Cynthia Hansen (Innovation Foundation): 4 lessons from scaling social innovation

0:00That's been one of the biggest learnings is that often people who are founders and are kind of experienced founders only know how to do it if they originate the idea. And so having someone who comes in and actually will build for us and build our idea, even if they haven't been there from the very beginning, that one is really hard. We spend a lot of time talking with the board about what does scale mean and that scale is really about scale of impact. So we've actually set up our own impact measurement framework because there wasn't an existing one that we felt suited us really well. We originally assumed that it would be really easy to find experienced founders who now wanted to do something purpose-driven who would come in and build with us.

0:34But we found that actually it's really hard to find the right combination of skills in one person. Someone who knows how to build a venture and build something that will be sustainable. Who knows how to work with a team and in matrix with other kinds of service providers and who understands the social impact side. We found it very hard to find like the perfect match. So what we need is good matches and then an array of other people around them, because you can't expect one person to be able to do all of that.

0:59Andreas Munk Holm:Welcome to the AVC podcast. So the ADECO Group is one of the largest workforce solution companies in the world. Inside it, Cynthia Hansen has spent the last several years running something most large companies don't have the patience to build, a social innovation lab that designs, tests, and scales employment solutions for the people that mainstream labor market has left behind. Young mothers, women re-entering the workforce, and people over 50. She spilled it from scratch, and she spent years getting this wrong. Today, she's walking us through the four things she would have done differently had she been about to start today.

1:33Andreas Munk Holm:Four learnings that now form the backbone of a Deco's 2026 to 2030. Motherfucker. 2026 to 2030 strategy.

1:47This show is not investment advice, and the hosts of this episode may be invested in the funds and companies featured.

1:53Andreas Munk Holm:Jaber, can you tell us a bit about ADECO and why you thought we should bring Cynthia on the podcast? What Cynthia is doing is in the vertical within corporate mentoring that we call BUILD. It is a very, very challenging part. What I think is extremely interesting is that ADECO is in the people space. So here now we have an expert with us that is building companies and has access to people within a very niche vertical of that as well. So we're really looking forward to learning more from Cynthia and what they're building at ADECO. Cynthia, before we get into the learnings, can you just give us one minute on the Innovation Foundation just for the listeners who haven't come across it before?

2:36Of course. The Innovation Foundation is the global corporate foundation of the ADECO Group, and that's what makes it slightly different from other kinds of corporate venturing within big organizations. So we set up actually not as a grant-giving foundation, not as an incubator or an accelerator, but rather as a social innovation lab. So instead of giving away money, all we do with the chunk of money that the ADECO Group gives us every year is to actually build these ventures, these solutions that serve a particular niche in the labor market and that help to bring in people who are otherwise excluded from the future of work.

3:13Andreas Munk Holm:Now, I want to guide us to the first learning, which is about picking where to focus. I want to ask you, and I want to get into this a little carefully, because the assumption you started with designing programs by demographic sounds completely reasonable. Youth, women, mature workers, you would think that it makes a lot of sense to build around that as your starting point, but you actually learned that that wasn't the right way to go about it. Exactly. So when we started up, when we actually moved to this model, which was around 2020, 2021, we wanted to start with a demographic that we knew. So we already had a footprint in the youth space, looking specifically at youth who were being excluded from the labor market.

3:59So we started with youth and decided, okay, within youth, let's drill down into who is the most excluded, who is most in danger of never joining the workforce. And so we built our whole kind of narrowing methodology around the idea of starting with a demographic and then narrowing from there. And we found interesting things. We then figured out who within that big demographic was most at risk, narrowing down to youth who are neat living at below the poverty line on the urban fringe, young mothers being a more extreme use case. And so all of that made sense. And then so we designed the first project from there.

4:37Then when we went through this, once that was off and running, with the next project, which was Women Back to Work, then we did the same thing. We started with women and then looked at who within women were kind of most in danger of either not joining the workforce or not being able to rejoin the workforce. That one was also sparked partly by data we saw coming out of COVID about high numbers of women who left the workforce and didn't come back. And so we'd use the same methodology and the same set of 15 drivers that we had set up originally. But then what we started to find was repeating patterns.

5:11If you're a woman, and if you're a woman maybe in a more male-dominated field, or if you're a woman and you have caring responsibilities, If you're a woman and you have limited education or you've always been in kind of low-skill, mid-skill jobs, then you have those same barriers to entering work. You know, maybe lack of confidence, lack of opportunities, lack of the right network around you, bias from the employer side. So we started to see those patterns repeat. So that one we designed slightly differently, often running and also youth at risk still running. Then we went into the third cycle and, again, picked a demographic, saw that mature workers, particularly this was starting about two years ago, was a fast-growing topic and one that was definitely hitting the global north but was starting to actually rise in the global south as well.

6:05Now it's a very big topic. So we started using that same methodology to start with mature workers and who is being left out. We looked specifically at men who had come up through the trades, et cetera, et cetera. But what we found, again, using that same set of the 15 drivers, which are things like education and socioeconomic level and family structure, et cetera, that the patterns repeated. And so then we started to see that actually if you have enough of those intersectionalities, that that's actually what tends to drop you out of the workforce or make it harder for you to get in or back in rather than the pure demographic you come from.

6:42So that actually led us to think differently about how we choose.

6:46Andreas Munk Holm:So what do you think caused you to start with demographics and what made it, so to say, take years before you realized that this was the wrong way to go about it? And how did you correct it inside of what is a pretty big ship? We started actually with choosing youth because we wanted basically one demographic as a test case. And so in the early days of the foundation, we were more a programming foundation. We were an umbrella for some things that had been taken out of the group and put into the foundation to run social programs, essentially. Two of them actually overloaded, overlapped with youth.

7:26So one called Youth at Risk was kind of a health and well-being program, but also that put money into youth programs. The other one, CEO for a month, was our own internal youth program. So we already had the footprint in youth. So when we went to the board with the proposition that we shifted into this social innovation lab model, we also made the proposal that we choose one topic and that we use that to experiment scaffold around. And we gave them an array of several different things. We looked at disability and women and youth. And with the board decided that actually youth would be the best one to start with because we already had the most knowledge and a footprint and visibility in that one.

8:02So that's how we started with that. And then the reason that it took quite a long time to see that the patterns repeated was that we're running really kind of starting a new project basically once a year. So you get one off and running, you do the scan and into the build. And then by the time you're in the build, then you start the next one with the scan. And so by the time we had done that three times, that's when we had run the scan three times and started to see that really quickly it was repeating. We actually changed some things along the way when we did the second project. So, for example, in the first one, we waited to do a lot of field research until we were in the build process.

8:40We saw that that was too late. So then the second time through with the women back to work, we already incorporated a lot more field research early, early on. But what we found was that field research then reinforced this idea that the patterns were repeating. So then with the mature workers one, we actually took a longer time to do the scan. We spent a whole year doing the scan instead of roughly four to six months we'd done in the past. so that we could do more field research and really dig into it and figure out what parts of the pattern were unique and weren't just repeating. And what we found was good uniqueness, but we still found repeating.

9:15And then that's what led us to the next stage, which is really the how do we actually design differently now? And I can tell you a bit more about that.

9:23Andreas Munk Holm:Yeah, and let me just stay on it because you started out focusing on demographics. Now when you're spinning up new projects, so to say, am I right in saying that you focus more on trends? Yes, exactly. So what we've decided now, and this was part of the big review that we did at the kind of halfway through 2025 in preparation for the new 2026 to 30 strategy, was to figure out, you know, what was working, what was not working. What we found was not working was choosing the topics by demographics. Our hypothesis was that if we actually saw repeating trends, that it would be more fruitful to actually start with the trends.

10:02So we retooled our whole approach to research to actually prioritize looking at trends. And so now we look at trends across the labor market and then look for a trend or multiple trends that are disproportionately impacting people who are already disadvantaged. And then we'll grab one of those trends and then do a deeper dive into that. So we've actually set up now a more tech-enabled approach to doing particularly that first cut of broad-based research so that now we can actually do constant trend spotting. At the moment, this tool that we've set up actually does a scan every two weeks. And then we can see across the things that we've to look at what are the things that change, what are the things that don't change.

10:46And then we can grab trends like shrinkage of entry-level jobs or increase in demand in the care economy. And then we can use the same tool to do deeper dives and then complement that with our own desk research and with our field research to figure out if that's actually a trend that we want to go deeper into and that we think actually lends itself to a real project.

11:07Andreas Munk Holm:I'm super curious, Cynthia, because a person I love in the venture space is Mike Maples Jr. from Plotgate. He has a framework around inflection points and coupling inflection points with adoption points or changes in the market as well. And then, of course, technology. And then he looks for founders that have a unique insight within that space. And it's kind of when all of these things come together, an inflection point in the market, an inflection point in the technology, inflection point in regulations, those things externally, so to say. And then a founder that has a unique insight from, he often uses the example of Morgan Driesen back when he was one of the very first people to be exposed to seeing coupled computers.

11:56Andreas Munk Holm:Then he could see how the internet was moving in a completely different way than what everyone thought at the time. Because we, of course, all thought that we would have this one line of communication, so to say that they called it the internet highway or is it something like that so similarly there was an assumption that the internet would be used exactly like flow tv that would just be sent out but what he realized was it was very much a peer-to-peer thing and people would actually enjoy sharing videos of or pictures that are loaded once once every 10 seconds of coffee machines that would be dripping and people would actually sit and watch that and that showed to him the power of the peer-to-peer network on the internet.

12:36Andreas Munk Holm:I'd be curious to understand if we looked deeper into how you look at trends, what do you look for? How specific do you go? Do you go at it in the same way as a venture fund would and look for these inflection points that cobble together with founders that have insights to build in it? Or how do you go about it? That's a great question. And we're still exploring this because the actual tool and the way we use it is quite new. But what we're finding is that we are looking for essentially inflection points, but we tend to think of them as aha moments. And it's usually something that will appear that makes us think, OK, there's something missing.

13:17So if you take the example of entry-level jobs and shrinkage of entry-level jobs, we started to look at this about eight months ago. And at that point, there was no data about it. So our assumption was entry-level jobs are starting to shrink, and this is going to put an unusual kind of unprecedented pressure on the labor market. But there was no data. And so we started looking then for that data. And then we saw that actually there was suddenly this huge influx of data, most of it coming out of the U.S., some out of the U.K., definitely global north. And it was all about AI and entry-level jobs, white-collar jobs, and university grads who didn't get jobs.

13:57And that is a problem. But then that was actually this inflection point when we saw, okay, this is now the narrative is growing, that the narrative is being hugely skewed in one direction. What's the piece that's missing? And that was that inflection point. Okay, now there's enough data to have something to push against. So then we actually did this piece of work at the very beginning of this year to say, okay, what's missing and what's the bigger picture? If you were to paint a more balanced picture across the world and across different demographics and across different industries, what does it really look like?

14:28So that was that inflection point of, okay, here's the piece that's missing. So we did that and looked specifically at, you know, how does that trend play out in emerging markets? How does it play out in non-tech industries? How does it play out for people who are or are not university graduates? And then what's this convergence of people applying for and competing for entry-level jobs who maybe shouldn't have them? And so that's that example of how did we take a big trend and then figure out the missing piece and then zoom into the missing piece.

14:58Andreas Munk Holm:And this was a lot about the scan phase. So to say you can often say that there's a scan, the build and the scale framework. Yeah. And you've built the innovation foundation exactly around this framework. Could you just explain to everyone what you mean by that scan we've gone deep into now, but then if we go to build and scale, what are those parts? So we set up basically a three-part model of the social innovation lab. The scan is the research part, which is where we use multiple types of data and our methodology to narrow down what is the problem that needs to be solved. Who is being excluded from the world of work and why?

15:33So that gives you the problem to solve, the demographic, and the geography in which to solve it. Because you have to solve for something more specific. You have to do something place-based, but always with the idea that it's going to skip. So we don't want to do something that's so narrow in that one place that it would never scale anywhere else. So once at the end of the scan, then we know what we're solving for and for whom and where. Then it goes into build. And build is the ideation and building of the real solution part. So it starts with ideation. We run a set of design workshops with a multi-stakeholder working group.

16:05You don't want an echo chamber. You want lots of diverse voices. You want a really dynamic kind of exchange. So we run that design process and basically cherry pick the best things out of it, the ones that have the most promise. And we use that desirable, viable, feasible framework and then choose two, sometimes three things we want to build. Then we build venture teams around those ideas, those solution concepts. The venture teams act like little startups. So this is where it looks like a venture studio. Those are entirely funded by us. So we're not taking anybody else's money. They are also not incubating somebody who's got an idea or has a startup or has already started a project.

16:46Because the idea is that we've designed the parameters of what do we want to build that we believe is additive and isn't going to duplicate something that's out there. And so then we bring in someone who is an experienced founder, not to build his or her own thing, but actually to build our product. And then we staff the rest of it with secondies from the ADECO group and secondies from outside whom we bring in. And then we buy in expertise where we need. And so that little startup, that venture team, then basically runs through sprints as you normally would and gets to the point where they have something that is tested enough, that is a good concept.

17:19They pitch it. We say, yes, that looks like it should go forward. and then once they get to the point of a decent MVP, then it will go into scale. And at any point along those stages, then we could kill it and we have killed multiple things. But if you've got a really good MVP, it's been proven in the market and we believe that it will live eventually, then you can go into scale. And scale is where you scale it in the geography and with the audience that you built it for. Then you have to scale in both directions, new audiences, new geographies, and then you have to scale it out of the foundation so that it can live on its own.

17:53And it could be for-profit or not-for-profit, standalone. It could go into another organization, government, business, civil society, academia. We're agnostic as long as we're convinced that it actually will live and thrive and continue to scale outside us because we're not a donor. We're not a funder. We're not an owner. And so that lets us kind of be a flywheel. And it also lets us be purer about what it is that we're trying to build because we're not building for the ADECO group. We're not R &D for the ADECO group. nor are we R &D for anybody else. It's really R &D for society, essentially.

18:27Cynthia, some of the companies we've had on the podcast before, like Bosch and Rehau, they all mentioned this challenge around the build phase, finding the right talent to basically build on the scans you have done. Could you go a little bit more in-depth about how you go about that process? Sure. There have been lots and lots of learnings in this because I think this is hard. And we originally assumed that it would be really easy to find experienced founders who now wanted to do something purpose-driven who would come in and build with us. It's paid. So we're not asking anyone to come and do this for free.

19:09And we thought that this would be an amazing opportunity. Hey, you know, I get a salary. I get a budget. I get a staff. I get the safety net of a big company. I can come in and do this. But we found that actually it's really hard to find the right combination of skills in one person. Someone who knows how to build a venture and build something that will be sustainable, who knows how to work with a team and in matrix with other, you know, kinds of service providers or, you know, the ADECO group being one of the stakeholders, and who understands the social impact side. So what we tended to get was either people who had been kind of pure for-profit founders who then needed to ramp up much more on the social innovation, social impact side, or people who had been social entrepreneurs who maybe had to ramp up on the other side.

19:57We found it very hard to find the perfect match. And in the end, you're never going to have a perfect match. So what we need is good matches and then an array of other people around them because you can't expect one person to be able to do all of that. And I think on a slightly smaller scale, it's the same issue in finding the fellows. is you need people who understand enough of business and social impact to be able to be light and flexible and kind of all hands on deck. What do you need? I can figure out how to do that. How do you, you know, what else do we need outside that? Let me go find that person or that resource.

20:33It's a different mindset and you can't expect that everybody has it already.

20:36Andreas Munk Holm:How have you solved it now? So two things. One is we needed a better way to source people and we needed a better way to train them in. So we've actually set up something now called the Foundry. And the Foundry is our new training mechanism. And it's all run by the Innovation Foundation. And it actually serves two different purposes. One is it allows us to open source and share what we do and how we do it, which is part of our mandate as a foundation, make it available to anybody. You can take it, you can use it, you can adapt it and use it to drive your own social impact. But then kind of on the back end, it also allows us to train people more specifically and in a hands-on way into our methodology so that then they can use it more effectively and so that we can see them doing it live.

21:29And so that allows us then to then talent spot people who would be good either now or in the future as a venture lead, as a fellow, as an expert. That allows us to then find more of the right people and to also essentially look for people with the right mindset and the right combination of ingredients that then we can kind of shape so that they can have a higher success rate once they're in the venture teams. Cynthia, when you started explaining about the kind of team composition side, are you leveraging the ADECO resources for that, the internal people, how do they flow into this setup? So from the very beginning, the first time we ran the venture teams, which actually started with the Youth at Risk project, we earmarked a certain number of them to be secondies from within the ADECO group.

22:21We started small and light with a small pool that we knew would likely have the right mindset. So we started first with people only from the International Future Leaders, which is the ADECO Group's High Potentials Program, and the CEO from One Month Program, which is that other kind of youth program that we used to run. So we pulled only, in the beginning, alumni from those two. Then when we ran the second one, we actually opened it up more broadly to people who were in all of the alumni of multiple years for those two programs. And then we also cherry-picked a few other people to put into the process.