[EP 265] The Evolution of Flippa.com and the Power of AI in Matching Buyers

Transcript below!

Download the PDF transcript of this episode

Note: This has been automatically transcribed so will be full of errors! We are not providing it to you as a word-perfect version of the podcast but just as an easy way to provide you with a different way to be able to scan for information that might be relevant to you.

Joanna:  

Hi, it’s Joanna Oakey here and welcome back to The Deal Room Podcast, a podcast proudly brought to you by our commercial legal practice, Aspect Legal now today get ready to be swept away in another insightful episode of our much loved innovation series. Let’s face it, innovation is the beating heart of every successful business. And in the world of business broking, an industry that shapes the lives of entrepreneurs and business owners as they navigate the thrilling realm of buying and selling businesses. Innovation becomes an absolute game changer. So stay tuned and buckle in for another innovation series episode right here at The Deal Room Podcast. Here we go.

Blake a huge welcome to The Deal Room Podcast. It’s absolutely fabulous to have you on the show.

Blake:  

Thank you so much, Joanna. It’s nice to be here and looking forward to the chat.

Joanna:  

Brilliant. Okay, wonderful. Now look, um, I guess before we get into some of the topics that I really want to dig into today around innovation, why don’t you just kick it off by giving us a little bit of backstory? So when did Flippa start? Why did it start? What was the opportunity that was in the market that it was looking to solve?

Blake:  

Yes, so Flippa is a lot older than most people think. So Flippa has been around for nearly 15 years, believe it or not, and Flippa is a marketplace to buy and sell online businesses. So imagine 15 years ago, how many online businesses there were and how many had reached a stage of size and traction that mattered to investors and potential buyers? The short answer is probably not many. But the founders of the business were ahead of their time. And the reason for that was that sometimes the greatest businesses do come about as a function of either other businesses or some activity in the market that they can capitalize on. And so in this particular case, the business spun out of a forum, the forum was called SitePoint. And that forum is still around today. And SitePoint was a developer’s community. And so developers are obviously, early adopters. They’re often builders, of businesses and or tech in their own right, of course. And so what they noticed was that within this forum that they had started, people were trading. And so the forum was there to share learning materials and then ended up becoming a marketplace. Because what happened was those developers would say, Hey, I built an app. It’s making money, does anyone want to buy it? And inevitably, what happened was people did. And so as a function of that, we had this fast growth community that we’re doing exactly what is happening on Flippa. Today, they then spun that out alongside another marketplace. So they are prolific marketplace entrepreneurs, The first one they spun out was called 99 Designs. I know that the business that recently exited to Vistaprint. And the second business they spun out was the one we’re talking about, obviously, here today, which is Flippa. And so the problem that it solves is two things. One is acquisition, entrepreneurship has been around for a long time people buy companies, but they do so in very fragmented environments. They do so without the assistance of sales data and benchmarking, without the assistance of insights without the assistance of an end-to-end, negotiation flow and or funnel. And then the second problem is that most business owners don’t actually know how to exit, where to exit or even that there are supreme amounts of buyer liquidity available in the marketplace. And so, therefore, the marketplace solves the problem of exiting and owning. And it Now does that globally. And it does that to the tune of 12 to 15,000 transactions a year.

Joanna:  

Absolutely amazing. So, and I think it’s, it’s sort of an interesting story that it started in Australia. But you know, there’s a large US presence now on the website of the businesses in the marketplace. You know, what, what’s the reason for that?

Blake:  

Yep, I look, I think there are a few things one, the Australian entrepreneurial community is obviously fast-growing, and it’s really dynamic. And it’s super successful, but it’s just not as big as what you get throughout the US and what you get throughout Europe in the UK. And so as a function of that, we had a marketplace, that was satisfying entrepreneurs, owners and investors, and the risk appetite of an American investor, owner, operator and entrepreneur is for the most part, bigger in number and bigger in quantum and quantum being the value that they can afford them a deal. It’s just bigger. And so if you’re going to be a growth, thriving business and marketplace, you go where there are, where the opportunity is, and often that opportunity just comes to you organically as well. And so, Flippa is an Australian-born company. Our HQ is still Melbourne, Australia, but by nature of the very large ecosystem of entrepreneurs and investors being predominantly North American, that just happens to be where the vast majority of our revenue is derived.

Joanna:  

Fascinating. Okay, now I want to turn to talk about, you know, innovation more broadly, but first, maybe looking at AI and then we’ll move further than that, but we’ve got to quote from you last week talking about the AI algorithms successfully matching an impressive 106,000 buyers to 1000 Different businesses for sale, resulting in the last month the AI recommender being responsible for five and a half million in deal value. That’s probably five and a half million US, I’m guessing in deal value. And so, tell me how so how is how is Flippa using AI? Has this evolved? You know, obviously, the world seems to have evolved in terms of AI just exponentially in the last, you know, few months, when AI is now taking centre stage of every business owner’s discussion point. How long has Flippa been using AI? How have you implemented it? What does the future look like?

Blake:  

We began using it. We deployed the product in February, we’ve been building the product for about three to six months. The idea came on about six months ago. And then from there, it was a matter of building now. We have always been in the business of matching. The problem with matching as a concept, though, is it is required historically, for the buyer to demonstrate huge amounts of intent. So they had to tell you exactly what they were looking for in order to match them to relevant businesses. With AI, you can infer intent. And so what we have done is essentially build a recommender tool that matches buyers, based on their latent and hidden intent with relevant businesses, which we believe and the AI believes very accurately will be they will be a good match for. And so this comes to say this is about providing value for buyers, making sure that business owners get every possible chance of selling their business. But to be honest, it’s also solving a real business problem, we were signing up between 25 to 27,000, new buyers each month, they come in with a full head of steam, and they hunt and search on the marketplace very aggressively. But Flip is not the type of company that you wake up each morning. You don’t like it each morning and say I better check out Flippa to see if they’ve got a new deal. It’s just it’s not that right, you don’t go to Flipper every day or week like you do go to Kohl’s to go buy a loaf of bread and groceries. As a function of that, you need a reason for people to come back to your platform and use it so it solves a real business problem. So essentially, what you have is AI helping out with M&A such that there’s greater efficiency in that initial matching process. Thus traditional methods have relied on professional networks and CRMs. My professional networks are those people I know who want to buy a business that looks like this. And so I pick up my iPhone and I go to the contact section, I start dialling those people, or I go to a CRM, and I send an email blast out and say, Hey, I’ve got this great business, would you like to buy it? That’s not really sufficient. Because that assumes that someone in your database is going to actually be interested in that particular business. And that is just like, that’s a shotgun approach. Business owners actually want greater efficiency in the process. And they want to know that when a buyer comes and acquires those buyers have a very high likelihood of being interested in that business. And so AI becomes increasingly important in revolutionising the way that M&A is done in the way that deals happen efficiently. So that’s, that’s what we built. So what actually is it? Well, it’s cutting-edge machine learning. So it’s a system that matches buyers based on latent and hidden intent with quality listings, that the AI specifically picks for them. And so what it does is it matches, actions and behaviours, and informs preferences. So it used to be that a buyer would have to create a mandate. Now that mandate is created dynamically for them based on actions and behaviors of them and other people who look like them.

Joanna:  

So what does that mean practically? So break it down, like what’s an example of a buyer that comes in how is that buyer, you know, I guess, processed in the Flippa system in order to identify these matches?

Blake:  

So hundreds upon hundreds of different variables, but let’s just choose a couple of really, really obvious ones. So the first one is that you could have a buyer like me, go and view an asset. Then you have another buyer, like me that views the same asset. The original buyer goes on to view 10 other assets. The next buyer doesn’t, but their behaviours on the original asset are similar, they inquire the size of the offer they put in is similar their engagement with that particular listing and the sections of that listing are also similar. 

Blake:  

So it infers that they will also be interested in the listings that that first buyer went on to look at, even if they haven’t demonstrated that intent, as an example. Now, it does that across hundreds of different variables. So what you end up with is this crazy-looking alien-like structure, which is called a graph neural network. And whether it’s density, that’s showing all of the buyers who have got very similar traits to each other. And then when there’s distance between those buyers, it shows that they’re entirely different in what they would want to look at. And by that also happens on the sell side. So it looks at similarities between listings. And once you do this, across hundreds upon hundreds of 1000s of buyers, Flippa has over 1.5 million registered buyers.

Blake:  

 And in any given time, we have six and a half 1000 assets for sale. So we, therefore, can run these graph neural networks on an always-on basis, and essentially get smarter. As Joanna’s activity increases on the platform, it starts to make, it starts to infer what Joanna would also be interested in, based on your likelihood of engaging based on other like buyers and their engagement. And so it just keeps on running that. And so each week, you’ve got approximately 100,000 matches happening in real-time. And what’s what’s interesting is, we knew we could build it, it would take some time to get the machine learning frameworks working, because it learns it actively learns. But what we needed to see was the data. And what we’ve proven is that people love it. 

Blake:  

So there are two things going on. That’s text messaging and email. And so the open rates, we sent hundreds of 1000s of emails a week, but the open rates on these emails are consistently extraordinarily high 70%, which is world-class, the click-through rate is above and beyond 15%, most people would say at two and a half per cent click-through rate was good. And then the engagement with the listings has skyrocketed. So we’ve seen more negotiations on our listings than ever before. And as a function of that, our business owners who we represent on the platform as a marketplace are excited because they’re seeing more activity on their listings. And so everyone’s kind of happy, the buyers get to see relevant stuff, the sellers get to sell to a relevant buyer. And then Flippa is happy too because we have a successful.

Joanna:  

Love it’s just amazing. And tell me how long has the AI component been live?

Blake:  

went live on February 5, and we’re on version four and a bit. Because it’s like 4.1 4.2 4.3. And so that says we add more variables. And sometimes you remove the variables because it gets inaccurate. And so we’ve been adding more and more variables into the mix. And the variables include keyword data, right? So I list myself as a private equity buyer, who is looking for E-commerce assets that are predominantly direct to consumers, and I like the category of health and wellness. But then weirdly, I go and buy an E-commerce business that has shoppers, Shopify predominantly direct to consumer, but he’s in the pets category. Okay, so you told us you wanted health and wellness. But you just proved behaviours differently the way you identified yourself. So the AI will take that into consideration. And it will then say we don’t care what you said about yourself. And this Joanna, you’d love to before we were doing our briefing call, this is analogous to dating, right? Because actually, I might say, I am interested in people who look like this. Then you put me on a date with somebody else who looks slightly different with a fantastic personality and all of a sudden I’m in love. Thus the reality is, that AI is far better at matching than the buyers themselves. Is that understanding their own preferences?

Joanna:  

Fascinating what you said something before about you know, there are some variables that that are working and some that are inaccurate if you got any examples of those inaccurate ferry boats,

Blake:  

I know I would almost need to go to the data science team to sort of understand that a little bit more for you. But sometimes you do have browsing patterns, which can misinform. And so it could be that we infer that a buyer is interested in a particular business model, when that what they’re trying to understand is the category. So they might be looking at SAS businesses or content businesses or apps that are in the pets category. In fact, they are definitely only interested in E-commerce businesses, but their behaviours infer otherwise. And so what you’ve got to do is make sure that you then throw in other variables to determine whether those behaviours are in fact, a good inference. And so then you would throw in messages. So you would say, Okay, well, they might be browsing certain categories of assets. But what types of assets do they actually inquire about? And suddenly, the AI would then say, Okay, well, the variable of browsing isn’t actually improving the AI. It’s another match. It’s the variable of messaging, which is a better indicator, and so the AI will naturally reconfigure itself and then kick that variable out.

Joanna:  

Love it. And you just mentioned data scientists, which matches the question I was about to ask, you must be collecting a heck of a lot of data there. In Flippa, what? Are there any trends in the market that you’re seeing through the data at the moment? Or if you don’t have them on hand? Don’t worry, we can talk about it in another session. We love. We love market trends, though here in The Deal Room.

Blake:  

Yeah, we certainly have a lot of data. And what I would say is other than the super sophisticated AI, the trends around buying and selling businesses are kind of quite predictable that you and I could get on a whiteboard, and just, hey, what’s the state of the economy? What types of categories would be interesting to people, you can kind of guess it, and the marketplace will tell you similarly. And so let’s face it, we’re in a high inflation environment, and you’re in a high-interest rate environment. And so as a function of that, people are looking for more predictable industries that have a history of being recession proof, and that have a history of being good performers, regardless of the state of the economy. And that’s what our buyers will go and seek. So they will look for finance-related websites, and they will look for categories like health and wellness because people typically still look after themselves in times of need. But maybe luxury goods-orientated assets are less preferable right now. So yes, we have a very, very an exercise, if you go to the flippa.com homepage and just scroll down, you’ll see the long tail of searches that buyers undertake on the platform. And so it will show you that SAS is number one. And that content is number two, but then it will break it down into a long tail around categories like health and wellness, finance, travel, pets, and even crypto right so that crypto and NF T’s literally dropped off a cliff about six, maybe about 12 months ago now. So yeah, we have that data. We watch that data really closely. It’s super fascinating and happy to do a session with you where we just screen share. And so you show you some of that.

Joanna:  

Oh, I’d love it. Well, look, I love hearing about data in this area. And what what we as we said, we might come back and do this on an on a separate episode. But just very quickly, one of the what are the key data points that you guys are looking at right now? What are the things in the data that’s that you know, you’re most interested in looking at?

Blake:  

So business multiples is the number one right? Because what we have to do is inform the business owner as to what their business will be worth and we have to get that reasonably right. So we will inform business owners on a real time basis based on our sales data, what the likely price is that they could expect similarly, we will tell buyers that as well. And so when you’re looking at an asset on flipper, let’s for argument’s sake talker, talk about a $250,000 SAS business. There is a benchmarking module that sits on the listing page and that’s dynamic. And what that shows you is how that asset is priced relative to other assets. like it that are either for sale or have sold. So that gives buyers instant visibility into whether this is overpriced or underpriced. And that’s number one, because buyers want to know whether they’re in the game and sellers want to know whether they’re accurately priced. So that’s number one. The second thing we will tend to look at is activity on the marketplace to predict whether something is going to sell or not. And so when an asset has a lot of interest, then we can predictably say that this business will sell and then we have forecasting around that for ourselves. And so there’s data for the purposes of externalizing, the use case and giving people value, and then as data we use for our own benefit, and so, you know, that’s a balancing act.

Joanna:  

Love it. And I guess just back to this general AI topic, what, you know, you’ve talked about where Flippa is right now, what are the biggest, you know, what are the short-term developments that you’re looking at for the use of AI or the, you know, the progress progression of the use of AI in Flippa.

Blake:  

So to be really selfish about it, we are measured by how many listings we win, and how many of those we sell. And the great thing about that is our buyers want us to win listings. So it’s very much inextricably connected to the value equation that they seek, and then our sellers want to sell. So that’s, that’s the conversion. So we’re, we’re very aligned, there’s great synergy between what we try to achieve and what our customers want us to achieve. So that’s, that’s critical. So then we’ll How do you apply AI to that? Well, you look at the problems that you have in each of those two needs. So the first one is, we want to win assets to our platform. And so much like Airbnb, when a host or a homeowner goes to put their house or apartment on Airbnb, there’s an onboarding process. And that the onboarding process is complex. So how can AI help to streamline that because selling a business isn’t easy, and most people don’t know what matters? And so you can actually do a lot of that work using AI. So the first thing is, if I say to you, explain your business to me as if someone wanted to buy it. People don’t know how to do that. But AI is really good at that. So we can take on all the data. And then we can basically say to the business owner, you can write your own listing description if you want, you can try to sell this to the universe of buyers on Flippa. But we think you’re better off utilizing AI to do that. So we do that. They can then hit a button that says generate listing description. And it basically writes this long form brochure that says, aren’t we great. And so that’s, that’s really, really cool. Then the other thing you can then do in the interest of conversion, is you can basically write a due diligence report that analyzes all the data from our 15 different data sources, Google Analytics, QuickBooks Online, Xero, Stripe, Shopify, whatever it is. And instead of just throwing that data at someone, AI will rationalize that data and spit out trends, seasonality, skew, which skew is good, which SKU is less good by percentages, all those types of things. So it’s dynamic in the sense that it enables you to take complex problems and solve them very quickly. And this use case is really relevant.

Joanna:  

Love it. Absolutely love it Blake. Well look thank you so much for coming on The Deal Room Podcast very sadly we’ve run out of time to talk today but in good news, we’ve got you back for another few episodes into the future just want to say a huge thank you for coming on to the show. I really enjoyed it. How do our buyers and sellers who are listening in I mean, I guess it’s pretty obvious how they find Flippa. You know, where did they go? Are there any you know, great things that they should be checking out along the way in terms of stuff that they can get on the platform, even if they’re not quite ready to buy or sell just yet.

Blake:  

Yeah, probably the best thing is just gotta Flippa.com and subscribe to the daily it’s an email and all it is is a list of deals that comes out each day. And from there, you can learn a lot you can learn valuations, you can kind of learn what sells and what doesn’t sell. And it gives you a sort of Intel. The thing is you can learn what what works well in certain businesses and what doesn’t work so well in other businesses. So it’s got some tips and techniques as a part of that. So I would do that if I was new. And that will that will take you on a bit of a journey. 

Joanna:  

I love it. Blake, huge thank you for coming on The Deal Room Podcast today.

Blake:  

Thank you so much Joanna.

Joanna:

Well, that’s it for this episode of The Deal Room Podcast. We hope you’re now primed for your next deal with these pointers and have enjoyed these fascinating insights.

Now if you’d like more information about this topic, then head over to our website at thedealroompodcast.com where you’ll be able to download a transcript of this episode as well as access any contact details and any other additional information we referred to in today’s podcast. Now if you’d like to get in contact without guests today and the services they offer, you can go ahead and check out our show notes for a link right through to them and their details. You can also book indirectly with our Legal Eagles at Aspect Legal.

If you’d like to soundboard your next steps, discuss a legal question or find out more about how we can assist, whether that’s with buying or selling a business or perhaps somewhere in between. Now, don’t forget to subscribe to The Deal Room Podcast on your favourite podcast player to get notifications whenever a new episode is out.

We’d also love to hear your feedback. So please leave us a review and rating. If you’re already one of our subscribers. Or even if you’re listening to this podcast for the very first time, every review helps our team produce valuable content for you. Well, thanks again for listening in. You’ve been listening to Joanna Oakey and The Deal Room Podcast, a podcast proudly brought to you by our commercial legal practice Aspect Legal. See you next time.

Our Business Sales and Growth by Acquisitions Services

Aspect Legal has a number of great services that help businesses prepare for a sale or acquisition to help them prepare in advance and to get transaction ready. And we’ve also got a range of services to help guide businesses through the sale and acquisition process.

We work with clients both big and small and have different types of services depending on size and complexity. We provide a free consultation to discuss your proposed sale or acquisition – so see our show notes on how to book a time to speak with us, or head over to our website at Aspectlegal.com.au