Ledge: Olin, it’s great to have you on. Thank you for joining us.
Olin: Thanks so much, Ledge. It’s a pleasure to be here and I’m honored and flattered that you invited me. Thanks!
Ledge: Absolutely. So, for those who are not familiar with you and your work, would you mind giving just a two or three minute introduction? Where you came from, where you’re going, and what you’re up to these days.
Olin: Sure. I’m a serial entrepreneur. This is my eighth startup. LeadCrunch is a B2B lookalike marketing platform. What we do is, we enable marketers to find the right audience and engage that audience with the right content to start prospects on the buying journey.
You can think of LeadCrunch as doing for businesses that sell to businesses what Facebook and Google do for marketers that sell to consumers.
Ledge: We have a heavily technological audience, so connect the dots to that world in the technology seat and the engineering seat to your world, to talking to marketing and sales folks.
Olin: Probably what would be interesting is, the genesis of our technology is very oriented in physics and natural language processing and work that we originally did for the military.
My training and background is in systems theory and artificial intelligence, and LeadCrunch originally started in a very different space. We wanted to find a way to accelerate medical research, and the insight that we had was is that language has patterns and those patterns are the way we derive meaning, and why aren’t we doing more intelligent pattern recognition among medical research to accelerate the way we find new drugs and so forth?
So we developed a very simple – what I think now is very simple – language model, and we data-mined about 20 million open source medical research papers, and we helped a friend of mine accelerate the discovery of a possible link between Type I diabetes and the herpes virus. That research was published. It got $11 million in funding, and we thought, great, let’s go use this technology to cure cancer. Let’s do something big!
What we found was is that the medical research community is really more interested in publishing papers and getting funding, and that they don’t have budget for our kind of technology.
So the genesis of our technology was really that we had what we thought was a great idea, and we had some technical success, but we had a complete and abject market failure because we’d never identified a market that would be willing to pay for a natural language processing engine to go out and find new research methods.
So we pivoted – and we’re about to shut the company down, we’re almost out of cash – and we were fortunate to win a contract with Lockheed Martin, the big defense contractor, where we beat IBM Watson and Palantir to win an R&D contract to develop target verification technologies for the Navy.
We did that for a while. It kept the lights on but, quite frankly, we were just not cut out to be military contractors.
So we pivoted and we looked for ways that we could use targeting technology in a commercial setting.
About that time, Facebook and Google were just crushing it with their lookalike platforms. Facebook Lookalike Audience. Google Audience Match. We thought, wow, we really don’t want to compete with Google or Facebook. Maybe we can do this for businesses that sell to businesses. That became the genesis of where we started.
Hopefully, that wasn’t too long of a story, but it shows that the way that we developed was through irrational exuberance and enthusiasm to overcome failures.
Ledge: I think we often get this idea, particularly on the engineering side or product side, of “Let’s pivot the tech,” or, “Maybe there’s something wrong with the tech,” Very often that product market fit, the thing is, the real challenge is you have an awesome solution you haven’t found the problem for yet. You know, backwards.
Olin: My background is engineering and I tend to get really enamored with developing the hammer, and I run around looking for nails. What’s really a better business approach is often to start with the problem, start with the nail, then you build the hammer specific to that nail.
I wish I was that smart. In the case of this company, we really started with the hammer looking for, where is this technology going to fit? What problems can it solve? There’s more wrong answers than right answers, and the proof of that is that we had about four major failures before we got to product market fit.
Ledge: Well, congrats on getting product market fit before you ran out of runway. I think that that is the big, most important lean startup objective – anybody that reads those books.
Let me ask you about audience matching. For those who are not familiar, what is that? And then talk about that in the context of B2B and how it’s going to be different than maybe the consumer mapping that you might do, or matching, with a Google or a Facebook.
Olin: B2B marketing is dramatically different than B2C marketing. In business to business marketing, you typically have five, six, seven, eight people involved in the purchase decision. So if Mary and Joe have very different ideas about what they’re going to purchase, that creates a lot of friction in the buying process.
If you compare that to the consumer decision, that’s made by individuals. The advantage that Facebook and Google have are data advantages, where they know what websites you’ve been to, they know what you like. They can create lots of data points to understand who you are as a person, and who you’re very similar to.
They have enough data on who you’re similar to, and they know that, for instance, if Ledge and Olin are very similar people, Ledge likes purple sweaters, let’s put an ad in front of Olin for a purple sweater and we’ll see that he has a much higher conversion rate. They do this through deep personalization.
If we did that deep personalization in the business world, what happens is the two people making the decision, you know, Mary and Joe, one might like red sweaters and the other may like purple sweaters and personalizing it actually amplifies their differences. Actually slows the process down. Actually reduces conversion rate.
So, what we do is we find the similarities between Mary and Joe so that we can build consensus. The idea is not to personalize to seven or eight people in the company, but rather to create a coherent message that brings together consensus around our customer’s brand.
Ledge: So, you’re looking for the ways to… Having been in B2B sales a long time I’m thinking, anybody who’s not familiar with that, you have this constellation decision making. You have people who are influencers, and you have people who can say yes, and people who can say no, people who just get in the way because that’s what their job is and they’re gatekeepers. People who can write a check. People who can make a decision.
I’m hearing you say that, what you try to do is take that crowd and find ways to grease the rails to get through all those buying or supporting personas?
Olin: Right. It’s actually a three-step process. First you need to find the right accounts to go after, and that’s fairly straight-forward lookalike marketing. We take a list of 50 of your best customers or your account based marketing list. Whatever list you want to give us, we’re going to find lookalikes to that list. And within that lookalike list with that audience that we build, there’s going to be thousands and thousands of people that we organize into what we call buying centers.
Buying center is an old idea, and that’s the different people that are involved and the decision makers. The decision maker, the gatekeeper, all the people that influence that decision. In sales, they oftentimes are called account mapping.
So, who’s involved in the decision and how do we reach the right people? Our AI models about 180 million people in America, and we can predict who is likely to be involved in a purchase decision and what they’re likely to be interested in about our customer’s brand. We can consult with our customer and say, “Look, we really think you need a white paper on this subject or a piece of content that would help, say a buyer’s guide, to move forward.” Then we syndicate that content into that buying group to build consensus.
Ledge: Talk about that syndication process. What does that look like? How do you get access to those folks?
Olin: Email. Display ads. Telemarketing. You name it, we’ll use it. What we’ve really found that’s interesting lately is display ads, because we can precisely target. We can suddenly change the use of display ads from awareness – where marketers typically use display ads to get brand awareness. We can now use those click-throughs to get engagement on a piece of content.
That’s s relatively new thing that we’re doing and we’re seeing tremendous CTRs. In one of our cases it was three times greater than the leading intent-based marketing solution.
So, we use a number of different methods to get engagement. Of course, we have what we call a response model for each one of these people. Everyone in our database, we have a model of who’s likely to respond in what medium.
Ledge: That’s fantastic. So, you’re talking to a lot of technology product, CTO-type folks and I think you probably know that in the technology seat sometimes marketing is this nebulous, weird group of people that tries to use technology that comes from 7,500 different vendors the last time you checked the list of all the logos. The purchasing decision alone for martech is complicated. The integration is complicated from a technology and data perspective.
Can you just talk about that? If you zoom out to the CTO seat, how should they and will they experience martech and solutions like yours in general?
Olin: Wow. That’s a vast subject. There’s about 7,000 martech platforms. We think there are so many that we don’t need to have another login. You certainly don’t need to go out and buy more technology. There’s plenty of technology out there.
There’s an old joke in the AI community that when artificial intelligence works we call it software, when it doesn’t work we talk a lot about AI.
In our case, we try to simplify the user experience to simply upload your best customers then we directly import engaged leads into the CRM.
I think that, from a CTO perspective, marketing has been making their own technology decisions now for a while. I think most people, including marketers, would agree that there’s far too much technology required, too little return on investment.
So, I think that the way to look at us is more like a service. If the technology works, great. You don’t even feel it. You don’t even see it. All you get is better results.
Ledge: Absolutely. I agree with that answer. You may be familiar with the DevOps world, and it’s facing exactly the same thing now with the tooling and pipelines and releases. There’s thousands tools trying to solve the operational challenges, and it reminds me a lot of that picture of the martech landscape where the logos now outnumber the problems. It’s harder to shop for the solutions and find the right tool for the job than it ever has been.
Look from your seat at CEO, and you’re trying to integrate and build all kinds of things to pull your company together and make your teams across functions more successful. What does it look like from the top down when you see your leadership team, and scaling up of technology in other organizations, and making them work together? How has that experience been?
Olin: You know, I have a lot of trust and confidence in my leadership team so they have a great deal of autonomy in what they select and how they go about buying and building technology.
There’s always the buy-build-partner-license question. You usually have four different choices for solving a given problem. As a company, we really encourage people not to build – let’s focus our build on our own solutions for customers – and let’s use the right set of tools for everything else.
What’s interesting is, we have situations where we’ve outgrown technology and we’ve had to, for instance, get locked into a subscription to something that we’re not using. That has informed the way we do pricing for our product.
So, for LeadCrunch’s customers, they don’t buy the subscription to what we do, we charge by the unit of value that we deliver – and that’s a cost per lead basis.
That’s one of the reasons why we’ve grown so quickly. We grew our revenue 20% per month for 30 months in a row. Oh gosh! I think we grew over 340% last year. From our view is that that unit based pricing is a really smart way to look at technology. In fact, we’re looking at an integration tool right now that’s just going to charge us by record. I like that model a lot because that enables us to do attribution of value, allows us to look at the return on investment in technology.
I would like to urge the audience to really think through how they measure the ROI of something.
Ledge: Is that going to be similar to what you’d think of as a utility based model?
Ledge: Immediately, everybody’s head goes to like an Amazon, and use tiny portions of stuff and pay for it as you go. Do you envision that all SaaS goes that direction?
Olin: I think SaaS is a really old pricing model that was originally developed to move things from capital expenditures to operating expenditures, and it had this side effect of predictable revenue. Right? Why SaaS became popular with investors is the idea of sticky revenue and predictability, and that your growth was accretive. That you’re only selling for new growth because if your subscription then the subscriptions don’t go away.
I think that that model no longer works for many of the places where it’s being applied. For plumbing types of applications like a CRM, SaaS works great. But you don’t buy your gasoline on a subscription, for very good reasons. The oil companies want to maximize their profit based on the spot value of crude oil, and what does it cost to refine that oil into gasoline.
Similarly, most companies, for instance ours, is more like a gasoline company in that you can’t run your business without leads just like you can’t run your car without gasoline. Our company is going to charge more for more valuable leads and less for less valuable leads, and you can’t do that with a subscription. Even worse for the customer, the customer can’t do attribution very easily on a subscription for what is the return on investment.
So I think that Software as a Service subscription model businesses became wildly popular largely driven by investor interest. But at the end of the day, I care a lot less about what investors think than I do what customers value. And we certainly drive our business based on customer value, not the proclivities of what the venture capital community thinks because they’ve got bosses too called limited partners, and they’re usually not customers.
Ledge: Absolutely. You use the analogy of a spot crew, which also adheres to commodity market behaviors with like sort of the global market – that more supply drives down the price. So if everybody starts doing this across every utility, do we start to run into a spot where you’ve commoditized the utility of compute, or mean value, or whatever your unit of measure is. Do you think about that at all?
Olin: I do. I think that commoditization in leads has happened for a long time. There’s this prevailing wisdom that more is better, which is actually completely wrong. The quality of a lead is determined by its conversion rate.
It’s one of the most powerful places to get value in a company, is to have a top of a funnel with very high conversion rates. Everything is run smoother. You have incredible leverage down at the bottom of the funnel in terms of your return on investment.
So, I believe your conflating two ideas. Quality is a differentiator, not the pricing model. The pricing model does not lead you immediately to commoditization. The pricing model really builds in the value of quality, unless everything is the same quality then of course it becomes a commodity. But you still have plenty of room for differentiation with an on-demand pricing model.
Ledge: Absolutely. What would you say then are the qualities of a high-conversion lead that you might come up with, that you could value it higher? What are the heuristics of that that you can know ahead of time? It’s tempting for me to say, “Okay, cool. So I get a lead, I might or might not convert it,” but I would only know after the fact.
What’s the predictive capacity and heuristics to know if something’s more valuable on the frontend?
Olin: Well, leads are really people, and people go through a journey to become a customer and those journeys have steps along the way. How much do they engage with the content? How much do they engage with your sales team? What is their fit to your ideal customer profile? There’s probably 20 or 30 different measures that can be used by a different companies. Different companies have different ways of measuring how well that funnel is working but, generally speaking, the higher the level of engagement, the better it’s going to convert.
Ledge: Very interesting. Have you evolved from what would have been the traditional… You’re describing components of what I would have read about five years ago maybe as, you know, inbound marketing. It was the big revolution. Every blog post was about that.
Do you think that it’s evolved from there now, or is this a smarter version of that original concept that we didn’t have the ability to support before?
Olin: It’s an ‘and’ not an ‘or’. Inbound is still incredibly valuable, and I would argue that inbound is always going to be your best quality of leads. The problem with inbound is you have no control over who is going to be the inbound. You can influence it.
Recently I got a letter from someone in prison who was very interested in my technology. So you can think of that as the inbound lead that is completely not qualified. He’s serving a life sentence for murder, so how is that possibly a qualified lead?
Whereas, outbound you have a much greater control over volume. You have a much greater control over audience selection. The idea of our technology is the lookalike audience. The lookalike lead. That gives you a high degree of control and precision over who you’re going after.
We couldn’t be where we are today without some of the great work that was done in inbound marketing, and I would call our technology a complement to inbound marketing.
Ledge: All right. I always finish up here with some lightening round questions. This is critically important stuff. I’m going to get this open here in my notes list.
Olin: What’s my favorite movie?
Ledge: We get way serious fast, past that. Let’s see. Here we go. This is really critical stuff. You’re ready?
Ledge: All right. Star Wars or Star Trek?
Olin: Argh! Equal. If I have a gun to my head, Star Trek.
Ledge: No pun intended, gun.io. No guns.
What are you reading right now?
Olin: The Coddling of the American Mind: How Good Intentions and Bad Ideas Are Setting Up a Generation for Failure.
Ledge: Wow! Intense stuff. I like that. We’re going to have to link to that one.
What can you not live without?
Olin: My road bike.
Ledge: Excellent. What is the last thing that you Googled for work?
Olin: My goodness! I Google so much I can’t remember what… The last one was about a customer. I was looking at how much funding one of our customers received. Didn’t really Google, I went to Crunchbase.
Ledge: Excellent. That’s also how I use it a lot too, particular of our guests. My favorite answer to that question was if someone said they Googled me to learn about me before the interview. I personally thought that was awesome.
I don’t know if you’re a fan of The Office, but there’s a classic episode where Jim, the office protagonist, is messing with Dwight, the office heel. And he’s sending him faxes from future Dwight. For example, “The coffee is poisoned today,” or what have you. I like to ask people, if I gave you one piece of paper and one of those big, thick, black sharpies and I said, “You are now future Olin and you are faxing back to yourself some advice that you can put on that paper. What would you write on that?”
Olin: “Don’t sell the Apple stock.”
Ledge: Well played. Well played. Awesome. Well, Olin, thank you for being a good sport. We always appreciate that last round there. Super-interesting stuff. Thanks so much for the insights, and it’s great to have you on.
Olin: Thank you so much. Really appreciate it.