Link To Guest Website:

Title: “Utilizing Machine Learning To Hyper Localize Advertising”
Guest: Sam Mallikarjunan – OneScreen.AI
Interviewers: Jeffrey Davis – MAGE LLC & Evan Macedo – Sapers & Wallack / FEI Boston

Click here to read the transcript

Jeffrey (1s):
Welcome Sam

Sam (3s):
Thanks for having me – I’m looking forward to it. How are things up in Boston?

Jeffrey (7s):
Very nice. What stage are you at?

Sam (11s):
We’re a startup. We launched last March, right at the beginning of the pandemic because of the pandemic. Raised our seed. Just about to close our seed round and just about to hit our first million dollar sales benchmark. So rounding error, probably for most people on this call, but a good milestone for a startup.

Evan (29s):
Great! And Sam, you know, starting up in the middle of the, at the beginning of the pandemic is something that a lot of people probably wouldn’t even think of doing. So why don’t you tell us a little bit about what made you decide to go take the, that foot forward or that step, that first step forward to go and do that?

Sam (49s):
I mean, it was, it was an accident, honestly, when the quarantines were beginning myself and some other HubSpot alumni, which is a Boston startup got together for a hackathon to try and help small businesses. And we’re in a cigar lounge in Watertown, just commercials running on the TV, but one, they saw two, the company’s not making any money off of it. So we had the idea what, if you could build like Google display network or Facebook audience network for screens in the real world and help small businesses just make more money. Right? There’s commercials running in there anyways. It’s just unlike online. They don’t get paid for it. And beyond those, like, you know, people think about billboards and things like that. Most billboards aren’t owned by clear channel or, or out front they’re owned by like smaller local companies, smaller businesses.

Sam (1m 33s):
So we, we built the prototype. Then we did what we kind of have jokingly called a reverse stealth mode. We called every company in the space, told them what we were planning to do to try and understand why they weren’t doing it because it’s so, it’s so manual, it’s archaic, but it’s $31 billion global industry. So yeah, we quit our jobs and said, what if we can build something that would make, you know, small businesses, a powerful marketing, medium, like how would that change the world?

Jeffrey (2m 2s):
Why was there some, like a seminal moment Sam, that sort of told you that this was the time to start this business? Or you were doing something that sort of triggered you with all of this?

Sam (2m 13s):
I mean, there, there really was the, the beginning of how do we help small businesses make more money, especially off of reduced occupancy. Like they used to be a bar in, in Boston called courtside rest in peace. But like, if you wanted to reach HubSpot employees, like you should’ve just run ads in there, right? Like don’t do all the other stuff that people try to do when they’re selling to HubSpot. But that’s never, that’s never existed before the ability to connect marketers like me with audiences in any screen, other than this one, the desktop or laptop, or this one, the mobile device has never existed before. So that was why we started it. And then the reason we decided to like do it full-time and lean into it.

Sam (2m 56s):
It was, I mean, you dream about like this I’ve spent so much time on a whiteboard trying to find an industry that’s at least $10 billion a year that just has an information issue, right? There’s no directory of every billboard in the United States and who owns them. If you want to buy something, you have to like call 40 different companies and they take six weeks. So there’s th there’s very interesting machine learning and AI applications, but it’s also just, it’s like being back in 1995 or the year 2000 when it’s like, Hey, what if we had a directory of things that people could buy and called it Amazon?

Jeffrey (3m 33s):
I’m good at reading minds. I know Evan has a question.

Evan (3m 37s):
I mean, you just mentioned machine learning and AI, and I know a lot of our listeners on this call like myself are, you know, running the finances of our companies, but you come from a tech background totally different than a lot of our listeners. I just wanted to hear from you, you know, what are you using for? Or are you using any type of artificial intelligence or machine learning in your current business? And I mean, what is AI and machine learning mean to you in your business?

Sam (4m 7s):
Yeah. So machine learning has been one of the most important developments for marketing and just company growth over the last 10 years, Facebook and Google, especially, you know, we were alert. We were all worried about truck drivers losing their jobs to AI, but Facebook ads managers are probably the ones who lose at first because now you just tell Facebook the outcome you want to have happen. Somebody buys something, go somewhere or call something. And then Facebook takes a bunch of data and learns from every time somebody does or doesn’t, you know, reach that objective for advertising the real world. It’s actually significantly more complicated. The Internet’s actually pretty flat and straightforward, but for example, Cappies liquor in Boston.

Sam (4m 48s):
If you’re familiar with them, one of our advertisers, there are 3% of people who go to Cathy’s liquor and then go to work. So where their cell phone spends eight hours a day of super, to, to find out who they are. But in general, the system’s learning. If it thinks you’re on your way to work, it shouldn’t show you an ad to get you to go to copy’s liquor store. So it’s trying to say, based on the what’s going on in your life and the, sort of the context of that billboard or that TV, or that, you know, Uber the screen in the back of the Uber, based on what the marketer wants to accomplish, who values that the most and who should we sell that to at what price? So it’s, it was, it was a big thing for the internet, but it’s going to be much more important to get right for, you know, pricing something that’s, that’s a lot more complicated.

Jeffrey (5m 32s):
That’s, that’s really something. And I think the AI intelligence is important. So I have a different question for you, Sam, if you don’t mind, you know, so many entrepreneurs that I’ve been associated with in my career, they jump in and they don’t understand the alignment between strategy and findings for achieving their longterm goals. And I’m wondering how you’ve addressed that issue. It’s something that I have heard that you’ve been able to look at, and I think it is important for entrepreneurs to understand this is an important area to have alignment.

Sam (6m 2s):
It’s literally the most important issue. What kills most startups is the is bad cashflow management, bad economics. If you’re familiar with HubSpot, that was a company I spent eight years at leading HubSpot labs. You kind of think of us as this like fluffy happy huggy, make marketing people love kind of company. No, it’s, it’s a cashflow machine. Every company meeting, we discussed the unit economics and the business economics, all of us on the team understood that. So I knew, you know, if I had a customer lifetime value to acquisition ratio of seven, that I should bring that down to three so that we could grow faster.

Sam (6m 42s):
Like it wasn’t just the finance team. The CFO was who was really setting the guardrails of strategy for what we were executing as a go to market team as an engineering team and telling us, you know, if we’re going to bring in money from investors, what is the best way to deploy it, to grow the enterprise value? And so as a marketer, being able to draw a line and make decisions, you know, I don’t necessarily have to understand all of the nuances behind the accounting and things like that. All I need to know is what is our goal? Is it to grow faster? Is it to have more profitable unit economics, et cetera.

Sam (7m 23s):
And then how can I analyze that? So I can make really, really good decisions. And we do that at the startup phase as well. Although my co-founder made me take cashflow is more important than your mother out of our investor pitch deck. You know, we have a very, we’re raising a specific amount of money to validate a specific unit economics thesis so that if I go and raise 10 or $20 million for our series a, I have an intelligent answer to what are you going to do with that? Well,

Jeffrey (7m 48s):
I’m okay with the cash thing. I’m sure our last guest today, Jay Toole, he’s going to agree when he comes on. I was always taught if cash is oxygen and without it, you’re choked to death. Evan, I got your signal, you know, through the osmosis. I know you have another question coming through our time.

Evan (8m 5s):
Absolutely. Sam, I know that the relationship between the CEO and the CFO is one that is vital to any organization. I just wanted to hear about what your relationship is with your CFO. You know, are you calling the shots and is your CFO trying to kind of help you out there, or is the CFO calling the shots and everyone’s kind of taken their advice to move forward?

Sam (8m 31s):
I would love nothing more than to have a CFO who could help be calling the shots. I don’t have a CFO right now. So it’s just me doing a poor imitation and approximation of it, you know, early stay sharp. We’ve got 11 people right on the team and most of those are engineers. So yeah, I mean, I will say I, I messaged HubSpot’s original CFO after I launched this and said, I never truly appreciated him until I had to be at my own company without a CFO. So it is actually going to be probably our, it it’ll definitely be our next executive hire is somebody who can, you know, not, not just like do the accounting stuff. There’s a big difference between like accounting and finance strategy.

Sam (9m 14s):
Somebody who can tell me like what’s the right decision to make based on, you know, sort of the economic frameworks of the business. I’m not going to get good at that. Right? Like don’t like, nobody can be good at everything I, I gave up on being a good programmer to about 15 years ago. And that’s a relationship that, you know, I’m looking for, if anybody knows anybody who is tired of like, you know, their gig and wants to go work at an early stage startup, definitely let me know. Cause it’s a, it’s a fun math problem to work on.

Jeffrey (9m 45s):
Wow. Very interesting. I’m just wondering, you know, going back to this whole startup thing, is there anything that you would’ve done differently?

Sam (9m 55s):
I would have the CFO earlier. I would have don’t over-engineer the problem. So we went into this thinking that we had to fight Facebook cause we know Facebook has a place-based advertising team cause they tried to hire our CTO, Andre to run it. But you know, the people in this industry are still using spreadsheets to manage their inventory, or it’s just, you know, if you want to buy, if you want to reach people who work at hospitals in Pittsburgh, what billboards do those people drive past, especially in, you know, tech startups. We like to build really cool, interesting technology.

Sam (10m 34s):
And when I think it was when Google invested in HubSpot, they threatened to tattoo. We are not the user on the inside of our eyelids. HR. Didn’t let them, don’t worry about that. But you know, really understanding what actually creates the most value for our customers and not just what we’re most interested in building would have saved me a lot of time and money on engineering. A product features that didn’t matter nearly as much as having like a directory of all the out-of-home agencies that people could work with.

Jeffrey (11m 4s):
So Sam, I know you, you haven’t seen this, but I’m cutting Kevin off and he wants to keep asking you questions. So you’re going to have to come back again and be interviewed with us again. Cause we find this very interesting. If somebody is interested in you and your services, how would they find you?

Sam (11m 21s):
Well, the good thing about my last name, Evan, you did a great job is SEO is very easy. Google, anything even close to <inaudible> and you’ll generally find me. You can also check out one, just don’t blame our new head of marketing for the website. He just started. And I built that, which is why suffice to say HubSpot had a very talented design team, but I wasn’t on it.

Evan (11m 43s):
Great. Well, I’ll see him. Thank you very much for being on this show. You know, it was very great to hear your story and I’m sure our listeners out there really enjoyed it as well.

Subscribe to our Podcast!

purple circle podcast icon

Apple Podcasts

green circle with white curved lines for sound waves


multi-colored vertical lines in a diamond shape

Google Podcasts


Find us on Social Media

rounded blue square with lowercase white letters "in"


rounded red square with lowercase white play button in the middle


rounded blue square with lowercase white letter f


rounded light blue square with a white silhouette of a bird flying