Salesforce & Microsoft’s AI Sales Challenges, AI’s Impact on Product & Sales Teams | Sep 16, 2025

0:13Welcome everyone to the information's TI TV. My name is Akash Pasricha. It is Tuesday, September 16th. We have got a busy show for you today packed with a lot of enterprise software software. Today, we are talking about Satya Nadella's new playbook from Microsoft's 365 co-pilot and how that's going. We're also going to bring on Index Venture's newest partner, the former chief product officer of MongoDB, who has just joined the firm. We're then going to be talking to the CEO of Clay, the company that sees itself as a go-to-market engineer. We're going to get into what that term actually means. But I want to start with a big story that we published this week about Salesforce.

0:55Salesforce has gone all in on AI with AgentForce and its tools. But this week, my colleague Kevin McLaughlin published an in-depth piece about some of the challenges the company has had selling AI to its customers. There are a ton of different angles to this story, from pricing to talent to competition to Mark Benioff's management style. I want to bring on Kevin to talk about his story. And I also want to bring on Adam Mansfield, who leans software negotiations at Upper Edge to give us a bit of his perspective. gentlemen welcome to you both it's great to have you thanks for having me yeah appreciate it uh appreciate the opportunity i will say i love the intro music it's one of my uh favorite things here might as well get ready for college football season so you know as a as a drummer and someone who who wished that they could have played in a drum line growing up i mean i we we had to get the drums in there so i'm happy it resonated with at least at least one person um i love it All right.

1:50Well, Kevin, let's go to you. So, you know, big picture, I think it's well understood that selling AI is really hard for enterprise software giants. But you wrote about Salesforce. And so the first question I have for you is, I mean, tell us about why Salesforce's challenges stood out to you. Well, I think Salesforce, perhaps more than any other company, has gone so hard into AI that they've become this sort of bellwether. And what I mean by that is other CEOs, in some ways, they've sort of couched their predictions about AI and what AI would mean for the company. Mark Benioff, CEO and co-founder of Salesforce, at one point last year suggested, jokingly, I think, that he might rename the company to AgentForce, which is also the name of the product.

2:39So I think that watching how their customers adopt AgentForce, which is their main AI product, is very interesting because Salesforce has a unique ability over the years to get customers on board with new products. And so for those reasons, they are a very interesting case study. Okay. And so are customers adopting AgentForce? How's that going? They are, but perhaps not to the level that Salesforce wanted at this point. Nine months in, less than 5 % of the 150 ,000 plus global Salesforce customer base is paying for AgentForce. There are a number of reasons for that. And it's not to say that AgentForce is not going to see adoption someday.

3:22But at this stage of the game, I mean, myself and my colleagues have written about this quite often in recent months. The ROI is a little bit murky for a lot of companies. In some cases, they don't have their data in order. There's a myriad of reasons why they're not adopting quickly. Like, give us some of the reasons that you've been hearing about. Well, cost can be an issue. um also like the first wave of of ai was about assistance and and sort of features and applications and the roi just hasn't been there for that which is why sort of as a separate topic we're starting to see ceos talking more frankly about ai as as labor replacement uh that wasn't really the focus of this story um but yeah those are just a couple of the things that we reported so adam your firm helps some of these companies buy software, essentially.

4:18My understanding is you basically help them negotiate better rates on these giant software packages for companies like Salesforce and a host of these big public software companies. What are they telling you about Agent Force and how they're finding the product and whether they want it at all? Yeah. Listen, there's a couple of things here, right? So when we work with our organizations, absolutely at the end of the deal, you want to make sure you have the right pricing, right? The Before you even get down that path, you have to start thinking about, well, why are you even considering it? Right. And that's where there's a lot of like blank faces.

4:54Right. There's this idea of like, well, it's AI. So from the top down, we want to start thinking about Salesforce and their part in this. What ends up happening is when they bring Salesforce in, the next conversation is also, but I'm spending a lot of money with other Salesforce products that I'm not getting the full value on. So I need to really rationalize this. So before I start jumping into AI, is Salesforce going to help me on everything else that I'm spending millions of dollars with on? Like, where is that happening? So I think that's slowing things down a little. I would also tell you that the business case component to it, I mean, Kevin mentioned this and it was well said in his article, right?

5:34Like these customers, a lot of them, you have to think about as individuals. These C-suite individuals are going to put their neck on the line, right? And this is not something that you just taste it and then maybe you can correct it later. These are people that have to say, I'm putting my name to this product. 100%. And on top of that, there's the licensing component to it. This is not by the headcount anymore. This ties actually really nicely into the labor piece, which I won't go too deep into. But when Benioff gets up there with a smile on his face and he talks about 9 ,000 people going to 4 ,000 people, that's a headcount thing.

6:09The industry, the financial community talks a lot about this with Salesforce and others and says, well, if companies are going to be reducing their staff and you sell products based on a per-end count basis, help me understand how I'm going to make money. But the way they overcome that is AI, consumption-based pricing. Right. But when I talk about putting your neck on the line, it's not just the dollars up front. It's also the flywheel effect, which, by the way, everyone within Salesforce who gets up on the earnings call, they talk about the flywheel effect as well. This thing starts spinning.

6:44No customer wants to then have this thing spinning and not really be able to KPI or metricize the value proposition that they're obtaining back. It's a very difficult thing. Right. And tell me a little bit about, you know, one thing that Kevin sort of hinted at in his story, and I wanted to ask you about, Adam, is sort of the convergence of these enterprise software companies with respect. Everyone is doing agents. And I don't know, it kind of feels like they all sort of do something around the same thing. Is there kind of a convergence of these companies in terms of the products they sell? And the secondary question there is, does that cause decision paralysis for customers saying, hey, there's so many options now, we don't know which one to pick.

7:27So it's a great question. So I definitely think that's creating paralysis for sure. It's like anything else in life. When you give someone a lot of options and they seem like they're good options on the surface, you got to take time to evaluate them. Number one, when you talk about millions and millions of dollars associated with these options, you're going to take even more time to evaluate them. When you're talking about all of these vendors likely already being vendors for you where you're selling other things with big cost profiles, you're going to take even more time before you give them dollars.

8:00So all that adds time. What I will tell you, though, going back to what I said before, when it comes to Salesforce, the real friction, at least the deals that we work on, the customers we interact with, large enterprises, It really stems from I'm looking at a base of products that I am spending millions of dollars on, and I don't really understand the value I'm getting out of the things I'm already paying you for. I also don't understand where my dollars are going to enhance the thing I already have. And don't tell me the enhancement is I got to give you AI. That creates friction. That creates delay.

8:42which is why, in my opinion, and the 5 % number is a great one, Kevin, but when you talk about numbers and you dive in a little bit more, what is it, 12 ,000 customers are currently aging for its customers? And then to their credit, they said only 6 ,000 are actually paying. So there's some murkiness in there too. Like, what does that actually mean? So Kevin, I wanted to ask you about some of these startups that are starting to encroach or trying to encroach on Salesforce's space. And we had Brett Taylor on, the CEO or former CEO, former co-CEO of Salesforce, who's now heading up Sierra. And, you know, you talked a little bit about some of the, I guess, the annual contract value that Salesforce is expecting from AgentForce.

9:25Company kind of says that's sort of a measure of annual recurring revenue. And it felt like that was kind of in the same ballpark as, you know, what Sierra is sort of approaching with its own revenue, right? So the question I have is, I mean, you've got a startup whose whole business is AI, and it's already approaching the scale of AgentForce's business. Am I thinking about that wrong, or walk me through that? Well, when Salesforce first introduced AgentForce last year at the Dreamforce conference in September, Benioff's pitch was, this product is part of our platform, and as a result, it will be easy for customers to implement.

10:05It's not like you're out buying some bespoke product that is separated from everything else. It's part of the platform. What's happening, at least with Sierra and another startup called Decagon, is that they're coming in and they're focusing just on customer service right now. And I think Sierra is also focusing on sales enablement, which is a separate topic. But basically, yeah, the startups are getting buzz. They're getting investment. They're getting attention. And they're getting revenue. Sierra is on track for$100 million in annual recurring revenue, it's been reported. Decagon, we don't know a little too much about what's going on with them.

10:44But the point is that they are not beholden to the same base of customers that Salesforce is. The pressure is far less on them. Of course, they're going to have an advantage. I mean, the question is, what's the long-term play for Sierra? And Brett Taylor on TITV a few weeks ago said that best of breed, which is basically the best technology, companies are not as focused on the cost as they are on the quality of the experience right now. So, yeah, the startups are having their moment, but I wouldn't say that that gives them the inside track to win this race. Right, right. Adam, last question for you.

11:26And then, by the way, we're going to keep you on because we're going to talk about another software company in Microsoft in just a minute. But very quickly, as it relates to CRM and competition with Salesforce, I mean, you deal with really, really large software contracts with your clients. How much are they paying attention to these smaller startups? Or is it something that, you know, again, they get a lot of news, they get a lot of buzz, but the large companies, are they really paying attention to them? Yeah, it's a really good question. So two things. First, like Sierra is the example. You know, Taylor has a pretty, if there isn't such thing as a Rolodex anymore, he has a pretty big Rolodex, right?

12:05So that's part of it. The other thing is, so he's getting access and they're getting conversations. And on that value, they're focusing on value, not, it doesn't feel like they're forcing it down the customer's throat. They're trying to earn the business. There's a difference there. In terms of competition, in full transparency for the deals, the customers we work with, when you think about a replacement for Salesforce, generally speaking, there's Microsoft is coming in, right? So their D365, we all know that they have their death star of cash that they can point at a customer and say, hey, well, what if we were able to do this?

12:36maybe we could get you to take some more market share in the D365 space or the CRM space. But there's other players, right? There's some of them. We don't spend a lot of time in full transparency on kind of that smaller market. It's mostly just Microsoft. But again, they're massive too. And how you approach them is critically important too. Right. Great. Well, Kevin, thank you so much for coming on. We really appreciate it. It's a fascinating story. And you write a ton about Salesforce. So I think we're going to have you on more and more as the story evolves. And Adam, don't go anywhere because we're going to bring you back on in just a second to talk about Microsoft.

13:11That is Kevin and Adam. Okay. Well, Salesforce is not the only company paying close attention to its AI rollout. Microsoft is also doing the same thing, and it is keeping a close watch on the uptake of Microsoft Office 365 Copilot, which is the AI-infused version of Microsoft Office. My colleague Aaron Holmes published a story this week about some of the ways in which Satya Nadella has reorganized the company to keep that product competitive. I want to bring on Aaron, and I want to bring back Adam, who also has experienced talking about Microsoft's products. Welcome to the both of you, Aaron. Great to see you.

13:46Thanks for having me. Okay, so let's talk 365 Copilot. How exactly is that product going for Microsoft right now? So, I mean, there are ups and downs so far. I think that, you know, Microsoft has made some progress convincing businesses that it can be worth the$30 per user per month upcharge for Office Copilot, which basically more than doubles the baseline price of the product. And, you know, I've heard from some customers who say it's great for things like taking notes in meetings and, you know, organizing their email inbox. But at the same time, I think Microsoft has struggled with some of the more advanced features that it promised when the product debuted two years ago.

14:26Things like, you know, creating PowerPoint presentations or carrying out tasks within Excel. And I think what I'm hearing from customers is that it's sometimes hard for them to explain what value they're getting for that$30 price point. And as a result, it's not really clear that AI features are helping move the needle or accelerate the growth of Office more broadly. So I think now what we're seeing is Microsoft is taking some big steps to try to get the products to a better place. we broke the news last week that they're about to announce a new deal with Anthropic to power some of those Excel features.

15:04And we're also hearing about a little bit more leniency on discounting the product, which I think points to the urgency to get this into more customers' contracts. Adam, what are your customers telling you about Microsoft? Yeah, a lot of the things that Aaron just covered, just to go a little bit deeper on that, I think, you know, number one, I think Microsoft is approaching customers originally from a hardline stance, right? Like, hey, we're Microsoft. If you're going to do AI, why wouldn't you just add Copilot, right? They led with that. They led with, if you want to do Copilot, well, you got to start with 300 as a minimum.

15:41So why, how wouldn't you want to start with 300? So they started to kind of ease up on some of these things. But I will still tell you that the ease up still isn't driving, at least the customers we work with to a place of saying, all right, great. We're ready to go from 300 to 5 ,000 or 300 to 10 ,000. We're not ready, right? We're seeing some incremental. We're not ready. Why? Well, part of it, there's two real main reasons. The number one is they're still struggling with the, what am I actually paying for conversation, right? So what happens is there's a big sticker tied to this, right? And this is important and I'll spend a second on this if I can.

16:24The co-pilot price is$30 per user per month unless you negotiate it down, right? Which there is abilities to do that. But the real way customers look at this, which is creating friction, is it's not$30 per user per month because you have to have the underlying Microsoft 365 subscription. So the real math is I'm giving this product, co-pilot now added to Sally in accounting and I'm paying whatever it is per user per month, which is very expensive. And what I'm hearing is that she's not using it or she's using it for note-taking. That is ridiculous. I am not going to invest more into this so that more of my employees, Steve, Bob, et al., have this very expensive headcount cost for something they're barely using.

17:09That's the issue, right? So is this kind of a similar issue to Salesforce now, which is that you're asking me to pay for something new and I'm not even getting the value out of the existing suite of tools that I'm paying for. Is that the common thread here? That is another layer, which is really spot on on your part. On top of this, so they're looking at Copan and going, yeah, I'm doing it for note-taking. And again, I think Aaron and I would both agree, like there's probably a little bit more other use cases, but that's one that we hear all the time. What value is there? Then the underlying subscription price for E3, which by the way, after November 1st, They're getting rid of all volume discounting.

17:46So that's going to get even more expensive. But what's happening is customers are looking deeper at their E3 or their E5 subscriptions. And they're going, wait a minute, I'm using 40 % of that. So I'm paying air for something that's very expensive. And then you want me to put something on top of it, which is also expensive, which is probably going to have more air in it. So to understand this, you want me to put my neck on the line to buy a lot of air that might become valuable. It might become valuable, but by the time it comes value, am I up for my next renewal in three years? Like, what are we doing?

18:21And so I'm not willing to put, just like on Salesforce, am I going to be the one who puts my neck on the line for that? I don't know. That's a tough one. So Aaron, how is it that Satya Nadella is rearranging the company to address some of those issues? I mean, he's been making some changes. Yeah, so I mean, we saw a big shakeup in the office unit earlier this year. Specifically, he brought in two executives who were not working on office previously. One is Charles Lamanna, who was working on business applications like Dynamics. And then the other is Ryan Roslansky, who is the CEO of LinkedIn. Both of those executives are now overseeing the office products, but Ryan specifically is still serving as the CEO of LinkedIn while reporting directly to Satya in that role, which I think kind of underscores the urgency that there was around finding new leaders to lead these products.

19:15And, you know, from what I've heard, they've been moving quickly. I heard that Charles LaMana was, you know, pretty closely involved in this effort to bring in Anthropic models to make some of these products work better. So I think now we're probably going to see some announcements between now and the end of the year in terms of new features or upgrades to the features in Office Copilot. But I think the goal there is to move the needle more on actually automating white-collar work so that companies can feel like they're getting cost savings in the form of not having to pay people to do as much human white-collar labor.

19:48Right. Adam, last question for you. I mean, you spend every minute of your day talking to customers buying from these big tech companies. You're telling us, hey, they're concerned about the value they're getting from their existing products. They're not convinced on the value of the new products. A lot of questions in the air. How do you think this plays out? I mean, you know, let's go forward three years. What's your prediction here on, are we going to see some major sort of re-rankings of the big tech companies? Are we going to see startups, these AI startups, you know, become, you know, big public tech companies to compete with them?

20:24How do you see this happening? Yeah, I think it's a good question. What I would tell you is that in this particular space, we talk about productivity suite. I mean, I know I'm oversimplifying, but like that's what this is attaching to, right? if we're just being candid and honest out there, at least the customers we work with, the only other name is Google, right? Workspace. And I'm not counting them out in this game. I just think that what I predict is going to more happen is that Microsoft is going to continue to find ways to get creative, which are usually not beneficial to the customer. They do it to create leverage, to force the customer's hand into where they want them to go.

21:01Two examples I mentioned in November 11th, all the volume discounting is going away. They're going to say to the customer, hey, well, what if I can help you on your E3 price if you added Copilot? I have put this out there. Others have as well. I would not be surprised if you started seeing something like, hey, E5 now comes with Copilot, right? They're going to force the movement to E5, right? They're going to remove optionality. So they're going to force customers into a path. In other words, if you want the other things that come with it, because you started using Defender, you started using the security pieces that come in E5.

21:38Now, in order to get, you're going to have to keep that, you're going to have to add Copilot, right? But do you think customers are going to stand for that? I think they're not. And what I'm already seeing is this. There's two things. Customers are spending their time focused very, very, very closely on where can I take my ball and go home? They're really focusing on the pieces where they can. They're being honest with themselves. Could I really rip out Microsoft 365 E3? Could I do that? Could I? I don't know if I can. And the answer, if I don't know, is probably no. But what can I do? Yeah. AWS, GCP, want my Azure business.

22:20Highly competitive. Maybe I'll go away on that. Let's talk about things like D365. On the last conversation we were having, you asked me about Microsoft potentially coming in and taking Salesforce business. Don't think Salesforce isn't going to come in and try to take D365 business. Unified support. This is a really important one and we haven't talked about, but I'm going to spend a second on this if I can. When you talk co-pilot pricing, you have to factor in unified support because if you're a unified support customer, it's a percentage of your net spend. If you add co-pilot, your unified support cost goes up as well astronomically.

22:56Customers can move off of unified support as a lever. And I think what you're getting at, and it's a great graphic in Kevin McLaughlin's story, which is the last story we were talking about, the Salesforce story. He has a great graphic that shows a lot of these big public tech companies and the products that they're all now offering. And there's a lot of overlap in them. And like you said, I think we're going to see a lot more crawling on each other's turf. Adam and Aaron, thank you so much for coming on the show. Adam, thanks for coming on for the doubleheader. Maybe we'll have to have you on again.

23:27That is Aaron, who covers Microsoft, for the information, and Adam Mansfield at Upper Edge. Okay, well, Index Ventures made a big hire this week. The company tapped MongoDB's now former chief product officer to be its newest partner at the firm. Sahir Azam spent more than nine years at Mongo, but he is now taking his talents to the land of startups. I want to bring him on to talk about some of his reflections and also the current software climate. Sahir, good to see you. Good to see you, Akash. Thanks for having me on. Congrats on the new gig. You must be excited. I am. It's already very busy and it's been great.

24:04You don't have to worry about quarterly reports and all those headaches. It's just a different set of things to worry about. Well, we're thinking long-term now. We're all about the long-term investing here, right? In venture. As a chief product officer, I was lucky enough to be able to think at Mongo, but yeah, definitely it's a different environment and a fun new challenge. So thank you. Great. Well, so, I mean, let's just talk about what you're, I do want to get some of your reflections on sort of how product management has changed in the era of AI, because I think that's something you, you know, you've, you've been in product for so long now, but very quickly, but what will you be focused on at Index investing in?

24:45Yeah, absolutely. So first off, like Index as a fund is a very flexible platform, which is really attractive to me, meaning I have the ability to invest everywhere from seed all the way through growth and across geographies. And we don't specialize down to a point where we limit partners to particular domains. Now, that being said, I have a lot of history, background, domain, and AI data infrastructure. So those are areas I'll definitely be spending time. They're only getting more strategic and also starting to dip my toe into security as well as a clear adjacency. I think that whole market is going to be very dynamic as people are trying to build these new style of applications powered by AI.

25:25And so those will be kind of my core focus. Got it. So let's get into this topic of the role of the PM in the era of AI. You managed a big team at Mongo, and I presume you're going to be advising a lot of these portfolio companies on how to build their product teams. How have you seen that role change over the last three years? And how might you counsel some of these younger startups on building these product teams with AI in mind? Yeah, that's a great question. I think there's no shortage of people out there that have rightfully said that the velocity and pace of execution right now in AI startups is just faster than I think anything we've ever seen.

26:12it's highly competitive. You're hearing kind of 996 culture. And in many ways, the product evolution is happening so fast that the traditional defensible moat of technical differentiation is really hard to sustain. And so agility of execution is sort of prime for that. As a counterpoint to that, though, it also means distribution and brand really matter. So as a PM, I think it's critically important to not just be a more classic inbound-oriented product manager working only with engineering as kind of your key stakeholder, working on prioritization and shipping, but really thinking holistically about the strategy of distribution in conjunction with VP sales, CRO, et cetera, and designing the right architecture for the go-to-market to be able to build that as equal and if not as important as a differentiating factor of the business as the core IP of the technical.