6 Weeks of Runway to Near 8-Figure Enterprise AI SaaS

Omer Khan [00:00:09]:
Welcome to another episode of the SaaS podcast. I'm your host, Omer Khan, and this is a show where I interview proven founders and industry experts who share their stories, strategies and insights to help you build, launch and grow your SaaS business. In this episode, I talked to Bob Hyman, the founder and CEO of Sapia, an AI powered platform that uses smart chat interviews to help companies make better hiring decisions. In 2018, Barb was brought in to scale an existing HR tech startup. But within weeks, she realized the harsh reality.

Omer Khan [00:00:41]:
The product wasn't working and the business needed a complete reset. She made the difficult decision to fire the entire team, including the founder. And with just six weeks of Runway left, she had to quickly raise funding to keep the business alive. The next two years were a constant struggle for survival, and some months Bob wasn't even sure if they'd make payroll. She and a small team worked to rebuild the product from scratch and they conducted countless experiments to find the right approach and landing their first major customer.

Omer Khan [00:01:09]:
Qantas Airlines took a series of 15 trial runs over several years before they signed an enterprise deal. And just as they were gaining momentum, Covid hit, making it even harder for them to close new business. But Bob and her team stayed focused on building a product that their customers would love. And eventually their persistence started paying off as more companies began seeing the value in their approach. They won contracts with some of Australia's largest brands and much of their growth came through customer referrals.

Omer Khan [00:01:37]:
Today, SEPIA is approaching eight figures in ARR with a team of 45 people and they've raised over $21 million in funding. In this episode, you'll learn how Bob navigated the difficult decision from fixing to completely resetting a struggling startup. Why focusing on product excellence over sales and marketing was a strategic gamble that paid off. We talk about what led to their failed US expansion and how they found success in the UK market instead. How personal touches like handwritten Christmas cards to clients became their secret growth weapon.

Omer Khan [00:02:07]:
And why Bob believes that having clear AI regulations actually creates more opportunities than constraints in enterprise sales. So I hope you enjoy it. Bob, welcome to the show.

Barb Hyman [00:02:19]:
Hello. Great to be here.

Omer Khan [00:02:21]:
Glad to have you. Do you have a favorite quote? Something that inspires or motivates you?

Barb Hyman [00:02:26]:
I have so many. So one relevant to our product is In God We Trust. For everything else, we bring data. And another one, which came from my experience as a chro, where I really struggled to be an effective chro, was someone in my team coming to me and saying, I don't care what you know until I know that you care. And I had always been someone that focused on being clever and solving problems. And that was a very confronting quote from someone in my team.

Omer Khan [00:02:56]:
Love it. So for people who aren't familiar, just tell us what a chro is.

Barb Hyman [00:03:02]:
Chief Human Resources officer, otherwise known as a head of hr.

Omer Khan [00:03:07]:
Great. And tell us about sepia. What does the product do, who's it for, and what's the main problem you're helping to solve?

Barb Hyman [00:03:15]:
Yeah, the problem we're trying to solve is how do you know who are the right people to hire and promote in your organization? Because your people are everything in terms of your success. And the way that you do that today is typically biased when you're using people, it's suboptimal from a data perspective if you're using resumes. And most assessments are experiences that people don't want to have. And so we solve the problem of understanding people through language via the medium of a chat interview.

Barb Hyman [00:03:45]:
We're the first and I think still only smart interviewer in the world. So we're using TrueAI. We've been an AI native company since 2018 and through that capability, everyone learns. The candidate gets feedback and coaching through that experience. The recruiter gets smarter. The hiring manager gets talent insights. The organization gets a whole lot of data around diversity, candidate experience, efficiency. And so it's a kind of a full 360 ROI experience.

Omer Khan [00:04:16]:
Great. And give us a sense of the size of the business. Where are you in terms of revenue, customers, size of team?

Barb Hyman [00:04:23]:
Yes. So we're about 45 people. We're a global team, a virtual team, which by the way, I would not recommend. If I had my way, I would insist that we all come back together though that's very hard when you're around the world. We are getting close to eight figures. We have mostly big customers. We're an enterprise based SaaS solution. And yeah, we've been around since, or at least I've been doing this since 2018.

Omer Khan [00:04:51]:
And you've raised just a little over $21 million.

Barb Hyman [00:04:55]:
Nowhere near enough.

Omer Khan [00:04:57]:
Yeah, Give Bob more money.

Barb Hyman [00:04:59]:
It feels like we're very frugal every day.

Omer Khan [00:05:02]:
Great. So the business was founded in 2018. Sort of. There's a story about how this whole thing came about. I know your background was, you know, you're at Boston Consulting Group as part of one of your HR roles. So just tell us about, like how did this transition happen? How did the idea for this business come about?

Barb Hyman [00:05:28]:
So I was approached by someone on the board of A company called Predictive Hire, founded by someone else to scale the business. And I spent three weeks over Christmas doing the 100 day plan. And frankly, the whole problem they were trying to solve, which is how do you reimagine the hiring process for everyone, was one that I'd experienced firsthand as a head of hr. And I had a vision of how to do that. That was very experience based, very human, very intelligent.

Barb Hyman [00:05:55]:
And I came in thinking that I could help to build on what had already been established and grow from that. That was my first very big, naive mistake. I really didn't know anything about AI. AI didn't really exist at that point and I certainly didn't have a view of I want to do something with AI. I think that's the wrong place to start. I just had a really deep empathy for the problem and a deep understanding of the problem. And in essence, it was a reset.

Barb Hyman [00:06:21]:
I had to fire everyone, including the founder, which again, I don't recommend. That really should have been something that the board took care of and start again. And it was really hard. Everyone who loves me said, don't do it and walk away. But actually the hardest bit was not trying to retain a business and reset the business and hire again and build the product. It was actually the. The cap table. I didn't even know what a cap table was.

Barb Hyman [00:06:50]:
And fixing that took three or four years, much longer than it took to actually build the new product.

Omer Khan [00:06:59]:
What was the issue, just at a high level?

Barb Hyman [00:07:01]:
Well, I had to raise money in six weeks because we were running out of cash. Again, I was naive about that. The only way I could raise money was through a convertible note. Part of that was because there was a share class on the cap table called convertible redemption preference shares, which means if any investor puts their money in, that share class can take that money out if the valuation falls below a certain amount. And the business had been valued beyond what it should have been. So I was basically completely hamstrung.

Barb Hyman [00:07:29]:
And so when you raise money on debt, who owns the business? It's the people who started the business. So you had this group of initial founders who had all of the ownership rights. And I was working with investors who'd put money in and they had no ownership rights. So it was a real balance of control and power.

Omer Khan [00:07:48]:
And you said you had to kind of clear house. And were there any of the original founders still around when you sort of.

Barb Hyman [00:07:56]:
Yeah, yeah, they were. And that was really tricky. Is a massive understatement. Again, I wouldn't recommend that. It Was, you know, really emotionally very difficult. I remember at one point we moved from the original office that I had joined, which was not what you would expect from a startup. It was very ugly. It didn't feel startupy, didn't feel, like energetic and vibrant.

Barb Hyman [00:08:20]:
And, you know, I had a whole come to Jesus conversation with the people who were then at the team, still with the team, you know, you're gonna get on the bus or get off the bus. And made it really clear what my expectations were. And almost moving offices to something that was really, genuinely almost like a garage. A true startup environment was a moment to draw a line in the sand and say, it's gonna be a new culture when we move to this new office. And this is what my expectation is.

Barb Hyman [00:08:46]:
So, you know, that was quite sort of a milestone in the reset of the business, particularly of the culture.

Omer Khan [00:08:53]:
And then what year Was this roughly?

Barb Hyman [00:08:56]:
2018.

Omer Khan [00:08:57]:
Okay. And then did you kind of build off the existing product or as part of this kind of reset? It was like a brand new product as well.

Barb Hyman [00:09:08]:
We were in the recruitment space and the thesis was that data is a better way to evaluate someone if you can find an objective data set. And could we use an assessment to do that, and a data set from an assessment. I won't get too much into the science side of it, but in essence, six months in, I hired an incredibly smart person who had been the lead data scientist at cultureamp and had seen the value of language data and the signal about people that lives in language data.

Barb Hyman [00:09:36]:
And he could see that NLP was the future. And so I had a vision around chat and experience, and he had the brilliance around nlp. And we just started on a journey of experimenting for two years. It took us two years to build the product. I remember when he first came in, he did analysis on is there any validity in the existing assessment? There was none. You were better off not using it. It was a real moment as well.

Barb Hyman [00:09:57]:
So when we look back, I actually, I think my naivety helped me because if I'd truly known how bad things were, I probably would have left. But once I'd raised money and. And then once I hired Booty and then he hired Johnny, who became our head of engineering. You know, incredible individuals. All immigrants were all immigrants. It was a very strong aspect to our culture, which is this feeling of not enoughness and this need to kind of prove something to people.

Barb Hyman [00:10:25]:
I couldn't leave because at that point I then made a kind of an emotional and moral commitment to go, okay, we're in this Together, we're going to figure it out. And frankly, the learning journey, you know, you have to really want to learn and constantly learn. Like, it feels as if there's no space that's been more dynamic in terms of learning and technology innovation than the space that we're living in.

Omer Khan [00:10:46]:
So how did you figure out, number one, what product to build? Did you go and talk to customers or potential customers? Did you kind of research?

Barb Hyman [00:10:57]:
I have my 100 day plan. You know, I had a vision of what it would be. And you know, it's funny, you go back to it and so much of it has been built, which is incredible. Makes me feel really proud and lucky. And so I knew what I wanted it to be. I wanted it to be chat. I wanted to be something that people felt safe, that they felt heard and understood.

Barb Hyman [00:11:15]:
You know, the need for people to feel like someone has noticed them when they're applying for a $10 an hour job is really huge. And so I wanted to solve for that. I wanted to ensure that everyone got something back, that no one was ghosted. I wanted to get rid of the resume. You know, I could see in all of my jobs that the resume is a recipe for bias. I didn't want to have video interviews, I wanted to be untimed. So I had very strong convictions around the experience.

Barb Hyman [00:11:42]:
And so it was very much experience based. And then booty had to figure out how to do it from a science basis. And that's really what drove us. But everything we did was an experiment. You know, I remember the first time when we pressed the button on this automated feedback for candidates and everyone freaked. You know, I even remember speaking to AWS and other tech companies in the US and they said, no, no, we don't like to share anything with candidates. We want to keep them in the dark.

Barb Hyman [00:12:08]:
Creates too much legal risk to share anything. And so there was this incredible aversion to giving people feedback. Like that was a massive risk. When Qantas went live with that, I remember it was a risk when we said it's not going to be timed. And people assume that candidates cheat. This is before GPT. Everyone thinks that people are cheating all the time. But actually when we started to flag that because we could identify plagiarism, we saw the cheating rates were very low in most geographies.

Barb Hyman [00:12:35]:
And so we were constantly challenging every aspect of how people thought about how you hire, what a assessment needs to look like, what the experience of the candidate needs to look like. And I had very strong principles that guided me around transparency Explainability, fairness, experience. And I just stuck to those. So our product principles that I formed year one are still the product principles that we have today. And that has guided us in terms of how we've built additional products in the suite.

Omer Khan [00:13:06]:
Right. But I guess I don't want people listening to this and saying, oh, great. Well, Bob just came up with his 100 day plan and then just executed on it. I think in many ways you were able to do that because you had spent so many years in this space actually kind of being on the other side of this. Right. In terms of hiring.

Barb Hyman [00:13:25]:
And, you know, I think that, you know, I don't know how to sell. Right. All I know is your pain as another HRD or a head of talent. And so I was successful in selling because I wasn't selling. I would go in and I probably wasted way too much time. In fact, not probably. I definitely did. I was in love with the product. I could see that candidates were in love with the product and I just wanted you to feel it and have it.

Barb Hyman [00:13:56]:
But the process of actually bringing in technology like this is incredibly tough for an organization. This isn't buying IBM. You can get fired for this. This is AI. This is really scary stuff. And so because I wasn't selling and I just led with empathy and curiosity and I was helping you solve the problem that was sitting in front of you. That's how we were effective.

Barb Hyman [00:14:20]:
And that has continued to be to our benefit Today as we go into our next phase of innovation where we're starting to bring generative AI into the product, that's the next hurdle to overcome. Because again, there's another set of fears that people now have around using Gen AI in the HR process. So I've always been passionate about enterprise empathy. Like really, you know, don't celebrate when you do the deal, celebrate when they see the value. Because as soon as you sign the deal, they're terrified. Now they have to go.

Barb Hyman [00:14:51]:
And you know, sometimes our deals are seven figures, right? Like, it's a lot of money. And so you have to make sure that you're giving value and they see value before you start waving the flag and ringing the bell from a sales perspective. So I think that's always guided a lot of ways in which we sell. We market, you know, we don't do standard marketing. Most of our opportunities come referrals.

Omer Khan [00:15:14]:
There's kind of an interesting thing I'm thinking about. Like today you can avoid AI right? Everywhere. It's just in your face. But four or five years ago, it was Kind of more like, yeah, people talked about AI but nobody really understood what that meant.

Omer Khan [00:15:34]:
And when companies were talking about using AI, maybe they were using machine learning and a bunch of stuff like that, but it was, I think it was an even harder sell than it would be today because at least people are now kind of opening up to it and there's still fears and so on. But I'm guessing back then, one, any company using AI, there was a lot of hesitancy. And secondly, for hiring people, that sounds really scary when you're already terrified about legal risks and the kinds of things you might be telling people.

Omer Khan [00:16:09]:
And then now we're going to put something in between us that's going to kind of God knows what it's going to be telling these people 100%.

Barb Hyman [00:16:16]:
And I love to say that it's got easier, but actually I think it's just as hard. It's interesting, I remember an investor around the table saying, I think it was four or five years ago, you know, Bob, we're going to get that return on investment and education the market, eventually it'll cotton on. And even when GPT came on, you know, it's very easy to get a meeting right. Everyone wants to hear about AI and is curious to learn, but it's a different thing to actually get a deal.

Barb Hyman [00:16:44]:
So a couple of decisions we made really early on was that because we're advancing the field of machine learning in what we're doing is that we were going to publish research. So we've had an R and D team from the beginning and that wasn't delivering returns for the first four years. No one came to us and said, oh, I saw your amazing paper published in Frontiers, I want to meet with you. Right. But now I used to say that our moat was data and our moat was the algorithms and what we've done with the data.

Barb Hyman [00:17:13]:
But actually I think it's our brand based in science, peer reviewed published papers that has given us a level of credentialing and credibility that is really now separating us from other, you know, the myriad other AI companies. So the other decision we made early on was that we were going to invest in publications, which is something that Meta might do. And they're a massive organization. We're a tiny business of a very small labs team. But we have pushed out three or four papers every single year.

Barb Hyman [00:17:45]:
And you know that that has really helped build our brand. Right.

Omer Khan [00:17:50]:
Was there any fear about like, you know, giving away the secret sauce by being kind of so much more open about what you were doing?

Barb Hyman [00:17:59]:
Absolutely. But we felt that if we wanted people to trust us, we needed to be transparent around the signs. And we had a very strong view that unless you had the data, you couldn't do what we did. Like, that's still a unique data set that we have.

Barb Hyman [00:18:12]:
You know, one of the things I remember in terms of research very early on when we were trying to validate this thesis that language had signal around people is one of our team members scraped 2 million tweets and ran an experiment based on how people self declare their personality profile to see if there's a correlation. We actually wrote a paper, this is way before Elon Musk, around the fact that you can't use third party data and generalize it in this context, because one, the way you tweet is not how you interview.

Barb Hyman [00:18:40]:
The second is Twitter is not representative of the general population, so you risk adverse impact. And we actually made a decision not to publish it because we were worried that even then Twitter was attracting a lot of kind of attention, that it would attract the wrong kind of attention. So, you know, there were choices like that we made where we just, you know, we wanted to make sure that we protected the brand in terms of what we were prepared to share externally.

Barb Hyman [00:19:04]:
But we've been incredibly transparent around the science and the various algorithms that sit underneath the technology.

Omer Khan [00:19:13]:
So you mentioned earlier that it took you two years to build the product. How long did it take to land the first customer? Was it like you built for two years and then you went out to try and sell it, or did that happen in parallel?

Barb Hyman [00:19:26]:
Look, we were just flying the plane. We didn't even have a plane. We had like a seat and we were trying to just stay alive. You know, every year my family would say to me, you know, what's gonna happen next year? And I said, look, I'm just trying to get to the end of the year. And I really had no conf. Not so much, no confidence, but I couldn't forecast if we would be around for another year. It really felt like we were going month to month.

Barb Hyman [00:19:52]:
And there was definitely a couple of months from a payroll perspective in the really early days where if we didn't get that customer to pay that invoice, you know, we wouldn't have able to meet payroll, which is your fundamentals. So it was really scary. And you know, I was doing payroll, which is never a good thing. And, you know, you're kind of doing everything in those early days. And I look back and, you know, buddy and I will often go and we do our all hands in person and we'll look around and we'll go.

Barb Hyman [00:20:17]:
It's a miracle. It's a frigging miracle that we're still here. So, you know, what we had to do is find customers with serious pain. And Qantas was our first customer, which is kind of mind blowing because they're such a brand and they're a very conservative brand, but that shows you that there is when you find real pain. And they were using our major competitor and they really didn't like the impact it was having on their brand and their candidates. So they wanted to find a different way. And there was nothing else in market.

Barb Hyman [00:20:47]:
And so we did 15 pilots in a row before they signed a deal. So we absolutely had to find customers. But we worked out pretty quickly the target segment that we were going to go after. What we haven't done well is stay really focused on that segment. We've probably stretched too far just to optimize for customer growth rather than the right customer growth.

Barb Hyman [00:21:10]:
And what matters in terms of what we're optimizing for has fundamentally changed, really, I think over the last 12 months, partly because of the tech downturn, but also partly because, you know, we've become much more disciplined about what drives our success and we've got to control that rather than just trying to go for, you know, let's add on the next 10, 20, 30 new customers.

Omer Khan [00:21:34]:
So you're, you're, you know, back then, you're a startup, you're building something which is disruptive, you don't have any customers, right? And then how do you get your foot in the door with somebody like Qantas?

Barb Hyman [00:21:46]:
Well, you, you, you, you, you know, you do what you do now like it is water, it's door to door combat, right? There is no shortcut in enterprise, right? And the thing that has served me really well is that I'm just relentless. I'm relentless. So, I mean, I will send 50 LinkedIn messages a week today. And I'm doing this, I've been doing this for eight years. I'm not waiting for a salesperson to go hunting because there's a completely different response you get when it's the CEO. You know, I can tell the personal story.

Barb Hyman [00:22:20]:
And so for me, you are fighting hand to hand combat every single day. And what you have to do is focus on those businesses that are firstly, big, big pain, right? In the case of Qantas, you can't get your planes in the air unless you've got the talent. And it takes so long to hire people because you're going through security, etc. If you can do anything to reduce the time it takes to actually identify the right talent, that really makes a meaningful difference to the business.

Barb Hyman [00:22:47]:
So you kind of think cleverly about for whom are we going to be absolutely core to running their core business, Right? And that's where we've made a mistake, is if we're just marginal, if we're kind of good, but not absolutely necessary, the risk of churn is too high. That's one thing we've really learned this year. We've had a lot of churn from smaller businesses where they can survive without us. Now we focus on only selling to businesses where they need us to survive. And so that was the first thing.

Barb Hyman [00:23:16]:
The second is we went for brands, consumer brands, trusted consumer brands, because we had no money for marketing. And so if we can say, wow, companies like abcd, who are the five most trusted consumer brands? Trust Sapia. It's very easy to get the next 10 brands. So we went with a deliberate strategy of going for brands. And the fact is, brands really care about the things that we solve for. So there's a high overlap between their problems and their desires and what our technology does.

Omer Khan [00:23:50]:
Now, you mentioned 15 pilots with Qantas. How long did that take? And I'm just wondering, do you get to a point where it's just like, are we just going to be doing pilots forever? Are they ever going to make a decision on this thing? Or did you feel like, okay, every time we do something, we're getting some good feedback, we understand what are the concerns we need to address. I'm just curious, like, it feels like a nightmare just drags on forever.

Barb Hyman [00:24:14]:
I mean, at that stage, it's just default alive, right? You know, it still is default alive. You're just trying to stay alive and you're seeing startups collapse all around you. So, you know, I think the other thing with a big company and procurement hasn't gotten easier is they absolutely wanted to do a deal. But Covid stepped in, became very hard to sign deals.

Barb Hyman [00:24:35]:
Obviously, everyone's in austerity measures and frankly, we were just happy to keep going and get some cash and have the ability to use them as a reference customer, and they're still a customer, you know, six years down the track, which is, you know, amazing. And I also think there's a kind of a pay it forward, which is fundamentally people are taking a huge bet of their own brand. Right? Like, you're putting your brand at risk as a buyer when you're using AI unless you're using Microsoft Copilot. Like why has that been so successful?

Barb Hyman [00:25:11]:
Because no one's going to be fired from using Microsoft Copilot even if it doesn't work. But you're bringing in a startup and you're using AI around hr. You got to really respect the trust that they're putting in you. And so for us we have always, you could argue overinvested and over delivered and overserviced because you kind of pay it forward. And what we have now is a pipeline of opportunity that is almost completely coming from our customers unsolicited most of the time. And so it's expensive but it's in the long run.

Barb Hyman [00:25:47]:
I haven't done the economics but it's probably a lot cheaper than spending on marketing which I'm sort of less confident delivers result.

Omer Khan [00:25:54]:
And that's because you don't buy into marketing generally or because just the complexity of trying to sell this type of product to enterprise.

Barb Hyman [00:26:02]:
Well, I think LinkedIn is a very expensive channel. Right. And you know, if you think about as a startup founder, you have to make choices about where you're going to invest your dollars. And I can't have someone brilliant in marketing, in sales, in, you know, we've got three product teams, we've got a traditional product team, you've got a people science product team, you've got a, you know, machine learning data science product team. Like that's expensive.

Barb Hyman [00:26:25]:
If you're not an AI business, you don't have that, you have just one product team and then you've got your normal engineers, your machine learning engineers, your DevOps, your security engineers. Again, you know, there's quite a lot of infrastructure there. So I've always felt because I came in and I inherited a product that didn't work, that I was obsessed about building a product that worked. And so I really over indexed on anything to do with product, not to do with go to market. Now that could have been a mistake.

Barb Hyman [00:26:52]:
But now we have a product that people love and have been using for a long time and our NRI is getting up to 110, 115%. Right. As a reflection of that. So my choice was that I'm not going to invest in someone who really knows demand gen or who knows marketing because I hope that the product and then the customer experience will kind of sell itself in market. So those are the trade offs you make is you can't have brilliance in every part of the business. And I wanted to have brilliance in product.

Omer Khan [00:27:22]:
How long did it take to close the Qantas deal from start to look,

Barb Hyman [00:27:27]:
I think the pilots, I'm trying to remember now, but it's a good few years until we got a enterprise deal with them. Typically now it's around six months to nine months to close a deal from beginning to end. And what's interesting is one of our largest customers is the largest retailer in Australia. And we did a pilot and we were about to convert and then they hired an AI ethics team and then everything went into a whole different world of pain from a procurement perspective. And we weren't prepared for that.

Barb Hyman [00:27:59]:
I think they weren't prepared for that. And that's a challenge right now, which is we're often the first AI technology that a company has procured and so they have to actually build their own AI governance framework, the AI governance committee, their processes as they're procuring us. And so that's adding on additional complexity. Now we've done a lot to educate them to find ways in an independent fashion, help them understand what this new world is all about. And, you know, but that's, that's really challenging, right?

Barb Hyman [00:28:30]:
When you're selling in technology like this, you're, you know, you're helping build their maturity around AI at the same time as trying to do a deal.

Omer Khan [00:28:40]:
Okay, great, so you got Qantas and then presumably it gets a little bit easier because at least you have one case study you can talk about and some happy customers. How long did you just focus on the Australian market before you decided, okay, we need to expand out, we need to look at places like the U.S.

Barb Hyman [00:29:03]:
so look, we had UK customers that we managed to get just through being in Australia. I think there's a great connection between Australia and the uk. And you know, we had inbounds and so we were able to do deals from Australia in the uk. So we built up a good portfolio of UK based customers without being on the ground. But you know, it's interesting how politics plays into your thinking around where you should grow next.

Barb Hyman [00:29:26]:
And so we did our Series A and the thesis was we're going to go into the us and my view around that was we're a deep tech company and if we want to be acquired and if we want really solid capital and solid valuations, we need to be where all the smart money is and the smart people are. And that is 100% in the US. It's not in Australia. And so I hired a team and we went and tried to make it work and it didn't work.

Barb Hyman [00:29:53]:
It didn't work because one again, this comes down to I'm not a seasoned founder, I don't know what an amazing CRO looks like, and I've never expanded into the US So I trusted my lead to own the strategy around that. And the strategy that he proposed was effectively spray and print. Let's just go after any sector and see where we can get a hit. And then that will help us learn which part of the segment, which part of the market's going to work.

Barb Hyman [00:30:24]:
When we knew that we had a product market fit in certain verticals, we should have. And I should have said, you know what? No, we're going to just go with these two and we're going to push into the market and we're going to go with for these 100 clients. So that didn't work. I think the other thing that didn't work is a lot of people fancy themselves as an entrepreneur and think that they're a founder.

Barb Hyman [00:30:43]:
And that's effectively what you're building when you're going to the US like you're starting again because no one cares about your Australian testimonials or your UK testimonials. And you really need to do hand to hand combat. You need to go and meet Omer in San Diego in person because crikey, you got a meeting. You're gonna go and put on your best clothes and make a great impression. You know, I'm sort of really obsessed about salespeople going and meeting people in person. Right. Because you're building a relationship of trust right from the get go.

Barb Hyman [00:31:15]:
And he just wasn't able to do that. And the disconnect between what he thought he was and how he actually played the role, it was very much a standard playbook. I've done this before at a larger business. I'm gonna do it again. SDRs, outbound machine, multi threaded sequences. It didn't work. And it's also very expensive because you're raising money in Australian dollars, but you're paying for salaries in US dollars.

Barb Hyman [00:31:37]:
And the other thing that happened is that DNI was a tailwind and then it became a headwind in the period of time that we had landed in the US So I pulled out of the US and we decided to make the UK and Ireland our second Geo because we'd already seen good success here. And so I'm here now and that's very much where we're seeing quite amazing growth this year. We're about to hit a kind of an amazing threshold of winning some incredible logos.

Barb Hyman [00:32:01]:
A bit like where we were in Australia when we won Qantas and Woolworths and Bunnings, et cetera. And it's been a lot easier to be successful here than it was in the us So I think you've got to have really deep pockets of the U.S. you've got to be hyper focused. And fundamentally, I think I should have moved there. Right. And I didn't want to move there. I hired someone and that was a mistake.

Omer Khan [00:32:24]:
How long did you try to get some traction in the US market before you eventually pulled the plug?

Barb Hyman [00:32:32]:
Too long. Too long. I should have done it within 12 months and instead I gave it a good 18 months. In fact, I think every single decision I've made, that was a hard decision. I made it too late. I should have made it sooner.

Omer Khan [00:32:46]:
Yeah, I mean, that's kind of easy to see in hindsight, but when you're in the middle of that, I mean, it's really hard, isn't it?

Barb Hyman [00:32:54]:
It's like, well, you know, you're a small team, right? And so when, when particularly salespeople. I have found hiring salespeople incredibly difficult, ironically, for a business that's in hiring, because they're so good at selling themselves. And I tend to fall in love with people. You know, I love the product. If they love the product, I'm like, wow, you know, we're in this together and I just, I can't see through the bullshit and. Because I don't know what great sales looks like, because I'm like an artisan seller.

Barb Hyman [00:33:27]:
And the other really successful salespeople we've had are artisan sellers. They're not machines. Right. They haven't grown up in a world of salesforce or, you know, rippling, et cetera. And so I don't know how to see that. And we don't have the funding and the capital to build a machine. Right. We don't have 10 SDRs and 18 AES and, you know, et cetera. And so I've really got it wrong there. And I wish I'd let people go much sooner.

Barb Hyman [00:33:54]:
I wish I'd been more ruthless around sales because there's no one who started average and ended up being anywhere close to great. And the thing is that when you let someone go, it's very visible in a small team. And we're quite a. We're a really close, connected team. It's a very strong culture. It's, you know, there's real heart to it. And, you know, you can't sometimes, even when someone's not selling it still Feels depressing to the team to let someone go because they're really fond of them.

Barb Hyman [00:34:26]:
So it's, you know, that's been really hard to make those calls and I've made them all too late.

Omer Khan [00:34:31]:
Well, I appreciate your transparency and being so honest about that. Tell me a little bit about when you went into the UK market. What were some of the lessons that you learned from the US that you did differently? So you mentioned you moved out to the uk. Did you then start to focus on. Okay, we're going to kind of focus on the same verticals that we've got traction in Australia. What was the approach?

Barb Hyman [00:35:01]:
Shrink the tam, Right? Just narrow, narrow, narrow. Yeah. Just be confident about where you've got product market fit. Recognize that the size of that market in UK is 10x what it is in Australia, at least, and there's plenty within that. And just stay focused on that and forget the noise outside of that. Right. So again, we have a lot of companies that come to us and say, I love what you're doing. This real differentiation, right? The fact that you're interviewing everyone, you're giving everyone feedback. Who doesn't want to do that?

Barb Hyman [00:35:39]:
We're not ready for you, or you don't have the right level of AI maturity. So lose fast and actually really make the decision after the first session and even cut that first session short, which is to say, Omir, it doesn't sound like your business is at a maturity level to use AI. One of the questions we ask really early now is when, where is the business? On a scale of 10, they're like, hell, yeah, I want to use AI and I've got a C suite person who's going to sponsor that to zero.

Barb Hyman [00:36:06]:
We haven't even started. You know, we'll just opt out at that point. Whereas, you know, I'm a kind of. I don't like to give up and I've wasted way too much time going. But I know that you guys will really love this and it's really going to make a difference. I'll help you through that. We don't have time. Like, time is your most critical asset. My time. I spend so much time now thinking about how I spend my time and being really conscious and thinking, this year I'm going to focus on this.

Barb Hyman [00:36:35]:
This month I'm going to focus on this. Right. Because that is what I have not been in control of nearly as much as I should have. And that is the most important asset that you've got to manage across the business, starting with the CEO.

Omer Khan [00:36:51]:
So when you talked about the going into the UK market. Did you find that? I mean, I talked to a lot of founders who are international and they talk about, I want to get into the US and you know, there's a lot of different things that you have to overcome, but culture, the cultural difference is also pretty huge. Did you find that for an Australian startup, going into the UK felt culturally a little easier maybe?

Barb Hyman [00:37:21]:
Yeah, look, I mean, you know, Americans love Australians and you know, there's a great sort of camaraderie. We're all part of the same sort of set of Western democracies, but definitely the connection to the UK is a lot stronger. And in the sectors that we focus on, there's a lot of talent that crosses over between Australia and the UK from a C suite level.

Barb Hyman [00:37:41]:
So all of the businesses that we work with in Australia that are incredibly well known, respected brands are known and respected in the uk, whereas in America, they don't care, they don't know about them. And there's a very high level of American exceptionalism. That means unless you've got an American logo, we're really not that interested. Right. You don't find that in Australia at all. I think the other thing is, frankly, the regulatory environment here with GDPR has built a maturity and a fluency around regulation that the US does not have from an HR perspective.

Barb Hyman [00:38:14]:
And the EU AI act that comes into play in the middle of 2026 is so much easier to navigate because it's a centralized regulatory framework. Once you go into the database, you're done. Whereas in the US you have to navigate myriad different states, all with their own regulations. It's very complex and very costly for a business. So I think America is actually doing itself a disservice. They've got state and federal because they're not creating a simple, you know, one country framework and they don't have the maturity to be honest around managing AI procurement.

Barb Hyman [00:38:48]:
Whereas here, because of gdpr, there are lots of, you know, everyone knows what a DPIA is, right. And then they go into an AI audit and they know what that is in the us, just that maturity and fluency is not there, at least in HR in my experience.

Omer Khan [00:39:05]:
Right. So it's basically like companies in the UK are willing to take a little bit more of a risk because they feel like, okay, if you're at least compliant with GDPR requirements with the AI, at least I have some guardrails, I feel a certain level of confidence.

Barb Hyman [00:39:22]:
Well, regulation creates certainty. Right. Like, bring on regulation is what I would say. You Know, the lack of regulation creates uncertainty and then your legal team will just veto it and say, no, we're not interested. You know, I remember in the early days, we actually pitched to Apple in Australia and we got as far as Cupertino. Now, this is back in 2019, and at that stage, Apple in Australia, their retail store, they were getting 50,000 applicants, seeing really high turnover.

Barb Hyman [00:39:48]:
They love the idea of using a technology that's a true learning technology that retrains models based on turnover data. And the legal team just said no, they weren't prepared to use anything like machine learning. This is way before GPT, right. And when AI became really trendy and even when Apple was using a lot of AI and they were just not prepared to go there. And it was a decision made purely by the legal team.

Barb Hyman [00:40:09]:
And that's the other thing about the us, the legal team effectively have veto, whereas in the uk, they're a stakeholder, but the business really owns and makes the decisions on advice of a legal team.

Omer Khan [00:40:19]:
I think it's also really interesting that sort of conventional wisdom is if you're in the early stages, you're trying to get some traction first 10 customers, whatever, focus on your early adopters and the fact that you are now, you know, six, seven years into this business and you still talked about how you're asking these questions and effectively you're identifying your early adopters.

Omer Khan [00:40:45]:
So people who are, like, either comfortable with the technology, they're fully committed to doing this thing, so you're not wasting a whole bunch of time trying to persuade people that this is good for them.

Barb Hyman [00:40:59]:
Yeah, we did that in the beginning, but we don't do that anymore. Now we're more selective.

Omer Khan [00:41:03]:
Just explain that a little bit. Like, what do you mean?

Barb Hyman [00:41:05]:
So in the beginning we would invest to get people comfortable with it and to educate them. Today, we've got enough brands that we don't need to sell to you. If it's going to be a really steep road, you know, an uphill battle to get this real business right, we'll move on to the next one who's more AI friendly and got more AI maturity. I think the other thing I'd say that has really set us apart is it's very personal. Like, I feel what we're doing is very personal.

Barb Hyman [00:41:36]:
So, you know, I write Christmas cards to every single one of my clients and our clients every year with a personalized message. We've just hired a salesperson who's American, actually living in the uk, and she's brought us this wonderful new Tradition of gift giving. And what she's done is, you know, every time we have a meeting, she brings something, she might beg something, she'll bring chocolates. But we've now got a gift register where if you can't accept gifts, we'll give a donation on your behalf.

Barb Hyman [00:42:04]:
But say you give us a referral or we had a meeting and you helped us out, we'll send a gift. And it's people just are so grateful. Right. So we don't just ask Qantas to give a referral and then go, thanks on an email. We'll send a gift, we'll send a note, thank you so much. It makes such a difference. And so there's something very personal. I think, you know, one of the things that people say when they join us is the level of intimacy we have with our customers is really noticeable.

Barb Hyman [00:42:32]:
And I think that comes. Maybe it's because I'm female and I'm created certain traditions in the business, but it's a different way of connecting with your customers.

Omer Khan [00:42:42]:
Yeah, yeah, I love that. It's kind of ironical. On the one hand, it's an AI product which doesn't have the personal aspects to it, yet the culture of the company is almost the other end of the pendulum in terms of. That's kind of so core to, I guess, what you have to do. Which is why, from what you were telling me earlier, that a lot of new business is coming from referrals from your existing customers.

Barb Hyman [00:43:13]:
Well, look, the product does have a very strong human element to it. So that's one of the big objections we've got, which is, but we're removing the human touch if we introduce AI. And what I say is, well, you're getting 100,000 people who reply and you're hiring 1,000. How's that working for you? Are the 100,000 getting a human touch? They're not. Humans don't scale. And this allows you to at least give everyone a fair go, to give everyone the chance to learn about themselves and know that they've been considered.

Barb Hyman [00:43:43]:
And if they don't get this job, it may help them get their next job. And so we feel like we're humanizing the process. And one of the things, again, that I've been obsessed about is language. So I learned from my last boss, when I was a chro, who was an ex Microsoft executive, she was obsessed about communications. And you can send an email out from her that she didn't review really closely. And she was all about authentic comms. And so I Speak in a very human way. You know, I'm not a polished.

Barb Hyman [00:44:09]:
Like, even my hair is always messy. And I wanted to bring that realness into the product. So in the early days, you know, I wrote all the questions, I wrote all the scripts, all the dialogue. And today, you know, I'm fanatical with the team about this is somewhat of our product that's a bit unscalable. Is that when you're creating this experience of, you know, ome, thank you for applying for a job at Starbucks. This is what we care about. Take your time, you know, like you have to.

Barb Hyman [00:44:35]:
It's curated, it's created by our CS team, that dialogue. But it's gotta feel real, and it's gotta feel real to their culture. You know, it's not a. Fill out these 20 questions and we'll let you know. So, you know, that aspect of the product, which we haven't figured out how to scale yet has very much come from me and I think is very much in part why people love it, is. It just feels more human, it feels more authentic.

Omer Khan [00:44:58]:
Love it. All right, we should wrap up. It's getting late. And you've had quite an adventure over the last day or two in that hotel, which. Why don't you just quickly just tell

Barb Hyman [00:45:13]:
us about that, My hotel story. So, Yeah, I arrived 5am yesterday, first night, 11:30 at night. I've taken my sleeping pill because I rely heavily on sleeping pills when I'm traveling internationally. And there's just this dripping in my hotel room. And I'm thinking, did I leave a tap on somewhere in my exhaustion? And then I turn on the light and there's just water coming down from the ceiling from the lights, on the floor, puddles, freaking out.

Barb Hyman [00:45:39]:
So I took a video and sent it to my husband, and he said, get out of that room and find a new room. So that was my adventure.

Omer Khan [00:45:45]:
God knows what was happening upstairs. All right, let's get onto the lightning round. So I've got seven quick fire questions for you. Just try to answer them as quickly as you can. Ready?

Barb Hyman [00:45:56]:
Go for it.

Omer Khan [00:45:56]:
What's one of the best pieces of business advice you've received?

Barb Hyman [00:46:00]:
Don't overthink it.

Omer Khan [00:46:01]:
What book would you recommend to our audience and why?

Barb Hyman [00:46:05]:
Amp it up by Frank Slotman? I think it is. I just think he encapsulates everything that I wish I was as a leader and wish I could be as a leader. I've gifted it to many different people.

Omer Khan [00:46:19]:
What's one attribute or characteristic in your

Barb Hyman [00:46:22]:
mind of a successful founder Madness, Relentlessness.

Omer Khan [00:46:27]:
What's your favorite personal productivity tool or habit?

Barb Hyman [00:46:30]:
Tripit, actually, because I do a lot of travel, I really love tripit. Just to keep track, because I'm very badly disorganized and I don't have a PA or anything, so. Yeah. And I've got to say, I use the clauds and the like every day.

Omer Khan [00:46:45]:
Of course.

Barb Hyman [00:46:46]:
Yeah.

Omer Khan [00:46:47]:
What's a new or crazy business idea you'd love to pursue if you had the time?

Barb Hyman [00:46:51]:
Well, everyone in the business, when we ask them this at orientation, says they wish we could take our capability and apply it to personal matching. How do I figure out whether or not you're the right person for me? So, extending it into dating.

Omer Khan [00:47:06]:
What's an interesting or fun fact about you that most people don't know?

Barb Hyman [00:47:10]:
I actually was in the Israeli army for a period.

Omer Khan [00:47:13]:
Wow. And finally, what's one of your most important passions outside of your work?

Barb Hyman [00:47:20]:
Well, my kids, my husband, my dog, and gardening. I actually love just being outside in the garden. Not that I do it very much, but yeah.

Omer Khan [00:47:30]:
Awesome. Bob, thank you for joining me. It's been a pleasure. Love sharing the story of you building the business. I love just how open and authentic you've been and willing to talk about your successes, but also, you know, the challenges and the mistakes. And I think that's often where, you know, the rest of us can. Can learn from. So I appreciate that. If people want to check out Sepia, they can go to Sapia AI. And if folks want to get in touch with you, what's the best way for them to do that?

Barb Hyman [00:48:01]:
LinkedIn.

Omer Khan [00:48:02]:
Great. We'll include a link to your profile in the show notes. Great. Well, thank you so much. Thank you for staying up late. Appreciate it. And I wish you and the team the best of success.

Barb Hyman [00:48:12]:
That's a pleasure. Thanks.

Omer Khan [00:48:14]:
Cheers.