surveys you do like ask it to like you know reach out to all the members that haven't purchased in the last three or six months with like a personalized message so the work profile is going to be very you know diversified so just like human beings the intelligent ones are being built like lawyers and are being built per hour ais will be built on token basis like how much work do you want me to do because i can do anything like as a lawyer i can i can help you get out of a ticket but i can also defend you in like a case or uh like a mna representation so i I don't know what that unit of work is going to be.
29:31Sean Frank:Sean here to tell you about Saris Analytics and Saris Pulse. Ridge is profitable every single day. And we've taken that super seriously since we built this business. We track contribution margin by day. We look at the SKUs we sell every single day. And we have to do this manually up until Saris Analytics came out. We take all of our SKU level data. We build it into the data warehouse. Everything that goes into making a true P &L, I get on a day-to-day basis. Sarah's Pulse gives you clarity so your CLO and your CFO and your CMO start speaking the same language. Contribution margin shifts teams away from hoping profits survive the season to managing them in real time.
30:06Sean Frank:Book a walkthrough with the Sarah's Pulse team today. Click the link in the description and thank you Sarah's for bringing you this show. It's an interesting idea that we all become so used to thinking and talking in tokens that our pricing expectations for software could actually just wind up being like, now I understand what it costs. Whereas in the past, I have no idea what the cost basis of Ridge Panel is to run that company. It was really hard for me to say, I want to negotiate based off of your cost. I have no clue. But everything's in tokens. I'm kind of like, it's pretty transparent now.
30:45Sean Frank:It's going back to the media buying model. It's like, I'm going to charge you 10 % of media buying. or whatever. It's like, look, the more media buying I do, the more your company is successful. It's alliance incentives that I want you to spend more money. Rightfully, there's a lot of problems with that. And we've thought about if you should charge percentage of ad spend or whatever. But the reality is the seat-based model's dead because I'm going to have less employees. And you're going to charge me for my agency? It just doesn't make any sense. And then I totally get where Sequoia is coming from.
31:17Sean Frank:As a customer, I would love for you to be like, yeah, I want a dollar per ticket. I understand that. But that's still, I think, old school thinking. Because like what you just said, what if the thing could do my taxes for me too? That's not a ticket anymore. You know what I mean? In your example, the lawyer, yes, it could be a lawyer. It could also be a customer service person. If you're on the Supreme Court, you're probably pretty good at do what others too, right? So I do just think they're probably thinking in a more limited stroke. I would love the idea of just all work being done on tokens plus, right?
31:51Sean Frank:Just like a commission basically. Yeah. It's like, it's cost plus pricing. It's just a new form of cost plus pricing, which is a very big break for the SaaS industry. Amit, do you, do you, something you said is like, so, you know, you could have your agent do all kinds of tasks when it comes to the customer. They can reach out and do a survey. They could reach out and say, Hey, I saw you bought this? How about this? Like all that stuff. Do you see a convergence then in consumer SaaS? Right now I have lots of SaaS that does all these things. I have post-purchase surveys. I have like, you know, cross-sell, up-sell, post-lake, all this is everywhere in my company.
32:31All of it is kind of, if I could group it all, all of it is talking to the customer in some way. Like there's some kind of conversation happening with the customer. So we have like lots of different software brokering messages to our customers for different workflows. Do you see all of this converging? Is there going to be like one master broker that's working with all these specialized tools? Or do you just sort of see like, no, it's just going to get so easy to do all of these tasks with such intelligence that, you know, Rich Panel will be in competition with other people who do messaging of other forms right now.
33:07Yeah, yeah, it's pretty fluid, Matt. So it's difficult for me to say with a lot of confidence. I do see like a lot of conversions happening. But by that same token, I also see that the capabilities will level up. So if you look at an e-com brand from like 2015 versus an e-com brand of 2026, there's a massive difference in the amount of things you have to do to earn that same dollar. Because yes, you guys got like Shopify, you guys got Kave, you got Richpanel, got you know like 50 other tools but that leveled up your capabilities like you didn't stop there because that's what we will do like in a competitive capitalist market when you have these capabilities available you'll just do more so by one logic all of the things that you're doing now there is room for like consolidation but because of the capabilities that will be available to you and to your competition you'll just be expected to do so much more to compete in this deal market piece.
34:08Okay. So how, I guess, what are you expecting on the brand side to change? So we're talking a lot about like RichPanel and SaaS and everything. So one theory is that we stop hiring like net new people to do customer service, right? CX as we know it today. But then let's talk about like the people that are left, right? So like when you have all all of this work being done by agents, what do you think our teams look like? Like what is the head of, like Sean's head of customer experience that he has stateside, what does that person's job evolve to? Yeah, so let's look at it like a couple of ways.
34:51I don't know if you guys are following like what's happening to Figma's stock. You will see like people are just skipping through, like in my company, people are just skipping through the Figma step. So first it used to take like a few weeks of design reviews like you know you come up you you talk to like 15 customers you come up with a use case you validate it then you do 50 more interviews you you know propose a feature then you do like figma design reviews this entire process is like three or four weeks now people are just skipping past that stage where you ask claude or gpd to be your customer and you talk to it like hey pretend you're this pretend you're that and you do those conversations again and again it gives you the insight and then you skip through the whole design phase of figma reviews and all of that you just have like a skill built in like a front-end skill that builds and mocks the whole thing and then it also builds the thing to take it to production so same thing will happen for the head of cx where they don't have to do all of this mocks and things it's an agent that decides what experiment to run today goes and like builds the landing page then does like 13 types of test like is it compatible on mobile it's compatible with all widgets that's the orchestration that's as people will have to provide because it's very easy to like build a landing page using ai tools even today so where we will stand is giving you the infrastructure and the protection that once you make that change it's being tested against like all ios safari chrome all those devices and we ensure that whatever you wipe coded did not break the thing then the second step you can think from there is it will run the experiment by exposing it to like 10 % of your audience and it will pick the winner and then pick the next experiment so it's going to be like continuously evolving right so that's that's how I see it happening because you know we as humans we're not going to stop just because I was able to do this thing so well because your competitor is going to say like oh Matt you're doing some agentic thing to like do this once a week I'm going to do it like six times a week I'm going to change my landing pages I'm going to run like 60 experiments at the same thing.
36:57So that's, that's what I see happening. And do you see like a head of customer service then becoming more technical or do you think that AI gets so good at just removing the technical needs that they're actually just like a, a, a great finger? Yeah. I think your head of customer, I think your head of customer service
37:18Sean Frank:just becomes your head of customer. And like, it is one person who manages 50 agents or one master agent or however you want to talk about it. And they'll manage customer service. They'll manage cross-sells, up-sells, look up with more better flows, look up with customer events. I think we end up spending more time taking care of our customers in this new world. And let's talk about the job displacement part because AI is going to be very bad for SaaS companies in the same way that cars were very bad for stable boys. There used to be a lot of stable boys. Now there's no stable boys and cars are to blame.
37:54Sean Frank:And that's going to happen to SaaS companies. I will take the argument that U.S. jobs are less at risk because if it hasn't been outsourced already, why would it be outsourced now? I'm not saying in four years, like, you know, U.S. jobs aren't at risk, but the first people who are going to lose their jobs are call center employees, are people answering tickets, are, you know, IT. So it's really going to hit the outsource economy the hardest i think you know if you're if you're somewhere which is like a va placement tool like and you're in you're placing vas who answer people's emails i'm like well no claude can do that right now so um anyway that's just to think about the job displacement i think the u.s won't be hit that hard yeah so so the job will get displaced irrespective like Like Dario himself said it, like half the software engineers, half the lawyer, half the financial analysts will get displaced in the next 12 to 24 months.
38:58Half, 50%. So the economy is different. You know, like if I'm running a law firm and I get to like remove 50 % of my lawyers or just to like 2x my throughput, I benefit. But the lawyers as a whole do not benefit. So jobs are potent. Like in both the cases, it's going to suffer. But when you're outsourcing, that company was probably only relying on outsourcing as a revenue. So there, the jobs will also be gone and the economy will also be disrupted. Both things will happen.
39:31Sean Frank:If you're scaling an e-commerce brand today, ads alone aren't enough. Aftersell focuses on the one moment that every brand already owns after checkout and turns the post-purchase moment into more profit. Monetize every order with post-purchase offers and thank you page experiences without disrupting checkout or hurting conversion. Enterprise-grade tech used by Gap, Ticketmaster, Macy's, and Target now driving results for brands like True Classic, Hexclad, Ridge, and Jones Road. I would know. This is the reason I ended up buying three pans from Hexclad instead of two. Aftersell has already generated over$1 billion in additional revenue for e-commerce brands, revenue that doesn't require more traffic or higher cap.
40:11Sean Frank:So check out AfterSell and tell them that the operator sent you. I guess, but hang on, can I push back on this for both of you guys? Like, is it more true that companies are simply going to go after cost savings? Or do you think that there's going to be companies that go after more land? Like, will they, like if I have a thousand lawyers right now, I could take two approaches. I could say, I want to go get 10 times the business with my thousand lawyers because I can now. or I could say like, I'm happy with my business at the size it is and I can just cut my costs by half. Like what's more likely to be true in the U.S.?
40:47Sean Frank:I think the best firms will 10X their business because, you know, the problem with lawyers is that it's very hard to become an expert. It's very expensive. It takes years and years of training. And if you could just have infinite lawyers overnight, you're going to file more lawsuits. This is what's going to happen. it. You're going to clog the whole system. But so if everyone gets 10x more productive, I just don't know if the world can tolerate 10x more legal output, right? Or 1 ,000x more legal output. I think at some point, it's going to both, the best firms will take over more land, but half the firms will just go bust.
41:25Sean Frank:So it's going to be - What do you think it would have been a consumer, Sean? How do consumer brands react here? like companies that have, like us? What do you think we do? I think this hurts SaaS and this hurts service businesses. That's what we just talked about, right? So SaaS is screwed because now everyone can make a SaaS product and there's market compression and service businesses are screwed because you make an arbitrage on people and the idea of people being workers has just broken because your computer is going to be workers and you're going to charge people on tokens. And so, you know, if you're in services, if it's account needs floors or whatever, you're screwed.
42:01Sean Frank:Dude, I think it's the best time ever to be a consumer brand. It's like, yes, there might be disruption in the labor market, right? But it'd be very difficult to, even in two years, to have the world's best AI agent and be like, make me a gummy brand that does$10 million a year in revenue, right? If you said that as a prompt, it's just, this is your cascading world of events that still intersect with the real world. because they're going to say, make me a gummy brand. They're going to have to find the people who make the gummies. You know what I mean? They're going to have to go through that entire world.
42:35Sean Frank:And then also earlier, Amit said this, distribution is still a moat. Like you have to get customers, right? And it's just going to be very difficult to break into consumer, I think. But I think it's a great time to be a consumer brand. So bearish on SaaS, bearish on services, super stoked to be a consumer. I don't, I, for some reason, I guys, I get really bullish on SaaS and bullish on services because I just think, I mean, this might be the optimist in me, but there, there would, in theory, Sean, there could actually be demand for 10X more legal services if the technology from a diffusion perspective is like evenly spread.
43:18So like if all, like if your landscaper can get more efficient and take on more business or if you're whatever, like all companies in all sectors, like if consumer can get bigger, if everything can expand, then there's demand for more of everything. Right. And that's very much like the other side of this, which is maybe there is actually like a lot more demand for more software because we're just going to need a load more of it. Oh, yeah.
43:43Sean Frank:So I say bearish. There's going to be way more software. There's going to be like everything will have software everywhere, right? But I guess when I say bearish, you won't have a company that does 100 million ARR be worth$10 billion or whatever, right? Yeah, the multiple compression is real. Do you think that the multiple compression is because of quality of revenue then? It's not that there's more software or less software. It's just more competition. I think it's going to be quality of revenue. And also it's going to be very hard to get those big, those big outcomes, right? We talk about Salesforce.
44:19Sean Frank:Salesforce is a multi-billion dollar a year SaaS company. And I think it goes to 200 million. I think that's just like the upper threshold. I think eventually people will replace Salesforce and it's going to take longer than two years. Like everyone's been talking about the death of Salesforce just because you have to get Walmart to get off Salesforce or whatever, and that's going to take five years or whatever, but they will go towards a gentic solution, I think. But yeah, so we've already seen this where there's companies like Lovable who get to 100 million ARR, and I think those cohorts look worse than the cohorts from a supplement brand.
45:00Sean Frank:I think those people are churning like 50 % every single month. I think consumer SaaS is just like a very, very blood blood in the water business. Revenue is no longer durable as it used to be because the barrier to entry is so low that the thing that made you so successful is also the thing that makes you so susceptible to disruption. So whatever you're talking about, like lovable and replete, that is 100 % true. It's not as durable. so amit how do you think about uh defending your business then like rich panel right now are you is the idea like we're just going to always stay at the edge of like building the best tools and the best agents for our customers and that that well that we're going to have to just like somehow triple down on product even though product i guess this is the thing that I'm finding too, like, again, as a power user of cloud code right now, I can build lots of stuff.
46:03But there's a big difference between the crap I'm building and, like, something I could sell to someone else. Like, there's a huge gap between those two things. Like, I can build crap for my team to use. But, like, it's not sellable software. And there's a lot more that's going to go into, like, how to sell it. So I guess until agents are buying software, humans are totally out of the loop, but I don't know when that is. So how are you thinking about this, Amit? How do you build RichPanel going forward? Tell us why we should continue to pay you. I think you answered it, Matt. The directive I've given to my entire team is our right to win is how fast we play in this new economy and how much value we deliver.
46:49So we have to be like always on the edge of whatever is new and, you know, help our customers get that vantage point by getting there quicker. Like 12 months before somebody else does, 6 months before somebody else does and continue to do that so that our customers have that vantage point or that advantage over their peers, which is why they should come to us. Like, hey, there's somebody who's thinking about CX all the time, you know, taking all the agent capabilities and giving me that advantage over what my peers have. But six months and 12 months before everyone gets there, because by the time everyone gets there, it's commoditized.
47:29It's no longer an advantage. So that's the role us playing. Yeah. Go ahead, Sean. I was going to say, one thing I'm very bullish on is just people being AI implementers.
47:44Sean Frank:Like, the reason I'm going to pay Rich Panel is they're closer to the source than I am. One, they're engineers. One, they're coders. Like, two, they're coders. And, you know, they are very incentivized to make sure that they are on the cutting edge of working AI tools for customer service. So it's like, look, the fee is cheap enough. It's like, look, to kind of outsource that thinking to them, I think it makes a ton of sense. And if you're a service business right now and you can make a ton of money just setting up OpenClaw for people to be like, Hey, I know exactly how e-commerce agency or e-commerce businesses should use OpenClaw.
48:16Sean Frank:I'm going to set it up for you. So I think there's, there's still a lot of money to be made right now and just being the expert in the hyper-specialized field of AI for play. So I'm, I'm a big supporter of that. Uh, Amit, this might be a super technical, like weird weird do you do you think that rich panel do you think you're going to build rich panel in such a way that like every single person every single customer using it is ultimately going to have their own version of that piece of software so like will it change itself based on our needs as a business like my needs are there's a lot of common stuff between sean and myself, right?
48:54As businesses. But then there's a lot of weird things that I have that Sean might not. And then vice versa, right? So are we just going to get to a point where I'm like, Rich Panel, instead of going and talking to your customer team, asking for a feature request, am I just going to tell Rich Panel, like, hey, I'd really love it if you could change how we handle this. And it just writes the missing software for us. Yes, that's the vision. That's our vision entirely and I see it being completely possible so the loop is going to be like first there is a support agent which is able to like resolve 99 % of your tickets that happens in the next six months simultaneously we'll give an agent which manages the knowledge base for this support agent which we are calling like CX manager like that's the head of your customer service department or the CX manager that is responsible for building the guidance for these people.
49:49So the reason we build a second agent is because this is the person you talk to and you eliminate that six week setup time because this agent knows how to manage other agents. So that's the CX manager. So this is stage one and stage two is you will be telling the CX manager, hey, why don't you go and like add this SOP why don't you you know build this integration out for me and stage three is where it has like a open claw style heartbeat where it comes to you every day and says that hey Matt these there are like 15 customers that are upset about this particular skew or variant we should consider sunsetting it right because it was able to like track that it will come to you and say like this is how the customers are reacting to something yesterday and we still had 230 escalations last week because you had to do some operations in ship bob or fulfill erp so can i go and build that integration so like yeah go ahead so when it goes it spends like a bunch of tokens spends like couple hundred dollars makes the fulfill erp integration with rich panel you already have a fulfill integration but this is just hypothetical so you just go like build your own stuff and says that I will no longer escalate this, you know, shipment delay or like reshipment queries to you because you were editing some button on fulfill.
51:10Now I have that capability to do it. So it becomes like this proactive person that's coming to you next. That's how I see it happening.
51:21Every SaaS company says they are AI powered, but very few can explain what it actually does for the revenue of my brand. This is why PostScript's approach stood out to us. They don't just build AI for demos or buzzwords. They built it to drive real incremental revenue. PostScript's AI called Shopper, it shows up inside of SMS at moments with real buyer intent when shoppers are likely asking questions, hesitating, maybe even about to drop off. Shopper can answer product questions instantly, answer questions about fit, availability, recommendations, order issues, the kinds of stuff that people usually bounce for.
51:53This means more conversions, higher AOV, less lost demand so you are driving more revenue and doing it more efficiently. Check out Shopper from Postscript. We use it at Pila, which is why I am telling you to check it out. Yeah, I just think it's a great time to be like the challenger SaaS company.
52:12Sean Frank:You know what I mean? We left customer, which was bought by Meta and it totally screwed up. They did not know, they did not have a customer vision or whatever. And they're lost in the sauce. We went to Rich Panel just because it was lightweight, it was new, it was cheaper, and they're delivering better services. I think it's just a great time to be that challenger because there's going to be so much innovation. And if you're already coming from a place of being a challenger, you have to embrace it. It wouldn't surprise me if you said 99 % of my tickets are done by the support agent in six months.
52:47Sean Frank:I mean, that is crazy if that happens, going to Q4 and I I don't have any more customer service reps basically. Like that'd be, that'd be, that'd be nuts. But, um, I do think it could happen. I think we've seen the technology that that's what's going on. I believe it's happening. Like I would be, I would be shocked if it doesn't happen. Yeah. Which is like, yeah. If you're not the ones to do it, I'll also say like, we have a lot of like steps to like get there, but it will be, I will be embarrassed if we don't do it. Uh, but I believe we are competitive enough to, be the ones that get you there.
53:23But yeah, it's happening.
53:25Sean Frank:I think this is, this will be the wake up call to the whole industry. I think it's like, look, if we have a Q4 where there's one human in the loop, that's crazy. It's like, then it's like, you know, there's a lot of just like people concerned about what happens to accountants or what happens to lawyers or what happens to whatever in this world of econ, what does it do to the economy? We're going to see all customer service reps be replaced in the next 12 months. And that is a scary, scary time. So that'll be the big wake-up call to the industry. But what, okay, let's end the pot on this. 10 minutes.
54:00Sean Frank:What do we think e-commerce brands should be doing right now? Like, should they waste their time to set up OpenClaw? Should they, like, how can they actually, like, be on the cutting edge and not just have anxiety for where the future's going? Matt, what are you doing? And how can you advise everybody? I think it's irresponsible for any founder to not be going as deep into all of this as possible. I understand that we can hire people. Dude, I am like, Sean, how many years have you known me? Like I am such a like delegate, build team, like all that, right? I am that guy. And I am right now telling everybody that I talk to, you need to get your hand dirty.
54:40Like you really need to understand how all this works at a pretty intimate level. I'm not saying that you got to go build everything yourself, but this is not a moment to like just assign an ai person and then hope that they do it like i think every operator and consumer right now needs to be going so deep in this uh and then yes hire out delegate like do all that crap but man you need to know how this works
55:07Sean Frank:i think this is like absolutely critical yeah and i i want to be kind of a a realist in this whole thing. It's like, I think your teams will get 5 % smaller every year going forward. I think you'll still hire highly specialized people, but just in general, I think your teams will get smaller. Um, you know, not to the point where there's no, maybe, maybe, maybe in 10 years, nobody works at Ridge, but like, you know, we were talking about this the other night. I'm like, look, I still need a core group of 20 to 25 people to continue to grow my business. Uh, Yes. And I like, here's an example. We have an amazing inventory manager.
55:44Sean Frank:I think she's going to have a job here forever because inventory has to roll to somebody. Even if AI is doing that in all the work, if it's doing all the projections, if it is doing everything, because it's so expensive, I just want somebody to click the button. Yes, we're ordering this. And it's like, if it's not going to be her, it has to be my CRO Andres or has to be me. And I like it rolling up to Emily right now. So it's just, you know, I think it is the people who are still employed are running entire functions in an organization. So anyway, I, you know, it's very easy to, to come on here and be like, every job is going to be replaced and everything's going to go out of business.
56:21Sean Frank:It's all going to blow up. I, I, I completely agree that CX reps are going to be the, the first wave in this, in this AI war, but yeah, Amit, what should people be doing right now? If you're an e-commerce brand, what should you be thinking about? Yeah. I second what both of you guys said. I strongly believe that all operators, e-com brand owners, need to be time boxing two to three hours, if not more, on a weekly basis to be spending time and exploring these tools and understand what is now capable. So, like, in both of your cases, you're probably not logging into Meta every day and, like, creating ad campaigns, but you do know how to do it, right?
57:02Like you do know what the capabilities are. You may not be like going into Shopify and building like a landing page yourself, but you do know what the capabilities are. And this is a new tech that you're working with. So it's very important to understand its edges, its capabilities. So you will have to like time box at like two or three hours. And unlike Meta and Shopify and all of these other tools or technologies, which are not changing that much, this thing is changing and like leveling itself up every three to six months it's a step change it's not even like a minor change so things that you were not able to do six months back you probably would be able to do those things now so you'll have to to stay competitive you'll have to understand what you're working with like one or two levels down and then encourage everyone in the company to be harnessing and using these tools to their advantage but yeah i think I think there's going to be a ton of work in consumer.
57:59I mean, I'm just going to talk consumer, not every other industry, but I think it's consistent in all industries. I think we all have a lot of work right now on the people side in identifying the people that can and want to participate in this new way of work. So who on our team is actually going to dig in and be curious and build with AI? Because every job is going to have to. Like my head of customer service, they're going to have to know how to use all this stuff, right? And if they don't, I'm going to have to get somebody who does. So I think we're just going to see, like, Sean, I'm with you.
58:34I don't think this is like some dramatic, I'm going to, like, there's just, I need a certain number of people. Like there's no dramatic, like, people change coming. But I do see over every year going forward, a pretty serious upgrade and replacement process. Like there's going to be people that just can't keep up. They just can't do this. this is just beyond them particularly the next two years while there's still like a bit of a technical hurdle to some of this like i showed uh my wife open claw and she's just like yeah i'm not doing that yeah it's like not even close is she kind of anywhere near that thing um and i think that that's pretty that's a reasonable response right but yeah i think i think that's going to be the work sean is there's just going to be a ton of people work for all of us and i think
59:20Sean Frank:everybody listening to this is familiar with agents and if you're not use manis so manis is on beta it is it is a baby agent and and whatever my baby agent it's a browser only agent so you don't have to download anything you don't have to set anything up it has security to this to the quality and standards of meta so like i'm comfortable giving manis access to my email to my slack to everything and it will do stuff for you like i've i i've shared this a couple and I'll share this, but like, um, I needed to build an image database of celebrities. I got a list of names of celebrities we're going to put in a, in a commercial.
59:55Sean Frank:I didn't know who they are and I wanted to see their faces. So I just gave the list of names to Manus and it built me a database with their, uh, you know, link to their Instagram and their followers and a bio and their last movie. And like all of this data, they're like, it would have, it would have took my assistant, a human assistant, three hours to do or whatever, right. Maybe longer, five, it could does it so it can do that work for you right so get started there if you don't want to download open claw and do do all of the hobbyist nerd just start using madness today and then and then for your for your teams get everyone access to claude and then if they don't know what to do have them ask claude what to do claude will tell them how to get started how to do their jobs it is the first technology ever that can teach you in layman's terms to exactly where you are it's why education is going to be upended, right?
1:00:47Sean Frank:Like it meets you exactly where you are and coaches you and it never gets bored. So - Do you know what's incredible, Sean, with that last statement in my company, at least? I don't know if it's been the same for you guys, but probably not you because you have so many technical people. But in our company, people are used to asking someone technical how to do something technical. So I have developers, right? And when we need to do something technical, what do people do? They go to the developers. now it's it's such a break mentally for them to just ask claude they're like well i don't know how to use claude well i'm like when i ask claude how to use claude like that's the the weird loop that you have to break is like you don't go to another person to learn how to do this like the number of times i get messages it's like how are you getting claude to do this i'm like bro ask clod like that's what i'm gonna do i don't know so it i find that that right there is a place to start with a lot of organizations it's like just kind of asking dude and i've brought this up we do every month or every two months now we do ai game building and you just yeah you give everyone a couple hours everyone builds a game it's a fun little thing and you can see how far clod has advanced doing that we used to be we used to build like you know pac-man simulators basically and now you can build you know full 3d shooters bro like it's crazy what you're able to do yeah so yeah i think we're we're gonna do more hackathons now like as a consumer company i never thought i would use that word uh but i'm gonna get like everybody to install clod code on a friday and we're just gonna start playing around and building stuff like everybody anybody who wants to participate that's going to be your signal that you tell me that you're curious and i'm going to show you how to do stuff but build your own little tools like little things that you can use but i think that's a great step too yeah look it should be mandatory and if people complain they should get fired it's like that's the reality it's like imagine imagine if you're in the 1990s or whatever and you're like hey this thing with the internet they're like ah it's boring i don't care about the internet like dude you're not going to survive man what's up operators welcome to the rich panel ad read Rich Panel has been a sponsor for over 12 months.
1:02:59Sean Frank:I've been a paying customer for over 12 months. And guess what? I just renewed to pay again for another year. We have cut our SaaS bill in half and automation dropped our cost per ticket by 70%. Our CSAT has also improved from 88%, which is still really good, to 96%, best in class, all powered by Rich Panel. I told them last year, hey, you guys need to do the same thing with returns. And now Rich Panel has a returns portal. It's built to cut down your tickets and convert more refunds into exchanges. They do the heavy lifting, data imports, self-service, retention flows, team training, all of it, and they'll be live in two weeks.
1:03:31Sean Frank:If you want to save 30 % guaranteed on Helpdesk and now returns, book a demo. Okay, so Amit, we're talking a lot about how Rich Panel can be disrupted, right? And you've been pretty honest with us and kind of what you're doing to defend against that and how you're just going to kind of do that to yourself, which I think is brilliant. end. How are you then thinking about, like, that's the defense side. How are you thinking about growing RichPanel, like growing revenue, growing the capabilities of the platform? Like, are you guys adding things or is it more like depth? Just give us a lay of the land.
1:04:10No, absolutely. That's a great question. And in fact, you know, when we initially started the pod, we said that the cost of writing code is going to zero, but there are like a couple of other things that remain relevant and maybe more relevant than before in the AI world. So one of that is like distribution. The second of it is the brand that you've built. In the industry, we've done a fairly decent job of earning the customer's love. And we want to extend that by launching set of new products. So when I was talking to Sean and others, I think this was like June of last year where Sean is like, dude, you guys should be building returns.
1:04:46And he said that in passing but I took that seriously I interviewed like a bunch of other customers and as it turns out that people do really want a new solution that is a has like the capabilities all the high tech capabilities as as some of the leading players out there but it's also easier and more cost effective so with that thesis we built our own returns portal this would have traditionally taken us like a couple of years to build it but we did it like over the last six months you know ironed out all the edges. We launched it in like Jan 1st and we've signed up like 31 brands already. So that is how we are thinking about using our existing customer base and the existing brand that we have to, you know, offer more services and help them like cut costs in like each of these other areas too, right?
1:05:37And by the way, there is no announcement. This is not even on the website. There's no even like a page on the website. It's just like talking to a few customers based on the capacity we have we're onboarding them uh on the returns platform yeah but i do see that
1:05:50Sean Frank:the reason why i wanted to return so bad is i was with loop and i hated it like i mean why yeah yeah like like loop loop i think is considered to have the best in class features and like this isn't a dig i like i like all the guys from chubbies and whatever but the problem with loop is they raised a bunch of venture capital and then the cost got crazy and like they were literally trying to charge me ten thousand dollars plus per month to use a returns portal and i'm like what the hell are you talking about it's yeah you know like and then we we shopped around other returns portals or whatever and like you know we went to a company i think they're called return go and they just didn't work it's just like it just didn't work it was just like it just wasn't built well um you couldn't do exchanges for instance like just just horrible, horrible software.
1:06:44Sean Frank:So that's probably when I approached Richpan, I'm like, guy, I don't know why it's so hard to build a returns portal, but somebody should do it and give me a loop's functionality for half the price and I'll be happy. So that was where it came from. Matt, what are your thoughts on returns as a service? I think it's brilliant, man. I come back to your comment, Sean, on this whole brands are going to have a head of customer and that head of customers going to have like a whole series of agents and things that that person is going to control. And I can just see a world on it where like RichPanel can be the platform that has like a whole whack of those agents.
1:07:19Like I don't know which ones, but anytime like that we want to talk to our customer, I could see a world where like RichPanel has the agents that let us communicate to our customer that we're selling something, supporting them, handling a return. I can just see that, that that, that has to be true. Like someone has to own that, that set of pipes. Right. And I love it.
1:07:39Sean Frank:And I just don't know why reviews has some people building in it. Like, uh, there's Okendo, there's judge.me, there's stamped, there's Yapo. I think Rebo has, has reviews now. It's like, everybody's building in reviews and I'm like, I don't know. I don't like us do that. Returns is a way bigger pain than me. I don't really care about how I get reviews on our website. It's all a cookie cutter to me. Returns should be just as cookie cutter. And for some reason, Loop was really trying to charge me a six-figure contract for the joy of running my returns through them. It was the most egregious thing I've seen.
1:08:13Sean Frank:So that's why I was begging people. There's not a lot of categories of SaaS in consumer that should charge more than Shopify. That's sort of my baseline for pricing. If you're asking me to pay more than shopify bro you better show me oh that's gonna make me more money because that's hard yeah and i brought this up shopify has tremendous restraint like people people love to talk about everything the most you'll ever pay shopify is forty thousand dollars a month that is that is yeah they have a variable fee and it's a percentage of revenue that's the most you can by contract most you can ever pay them you could be doing a billion dollars a month and that's the most they're gonna charge for you.
1:08:52Sean Frank:I'm like, it's a pretty sweet deal, guys. If you send too many emails in a month, you might pay 40 grand. So yeah, anyway, I got pissed off at Loop Returns. So that's why I'm trying to have people build it. And thank you, RichPanel, for building the cool software. Ahmed, I think it's super smart, man. That's my short answer. I think it's smart that you guys are doing that. That's great. I was saying that's the exact brief I gave. Like Sean said, build a better Loop Returns platform about half the price. So that was a brief. Feature brief. Loop is too expensive. Make more mistakes. Make more mistakes.
1:09:24You don't build this.
1:09:25Sean Frank:Yeah, yeah. One shot at Loop, dude, in a clawed prompt. Okay, a$10 ,000 bet to charity. Rich Panel is going to deliver an AI agent that does 50 % of my tickets for this Q4. And if they do, I will give$10 ,000 to the charity of your choice. But if you fail and you only get to 49%,$10 ,000 is a charity of my choice. I will I'll take that back and if you want 50 % of your tickets answered by AI by Q4 you can go to richpanel.com he's got you taken care of I appreciate you guys I'll talk to you later
From the publisher
What happens when the founder of a software company admits his own product is a sinking ship — and doubles down anyway?
Sean Frank (CEO, Ridge) and Matt Bertulli (CEO, Pela Case & Lomi) sit down with Amit RG (CEO, RichPanel) for a conversation about AI agents, the death of traditional SaaS, and what it means for ecommerce operators building teams right now.
Amit doesn’t sugarcoat it. He’s walked through the five stages of grief about his own business, come out the other side, and is now betting that moving faster than the disruption is the only play.
The three dig into Dario’s AI phases and what they mean for your tech stack, why customer service reps will be the first wave of AI-driven displacement, and how Sean frames every job — from Philippines call centers to legal firms — as a token problem waiting to be solved cheaper. They debate token-based pricing versus per-ticket models, explore what happens to platforms like Klaviyo in an agent-to-agent economy, and make the case that consumer brands may be the last ones standing while SaaS and services absorb the first hit.
Amit closes with RichPanel’s new returns portal, already live with 31 brands, and Sean puts $10,000 on the line in a charity bet.
Powered By
Fulfil
Northbeam
Saras Analytics
Aftersell
Postscript
Richpanel
Operators Newsletter

