He made 2 key changes —then grew to $100M ARR in 2 years & exited for $2B. | Harish Abbott, Founder of Deliverr & Augment

Harish Abbott (00:00:00) :
The moment these two things happened. I mean, I'm kidding, you know, we went from trying to find the next merchant who would use us to thousands of merchants.

Pablo Srugo (00:00:07) :
Wow.

Harish Abbott (00:00:08) :
We were short on warehouse supplies and it became a supply constraint. I mean, I think this in the first year, we barely made any money and I think by year three, we were over a hundred million plus.

Pablo Srugo (00:00:20) :
Wow.

Harish Abbott (00:00:20) :
I think we went in two years from almost zero to a hundred million. In certain businesses, it's not product market fit, it's product-price-market fit. Your product could be fitting, but the way you price it may not create the fit and so you have to deeply think about, am I pricing it to really enhance my product market fit?

Previous Guests (00:00:42) :
That's product market fit. Product market fit. Product market fit. I called it the product market fit question. Product market fit. Product market fit. Product market fit. Product market fit. I mean, the name of the show is product market fit.

Pablo Srugo (00:00:55) :
Do you think the product market fit show, has product market fit? Because if you do, then there's something you just have to do. You have to take out your phone. You have to leave the show five stars. It lets us reach more founders and it lets us get better guests, thank you. Harish, welcome to the show, man.

Harish Abbott (00:01:12) :
Great to be here. Thanks for having me.

Pablo Srugo (00:01:14) :
Dude, so you're a multi-time founder and two. Well, one massive success, a company called Deliverr. Which raised like $490 million, was acquired for over $2 billion by Shopify and now you've started something that was kind of more in the fulfillment space. We'll get to that, then you started something that's more AI for logistics. As I understand it, literally twelve months ago today, right? So last October, and already raised over $100 million. Which is just an incredible feed by any standards. So excited to chat with you today. Let's go back to the beginning, right? Maybe give us just a little bit of background before starting to Deliverr. Who are you? What were you doing at the time? And maybe just the origin story around that.

Harish Abbott (00:01:50) :
You know, I started off my career at Amazon. Back when Amazon just sold books. Biggest bookstore was our tagline.

Pablo Srugo (00:01:57) :
What year?

Harish Abbott (00:01:58) :
'99.

Pablo Srugo (00:01:59) :
Oh man. 

Harish Abbott (00:02:00) :
Yeah.

Pablo Srugo (00:02:01) :
What a time to be at Amazon, that's crazy.

Harish Abbott (00:02:03) :
Yeah, it was great. It was maybe a few hundred people in the engineering side. So it was very small teams at that time. We just got the chance to go build a lot of software, which now runs Amazon Fulfillment. Which is presumably right now the biggest fulfillment company, at least in North America.

Pablo Srugo (00:02:22) :
How long were you at Amazon for?

Harish Abbott (00:02:24) :
I was there for about five years.

Pablo Srugo (00:02:26) :
Okay and is it true that during that time, like Amazon is such a behemoth today. It seems inevitable but as I understand it. Especially after the dot-com bust. It wasn't clear that Amazon would even survive. Is that true? What was it like inside?

Harish Abbott (00:02:38) :
I mean, I think the outside world was all about. I remember Time publishing an article saying Amazon.bomb. Which was like, hey, it's all going to blow up and it's a bubble. But inside, I mean, I have to give credit to the leadership there, Jeff Bezos, Jeff Wilke. They continue to see the massive opportunity that internet will afford for online commerce and almost doubled down on innovation. When everyone else was pulling back. You know, like in 2000, when the stock market had crashed and the internet stocks, and the darlings had all. For one time, really retracted back, Jeff Bezos and Jeff Wilke, they just said, hey, we're going to double down. Because we believe in this long term. This is a better way to shop. This more convenient, you can have more selection and you can have better prices long term.

Pablo Srugo (00:03:29) :
And they found a way to kind of stay profitable through it, right? To not fully depend on markets. Is that correct?

Harish Abbott (00:03:34) :
I wouldn't say profitable from the very beginning, but I think the leadership there really had a very good plan to either raise capital or raise debt. But it was in a very disciplined way. It wasn't like, hey, let's do this and maybe we'll figure it out if it's going to work or not, you know? It was very like, hey, if we get to this scale, our cost of goods will go down to this level, and then we will be clearly profitable. So let's get to that as fast as possible. So our cost of goods, you know, the way we were buying our books or at that time, even CDs or DVDs and so that I think is. Actually a big learning, I took away from that, that you can have responsible, fast growth. If you have a very disciplined way of looking at your numbers and how the numbers are going to trend up as you scale. And use those economies of scale to continue to drive your unit economics over time. Actually it helps you build a more defensible business, right? At today's scale, I would argue Amazon is an extremely defensible business, you know, for someone else to go build that scale in the online world extremely difficult

Pablo Srugo (00:04:39) :
I think its true. Obviously there's a trade off between growth and profitability. But sometimes its over stated and I think its a luxury. But some companies are in a position, where you can have both. You can have sound economics and either profitability or close to it. And still have pretty exceptional growth. Certainly from a compounding perspective. Because if the hold up fifty to a hundred percent. A few years straight, you get pretty big.

Harish Abbott (00:05:00) :
That's right, yeah. Especially true when there are technology transformations that are happening, you know, at that time. The internet transformation was happening and even during those times, I think lots of new companies emerged. Like what we're seeing today in AI, hundreds and hundreds of startups emerge. Lots of capital flows in but as the dust settles, I would think the best-run companies stay. I would argue companies who focus the most on the customer stay, right? But you have to continue to grow through this in the transition period, because you could be focused on customer. But if you're not in the top two or three then, I think you get left out of the race.

Pablo Srugo (00:05:40) :
So what happens after Amazon?

Harish Abbott (00:05:42) :
So I built a few businesses since, I went into the entrepreneurial side. I thought I had learned a little bit about how things work and was very inspired confidence in delivering good software. So I built a business which then merged with called Lulu.com. One of the largest marketplaces for independent authors. I build a business called Symphony commerce, which was this middle infrastructure layer called order management layer for midsize and large brands. And the last one that I built, which was called Deliverr. Which was very much like Amazon Prime for small merchants.

Pablo Srugo (00:06:18) :
And that was your biggest one, right?

Harish Abbott (00:06:20) :
That's my biggest, yeah. Biggest business and Amazon is like, America is a very large country. And we have a very low density per square mile. And the only way to affordably deliver e-commerce goods fast, predictably and reliably is by having inventory in close proximity to demand, right? So if somebody's in Chicago and they order, I don't know, Tide detergent. If the Tide is in Chicago, then I could use a local delivery provider and get it to them in one day. And it actually turns out that's the cheapest way to get it to them versus having the Tide in, California. And having it flown all over or driven all over the country. And that gets very expensive.

Pablo Srugo (00:07:03) :
Just from a timing perspective, you started Deliverr when exactly?

Harish Abbott (00:07:06) :
2017.

Pablo Srugo (00:07:07) :
Today, that's widely known and understood. I mean, even in Ottawa, we have on the East End. A massive Amazon warehouse, and on the West End, another massive Amazon warehouse. So I think everybody gets that, but in 2017, was this kind of a bit more of a unique insight. At least in the e-commerce world, that you needed to set these things up to really do fulfillment at scale?

Harish Abbott (00:07:27) :
I don't think it was a unique insight. I think Amazon was doing this for a while. I think what was unique was, could we apply it to a small merchant?

Pablo Srugo (00:07:35) :
I see.

Harish Abbott (00:07:35) :
You know, and what was unique is that thanks to Shopify. The number of small merchants who were building businesses online were just expanding rapidly, right? The number of new brands that were emerging was expanding rapidly, and so the combination of that. This new emergence of brands online. Through these very highly democratized platforms like Shopify were coming up. But then they also needed to offer something fast, affordable, reliable to their customers and how could you offer this service. Which inherently distributes your inventory five, six ways, maybe eight ways in the country. For a merchant who may not have that either level of inventory. Suddenly doesn't have that infrastructure themselves. So the only way is you have to pool thousands of merchants of inventory to then be able to forward deploy it and distribute it. And offer every single one of them a service comparable to what Amazon offers, right? I think that's sort of the unique insight that I would say at that time. But the method had been pretty well mastered by Amazon by that time.

Pablo Srugo (00:08:40) :
Is this the same thing as third party logistics, 3PL? Or is this your one version of that? Or was that a space that just happened over the last ten years?

Harish Abbott (00:08:48) :
No, we had a version of it. We were just very focused on delivery speed. So we just said, you know, a lot of 3PLs is like, hey, I don't want to pick and pack or store my things. So I'm going to give it to a business. We were saying, hey, that's okay, but what is really needed is to be able to offer a one-day and a two-day reliable delivery to your customers. We suppose if you do that, then your sales go up.

Pablo Srugo (00:09:13) :
Yes, right.

Harish Abbott (00:09:14) :
Where do online brands spend most of their money? Marketing, right? Either for most brands, that's the number one line item or the number two after cost of goods sold and one way to improve your marketing spend is that. When people land on your website through an ad and they see that you can get this item tomorrow. Their conversion rate goes up, right? Versus saying, hey, this item will show up in seven days and they're like, I don't know if I want to buy it.

Pablo Srugo (00:09:40) :
Would you say that's one of the things that you understood better than, let's say, most at that time? Today, it's taken for granted.

Harish Abbott (00:09:46) :
I think so, because I saw that power in Amazon. In Amazon, every time we dialed the speed down, our sales went up. There was never a time when we dialed our speeds down to say like, oh, could we do three day versus two and a half day, versus two days versus one and a half, versus one day. Every single time the delivery probably shrunk, the sales went up and it's just like, it's very common sense, right Pablo? Nobody says, oh, it would be great if I could get this item in ten days. Nobody likes slow shipping but everybody likes that, hey, what would be super cool is I can get it today, I can get it tomorrow, and I can try it. And if I don't like it, I can return it.

Pablo Srugo (00:10:28) :
I think the idea is obvious, but just the importance of it, and especially back then. I actually remember, funny enough. My last startup, we brought in a professional CEO at one point, and he loved Amazon. Actually back then, this is literally 2017. I remember one day he said to me, what other merchants don't get and what Amazon gets so well is, the sale is not done till the thing is at my house. Like, I'm buying something, I want it now. You know what I mean? They get that and it's true. If you have to wait, the old school way of thinking would have said, well, whatever, you know, you don't need the thing tomorrow. I'm not talking about a meal that you're gonna eat. It's fine, it comes in seven days. But there's something about the UX that makes it so different when you get it tomorrow versus seven days. Now it's fully understood. Being able to power that eight years ago. There's some alpha there. There's some kind of unrecognized value, I think.

Harish Abbott (00:11:12) :
There is, you know, if you think about what is online commerce in some ways competing with, is offline commerce and what do you get in offline commerce? You get instant gratification, like, you walk into a store, it's inconvenient, you have to park, your selection is somewhat limited, you may not find your size or the color you were looking for. Online solves for all of that. Where offline is still slightly better is, hey, when you do find the size and the color. You're like, you pick it up and you love it. You're like, hey, I've got it, you know? And so how do you counter that? You counter that by reducing your delivery speeds, right? You say like, hey, if I can get it to you today or first thing tomorrow. That's pretty close and you can try it on in the comfort for home. And you don't have to go find parking in a parking lot or, sift through thirty minutes to find your size and color. You can pinpoint that. It's a better way of commerce if your delivery speeds are faster and faster, right? I think that was a unique insight and the other insight was that the online brand space will continue to grow. Shopify itself now is, I believe, twelve percent of all U.S. e-commerce. You know, just all the merchants on Shopify. It might be more or less, I don't remember the last number yet.

Pablo Srugo (00:12:22) :
But it's $100, $150 billion company today. Back then it was what, $2 billion? It was pretty small, I think, in 2017.

Harish Abbott (00:12:28) :
It was a smaller company. I don't remember the size of it, but it was a smaller company.

Pablo Srugo (00:12:322) :
Yeah.

Harish Abbott (00:12:33) :
But you could see how these brands were coming up and they were building higher quality product. And they were just becoming this thing that most people started to gravitate to buy from. And what they had going for them was just higher quality product, and amazing brand story. They just didn't have great fast fulfillment options. So we were like, okay, Deliverrcan fill that need.

Pablo Srugo (00:12:53) :
And so how do you start? Because it's not a traditional SaaS business at all.

Harish Abbott (00:12:58) :
It's not, yeah. So, well, I think we said, hey, what are the things that we're going to be good at and what are the things we're just going to use partners, right? Because it's very complex to build an infrastructure like this and to execute it for a young company. And so we just said, like, hey, we're going to be good at one thing, which is placement of inventory. We will predict what New York is likely to buy of an item, and we will orchestrate getting that item in New York. And then we'll be partnering with warehouses. So we won't run the warehouses. Other people will run the warehouses. We'll put our software there. We will tell them which items are coming in. What do they need to pick and pack. When we would use their space, their labor and we would use third party delivery companies, and orchestrate, and create a network of last mile delivery companies. Because we would have forward deployed inventory anyways. So now we're not doing as many cross country shipments. We're doing much more local New York to New York shipments or Chicago to Chicago shipments.

Pablo Srugo (00:13:55) :
And your customers would be the merchants? Or the warehouses? Who would pay?

Harish Abbott (00:13:59) :
Our customers are the merchants. So the merchants will give their inventory to us. We would decide how we are allocating that inventory.

Pablo Srugo (00:14:07) :
Gotcha.

Harish Abbott (00:14:07) :
The warehouses will be one of our vendors. Where we would pay the warehouses to do the service. Very much like how Uber executes, right? The customer is the person who's getting the ride and the driver is one of the vendors of Uber who, you know, they pay for the service level.

Pablo Srugo (00:14:24) :
That's kind of the idea. How do you start? Do you raise money right away? Do you build a product? How do you get your first few merchants on board?

Harish Abbott (00:14:30) :
I think we did a bit of both. So we raised some money with this idea. That there is a need for an Amazon-like infrastructure for this very emerging world

Pablo Srugo (00:14:38) :
How much did you raise?

Harish Abbott (00:14:40) :
I think I started off with $7 million, my first raise.

Pablo Srugo (00:14:44) :
And a lot of that, I assume, had to do with your background at the time. It made it relatively easy for you to raise.

Harish Abbott (00:14:48) :
Probably, yeah. I mean, that helped a little bit, but I think the ideas were timely. In fact, we raised it from a fund called 8VC. Which is founded by Joe Lonsdale and they were coincidentally thinking about the same idea. They also incubate companies in-house. When I had pitched them, they're like, hey, we're thinking about something similar. Your thinking is a little bit more evolved than what we were thinking, because you're from the space. So instead of us incubating it, we would basically just back you. So some of it was coincidental. We basically said, we'll have four warehouses, one in the Northeast, one in Texas, one in the West Coast, one in the Midwest to begin with and so we went, and got four warehouses. We wrote a bit of a software that can now take a merchant's inventory, split it four ways.

Pablo Srugo (00:15:32) :
And by the way, you're guessing or I should say estimating. What gets sold where just based on their past sales and time of sales?

Harish Abbott (00:15:39) :
Yeah, we basically hooked up into their data, whether it's through Shopify. Only to understand where their demand is. We started to basically tell merchants, hey, if you use Deliverr, you're going to get faster shipping, and with that, your conversion rates will go up. And so we tried that, I think for six months. I think we got like ten merchants to use us. There's not much uptake into the idea at all.

 Pablo Srugo (00:16:05) :
Yeah, were they at least big merchants? Did they need to be of a certain size?

Harish Abbott (00:16:08) :
No, very small. I mean, the big merchants wouldn't trust a new company with their fulfillment. You know, think about from a merchant's point of view. One of their biggest assets is the inventory, right? If you're a merchant and you're selling $100 million worth of inventory. Parting your inventory to a startup that you've never heard of, it's a pretty big deal. You need a lot of trust for them to be able to say, oh, I'm gonna give you my $50 million worth of inventory and make sure you will take care of it, you know?

Pablo Srugo (00:16:34) :
But funny enough, I have to assume your software would probably add the most value to the big merchants. Just because of the amount of data.

Harish Abbott (00:16:40) :
It would, except that, you know, how do you build trust? When you don't have a fully functional product, you haven't taken out the kinks, you don't have a brand name and so you have to start with smaller merchants. Where we can handhold them and we can say like, when things are not fully working, we will make sure your orders get out, we'll make sure your inventory is saved, we'll give you daily nightly inventory reports and that's better than maybe your in-house half a person team that you are putting on this performance system of yours. But man, the first nine months or so, we didn't really get much demand on it. It was very small merchants. So we basically interviewed lots of merchants, talked to a lot of them, both our customers and people who said no when we pitched. Like, why is that? Why are you not interested and I think the key insights that we got from that were. I think, twofold. One was the idea of a fast promise is good, but I need to tell my customers and we were not able to offer like, hey, it's two day nationwide. Because they did not have enough inventory. I cannot say on my website it's two-day nationwide. It's sometimes two days in New York or around New York, two days in California, or one day in San Francisco. If I can't tell my customers, then I'm kind of paying for something that my customers don't know or care. So you've got to solve for that.

Pablo Srugo (00:18:01) :
You only get it at the absolute last part of that funnel. When they put in their delivery address and then you say, oh, it's two-day shipping. But by that point, they're already buying anyways.

Pablo Srugo (00:18:09) :
Exactly, so that was one big problem that they all highlighted. I think the other big problem was they said like, hey, the only way we deliver this one day and two day is we would distribute inventory. So they would give us inventory in LA or the ports and we will take it out. And we'll basically run some trucks that would get that inventory out to these things. Our pricing was, we will charge you for fulfillment, pick and pack, and shipping. You would also have to pay for the distribution. Because we're taking your inventory from LA and putting it in New York. And for them, it was like, it's too complicated. We don't know the benefit yet and now I have to pay you for moving my inventory to New York or moving my inventory to Chicago a little bit. Should we, should we not? Is it going to boost my sales or not? You know, so very, very valid concerns, right? In our minds, we were like, hey, we're doing fast delivery. The only way we can do this is by deploying inventory close to demand and they should pay for it. It's a fair game, but their perspective was very valid. You know, when you really peel the onion and you get a few very open customers with you. It is night and day difference in your thinking. And so we made I think two big changes, based on these learnings. I think first change we made was, we made a prime like badge that you can deploy on every product. So we would then detect if the person is viewing it in New York. We would turn that into a, oh, get this order in, two hours, get it by tomorrow, right? What you see on Amazon today everywhere. Amazon detects where you're at and based on that. They map it to where the item is at in the warehouse and then they give you a real time promise when you're looking at a product.

Pablo Srugo (00:19:53) :
Right, you moved it higher up in the funnel. So you actually can add to conversion.

Harish Abbott (00:19:57) :
Higher up in the funnel, exactly, exactly. We added into the filters, we added into the search. So that people could filter and say, show me all items that I can get tomorrow. So a lot of that visibility started to happen and because the promise is dynamic. Which is in Amazon's case too, is very dynamic promise. If the item is not in Texas and you're in Texas, they might give you a three-day promise. The item is in Texas, they might give you a one-day promise. So the promise was very dynamic, but you had to show it to the user. Who's viewing the product really up in the funnel. The second is, we changed our pricing to be flat. So we just said, for an item depending on their weight and size. You just pay us this. We're going to take the risk of distributing your items. Everything is inclusive.

Pablo Srugo (00:20:43) :
Including the final delivery?

Harish Abbott (00:20:44) :
Everything is one price, right? We just pay one price and in the fulfillment industry. There's sort of two things are true. One is the pricing is usually done by pick, pack, and ship. So, you know, the warehouse charges you and then the carriers charge you on a zone basis. Like, hey, if your item is going to zone one being short distance, zone five being long distance, you pay this much by weight of the item or volume of the item, and we sort of went a little bit further. We said, we're going to combine the two. We're going to eliminate the zones itself.

Pablo Srugo (00:21:17) :
You're there, everything is zone one. That's kind of by definition.

Harish Abbott (00:21:20) :
Yeah, distributing inventory. So if we are wrong at distributing inventory and our algorithms are messing up. That's on us, not on you and if we have to do more zones for your items. That's on our systems, not on you. You now get predictability of pricing. So you can now model your P&L and say, if I ever sold this item. This is my COGS, this is my fulfillment, this is my marketing, this is what I make and it turns out, that was a huge need in the market.

Pablo Srugo (00:21:48) :
Yeah, that's huge.

Harish Abbott (00:21:49) :
Right? Because now every business can very quickly say, I sold this item for $10 or $15. I paid $5 for fulfillment. $4 for my, I make this much money.

Pablo Srugo (00:21:58) :
Do they pass that on to consumers? Where it's like $5 shipping or most people just include it?

Harish Abbott (00:22:02) :
They could choose to, it's up to them. We left it up to them, it's their decision and even if we took the speed argument out. Just the flat shipping argument or the flat price argument, like changed the whole game, right? Even that was better than their status quo where they would ship it to one warehouse but then they would be surprised every month on how many zones did I ship? Oh, this one shipped seven zones. Oh my gosh, I lost money on that order. Because it shipped seven zones or this one shipped three zones, oh my gosh, this was great. It was just like, they couldn't model their business very well and so by taking on that risk. We were able to give them predictability of their financials. Which was, huge and trust me. We did these two things and then we did a little partnership with Walmart. Where we said, hey, we have fixed pricing. Your merchants at that time did not have FBA like equivalent. Where Amazon guys have FBA. You can use Deliverr for Walmart merchants to sell on Walmart but get one in today and Walmart obviously likes that. Because If their customers get to see a fast shipping promise, they'll buy more and we did that. And suddenly most of merchant sales was like 2x, 3x. Because the delivery promise was there. Plus, the pricing was flat. The moment these two things happened, I mean, I'm kidding. We went from trying to find the next merchant who would use us to thousands of merchants.

Pablo Srugo (00:23:25) :
Wow.

Harish Abbott (00:23:25) :
You know, signing up too. We were like short on warehouse supplies. It became a supply constraint

Pablo Srugo (00:23:32) :
And this is what year, at this point?

Harish Abbott (00:23:34) :
I think it became like 2019.

Pablo Srugo (00:23:35) :
Two years in.

Harish Abbott (00:23:37) :
We became supply constrained at that time.

Pablo Srugo (00:23:39) :
What was revenue growth like? Once that inflection point happened?

Harish Abbott (00:23:43) :
I mean, I think that's in the first year we barely made any money and I think by year three. We were over a hundred million plus.

Pablo Srugo (00:23:50) :
Wow.

Harish Abbott (00:23:50) :
I think we went in two years from almost zero to a hundred million.

Pablo Srugo (00:23:53) :
What are margins like in this business?

Harish Abbott (00:23:55) :
They're low, very slim, right? So on a unit economics basis, you're looking at fifteen percent. In the beginning, they were even lower, but then as we started to gain scale. We started to gain some margins. I think at scale, maybe you can get up to twenty, twenty-two percent in that range. Yeah, so it was a big, big lesson for us, right? There's these little things that you presume, but it's not clear to the customer and once you make that clear. But the reality is that to do flat pricing, we had to rethink the whole system, right? We're now deeply investing into how the items are distributed. We're deeply investing into making sure we have space available in all parts of the country. We then invested into building a system that could crosstalk inventory efficiently. So we would run trucks nightly across the whole country.

Pablo Srugo (00:24:43) :
I mean, you shifted the risk. You took on the full risk at that point. Because if you do it right, it's great. But if you make a mistake, you're eating all that cost.

Harish Abbott (00:24:50) :
So you had to re-architect, you had to rebuild your engineering teams, your infrastructures to focus on these type of problems. To the merchant, it looks very simple, right? Like, hey, I've got a one-pound item. I always pay you six bucks and never have to think about it. Plus, I get fast shipping. Hey, great, sign me up, and I think there was this separation between the product is what the customer sees, right? Underlying, there's just a lot of complexity of algorithms, pricing, capacity management, running intra-city trucks all the way every night and we had to keep taking that on, and on to build predictability. And maintain good margins for us. But yeah, it was a good lesson in product market. Which is just this little less change. The other big lesson for me here, Pablo. Which in certain businesses, it's not product market fit, it's product price market fit, right? Your product could be fitting, but the way you price it may not create the fit and so you have to deeply think about, am I pricing it to really enhance my product market fit?

Pablo Srugo (00:25:55) :
We have tens of thousands of people who have followed the show. Are you one of those people? You want to be a part of the group. You want to be a part of those tens of thousands of followers. So hit the follow button. You could argue that pricing was so important to your product. It was almost part of the offering. Like it's part of the product. The value you're delivering is the fact that it's a flat price for delivery. Which is something they're just like not used to.

Harish Abbott (00:26:17) :
Yes, I ran pricing for the business till we got acquired. It was so important. Every single pricing change I literally obsessed over, because it changed everything. What we invest in internally and then you of course have to build systems when you have thousands of merchants joining you. You can't serve them, with people. You have to build extremely self-serve systems, right? Which is another big, at least for me, learning is that as you started to get demand, we're like, hey, we can't have a person on a phone call or an email now onboarding a merchant. But now you're onboarding a merchant who's giving you $10, $20 million worth of inventory without ever talking to anybody.

Pablo Srugo (00:26:57) :
Wow.

Harish Abbott (00:26:58) :
Right? So how do you build very high trust onboarding systems that are self-serve? At one point we had 7, 8 thousand merchants or something like that. How do you onboard merchants where they're never talking to anybody, just going on and sending you inventory. And then you're telling them like, hey, you sent us fifty thousand units but we got forty-five thousand nine hundred, here's the proof. So you had to build a lot of self-serve systems for it. But systems that inspire trust, because systems that are very high visibility and transparency inspire trust. Systems that don't have that do the opposite. So our product roadmap started to change to do that, right? How do you inspire trust in your customers? And so any signal we got, any time we received inventory, any time there was defects, any time there were damages. Okay, we had to get pictures, we had to upload them, we had to share it with the customers or merchants. They had to be able to see it. They had to be able to dispute it with their manufacturers. That, hey, this whole box was defective. Because sometimes they would never see that inventory, before they came to us. They would just tell their manufacturers to send the inventory to us, right? So I think once you get on something, your product direction and roadmap changes with that. You have to follow it since it's entirety. You can't give it up halfway. Even if you had given up, in my view. The idea that this whole transparency visibility thing, we're just going to run it through customer service, right? I think we would of failed at that point, right? Just the cost of that and then people not trusting us. And saying, oh now I have to talk to somebody, and I don't know if they understand it. Now you have to train your people, versus just building systems where everything is there for them. They can see everything, they can get alerts on everything and vice versa. I think my other learning was that once you build self-serve systems. Man, they create growth for you like you sometimes don't imagine, right? Like somebody at 3 a.m. getting frustrated with his fulfillment company and finds Deliverr, and in two clicks can onboard themselves, and sends inventory. Great, like, we had merchants starting to sign up from Thailand and China and places we would have never imagined to go. And they would have these things where they would translate our English into their local language. Like a little widget or a Chrome extension and they would run the whole thing. Their whole business.

Pablo Srugo (00:29:18) :
Crazy.

Harish Abbott (00:29:19) :
And we would have never spoken to them. So I think self-serve systems are beautiful things and I mean, companies like, Stripe and Shopify, and others have really shown the word on how powerful self-serve is, if you get it right.

Pablo Srugo (00:29:32) :
So I want to make sure we have enough time for current company, Augment. But let's just fast forward and just maybe take a couple minutes to talk about. Because everybody always wants to know how the story ends, right? So how did the Shopify acquisition happen? Maybe just set some of the context of how those discussions started. Getting an acquisition actually done, especially at that scale. Multi-billion dollar scale, is not trivial. No matter how well your business is doing.

Harish Abbott (00:29:54) :
Yeah, so I think. Listen, an increasingly large number of our customer base was Shopify. Because Shopify was growing so fast and their merchants needed this service the most, right? We did have lots of merchants from marketplaces like Walmart and eBay. Even some Amazon merchants who would use us as a backup to FBA. But I would say a dominant percentage of our merchants were Shopify and so as we were scaling. I'm sure like Shopify being such a data-driven company, they were seeing more and more merchants using us. Shopify is a very merchant-obsessed company. I actually never met a company as obsessed with their customers and merchants specifically as Shopify is, and so from their perspective. What was interesting is one of the big bottle necks for you in growing your business. Because if their merchants business grows, clearly Shopify business grows too and Shopify as a platform has addressed payments. And addressed building a really seamless software, but fulfillment or physical world was one area where merchants would routinely fall behind. So the idea was like, hey, with Shopify scale and might. Could they elevate, Deliverrto a level that would really boost their merchants' businesses and that would really help their merchants, but also help us. And we had spoken, on and off for a while. Either for investment rounds or partnerships before.

Pablo Srugo (00:31:16) :
Is this with Toby, Harley? Or is this like corp dev teams, these kind of discussions?

Harish Abbott (00:31:21) :
I think we started with Corp Dev first, for some partnership stuff, and then their product teams as well on the partnership side. So that's how we were discussing and then more towards the end as we were getting closer to this. I spent some time with Toby and lots of product leaders there. Shopify is also a very product-centric company, right? Product and engineering really is a big powerhouse, and how things are gonna come together. Technology is very important. Toby himself is a very product-centric, engineering leader, phenomenal leader. Yeah, so as we started to discuss, I think it just made more sense to do sort of a partnership. Be part of Shopify, and add the superpower to their merchants, you know?

Pablo Srugo (00:32:03) :
Presumably, did you stay there? You stayed there probably for a while before you started Augment?

Harish Abbott (00:32:07) :
That's right. I stayed for about a year, year and a half. And then I think as AI was coming along. You know, if you look at my journey. I'm a technologist, but then have stumbled upon this world of logistics, supply chain. I've learned this space. It's a very humbling space too, Pablo. The moment you think, you know it kicks you in the butt, like, oh my gosh. It's very complex, you know, physical world is. As engineers, we think writing code is transactional in nature, right? So you can write something and close a transaction database, and everything is all set. The physical world is non-transactional, right? So you have things that you think are going to happen but don't happen. Somebody could pick an item in a warehouse and scan it. And just leave it or the shift ends, they just leave it and move. And now in your systems you had that item attached to an order, but in reality it's just sitting in a warehouse. And you add that up over, let's say, ten thousand items in a warehouse. It's certainly like, oh my gosh, this is a huge problem. So marrying the physical and the digital world is a very complex endeavor. I don't think anybody has truly solved it and as AI was coming along. I'm like, hey, this is an interesting thing where we have one of the largest industries, logistics, and supply chain in the world. That is still mostly run on emails and phone calls, and text messages. And partly because it's very fragmented. There's so many parties and there's only common ways for them to exchange information with each other is through these common mediums. Which is very draining for operators, but also very wasteful for the whole system and so. The idea was that I was both a builder and a buyer of logistics before, and I've seen these pains for ourselves. I could deliver, we were buying tens of millions of dollars worth of trucking. For instance, every year and it was very painful, to buy and to get visibility into trucking.

Pablo Srugo (00:33:58) :
Right.

Harish Abbott (00:34:00) :
I'm like, oh, AI is a unique way where we can at least solve this problem by meeting people where they're at. So far, the attempts have been, hey, there's a problem. It's well-recognized. Logistics is highly fragmented. It runs on emails, phone calls, and text. We're going to solve it. But the way we were going to solve it is, we're going to build a new system. and because it's so fragmented, you just cannot have everybody join that new system. AI was a way where like, hey, we can meet you where you're at. It can do your emails, it can do your calls, it can do your text, it can interact with your systems of record. It can start to take away all this busy work. So you can focus on more creative problem solving versus chasing documents, chasing emails, tracking trucks, setting up appointments, things like that and that opens up windows to do a lot more. Like, you have lots of interesting data about that business. You can now have AI on two sides collaborate synchronously. You can optimize chains better. So it just made sense to say this is a transformative moment in time and in an industry that has mostly resisted tech. To actually go transform it by building tech that meets them where they're at.

Pablo Srugo (00:35:12) :
Got it. I'm curious, what keeps you coming back?

Harish Abbott (00:35:15) :
I mean, I think it's just the love of the problem. Sometimes you're in a video game and you keep playing it and it's just the love of the problem. Can I crack it? It's not like, I would say, the love of the solution. I like people who just obsess about the problem and then there's a possibility that the solution they start off with today, is just no good tomorrow. Because tech has changed so much, but the problem remains and this is a pretty gnarly problem. Logistics is a very tough problem, but it's also a consequential problem, you know? The coffee mug on your table, like, twenty different companies had to collaborate for that coffee mug to get on your table, right?

Pablo Srugo (00:35:49) :
Right.

Harish Abbott (00:35:50) :
From maybe Asia to Canada, and lots of things can go wrong in that. But it's consequential because everything you touch, any store you go to, or you look around your home or your office, anything you see. Logistics and supply chain had to touch it for it to be where it is at today. So I would say it's just the love of the problem, love of building, it's that and also at this time. It was a bit different, where we're going through such a big technology transformation that I think the biggest in my lifetime. That I would regret not building with it.

Pablo Srugo (00:36:20) :
How do you start with this one? I mean, and you're building AI or voice AI specifically into logistics? Just to be clear.

Harish Abbott (00:36:26) :
No, we're building AI. So we're building, our product is called Augie, which is an AI teammate. Think of it as almost like a remote employee that you can have. It can do emails and phone calls, and text, and it can interact with your systems. It can message, just like a remote employee could, right? They could do all of that and you can give Augie work through SOPs. Just like you would give work through standard operating procedures to remote employees. So it's very no different. You just get Augie that works 24/7. Apparently like thousands of people and they can really get work done for you. Regardless of modality, regardless of the systems you use.

Pablo Srugo (00:37:03) :
And specifically for logistics companies?

Harish Abbott (00:37:06) :
Specifically for logistics companies, yeah. So it understands the word of logistics really deeply. Logistics is also very large, right? It has warehousing, it has trucking, it has freight forwarding, and distributors, shippers. And we're starting off with freight, which is trucking. Within that, there are brokerages, there are fleets, and there are shippers. So we are targeting that subsegment.

Pablo Srugo (00:37:27) :
We had a company called Happy Robot that came on here. They were doing voice AI for, I think, for forwarding. Are you familiar with them?

Harish Abbott (00:37:33) :
Yeah, yeah, yeah. I think they're very much more focused on the voice side.

Pablo Srugo (00:37:36) :
Yes.

Harish Abbott (00:37:37) :
I think our view of the world is that it's a very multi-modal world. Some problems are better solved through text. Funny enough, some problems are better solved through Telegram. In this business, there's a lot of dispatchers in Eastern Europe who only use Telegram. Similarly, there's lots of people who support this industry in India. They only use WhatsApp. They run their entire business on WhatsApp, right? And sometimes some problems are better solved through email, some are better solved through internal messaging. And so ours is like, hey, it's a multi-modal world. And then also end to end, order to cash type of surveys where it's not just like, we're tracking loads for you or we're taking bids for you. It can do all that. It can make calls, but how do we look at order to cash processes for your business? Because we believe if you do that then you carry the context through and you don't have to have five agents who just don't talk to each other. You know, it's like imagining if you have fifty very smart employees, but they never talk to each other. How chaotic would that be? So what you do want is the AI that is helping you in one process and after they're done. Transfer that context to the next process and so on and so forth. So then they are in sync of what is happening for that shipment or for a business, or a contract, whatnot, you know.

Harish Abbott (00:38:50) :
So walk me through one of the things that people want to know the most about, is going to market, right? How do you get those customers? And in your case. I mean, you've had so much experience and you know this place so well. Once you have this idea that AI can transform a lot of the pieces of the puzzle in this logistics world. How did you decide to go to market with Augment?

Harish Abbott (00:39:06) :
In this case, I think what we did is we went out to get. I would say, a few design partners slash customers, right? Because we didn't want to pretend to know the whole world. I mean, I think I know the world, but once you get to second, third, fourth layer. The only way to know is to let breathe, the operator, that business. We basically went to a few customers and said, hey, if you become our design partner. We have ideas, we have a little bit of a product working. It's not there yet, AI is going to be important. If you follow our vision, this is how it could look like in a year's time, two years' time with you. What we would want from you is time, of your people that we can sit down with and truly understand their word. Deeply understand, not from a CEO's perspective, not from a VP's perspective, but actually from the operator's perspective. So we want to be able to do that. In return, we will build software to your needs, AI to your needs, and you will be one of the early adopters

Pablo Srugo (00:40:04) :
Was there resistance from the end users? Because everyone's scared of losing their jobs to AI?

Harish Abbott (00:40:09) :
I would say there's a little resistance, but not a whole lot. Because I think the right leaders and us, we set the expectation that. I think there's sort of two pieces there, right Pablo? One is, it is going to happen, right? If you resist it, the only thing you're hurting is now you would be less employable. Because you don't have AI. Treat AI as a skill that makes you more employable, and you have a golden opportunity to be at a company that is adopting it early versus later. That builds your skill sets on how do you work with AI in a meaningful way. That actually makes you very employable. If you don't do that and you flip on the other side and you resist it. Well that certainly hurts your chances of employability, right? But I'm not saying it's easy. The other piece is this industry is overwhelmed. I think it's not about AI taking on too many jobs right away. I think right now it's getting work done. Because it's also been in sort of a freight recession for three years. People are doing way more with less number of people. Everybody's just very overwhelmed with the amount of work on their plates. So any help on tedious, repetitive, mundane work that nobody likes to do is very welcome, right? Nobody likes to chase documents. Hey, can you send me your delivery document? Oh, can you send me the invoice? Oh, your invoice is wrong. Oh, it's not signed right. It doesn't match the load number. Hey, you forgot this one. Like who loves that, right? Nobody in the company loves that, but you have to do it to get paid, to pay people. Oh, AI is perfect at that. Let's put Augie to work, it can get you all the documents, it can check the documents, it can check the invoices, it can get your billing cycle sped up. So you get cash fast, you can pay cash fast.

Pablo Srugo (00:41:49) :
Does it do accounts receivable? All that stuff as well?

Harish Abbott (00:41:52) :
Right now, it's chasing a lot of documents. It's not doing entries into the books or whatnot. But it is like, hey, a shipment got delivered, we need to collect documents, we need to collect invoices. Sometimes there's accessorial charges, sometimes there's lumpers. It can read the documents, it can see if the documents are the correct documents. If we need to collect additional documents. It is going back and forth on emails with companies all day long. And when it knows it has everything, it can forward or upload it to the system. So then the financial systems can take on and start to invoice, for instance.

Pablo Srugo (00:42:24) :
And just to get hyper-specific, do you remember anything in this design partner phase? You're talking about, you understand the top, but how do you understand all the subtleties? Do you remember anything in particular that you learned through that? That then you implemented and maybe made a big change or whatever it was. But something important that you learned through that design phase of really being there and living?

Harish Abbott (00:42:43) :
Yeah, so I think what we did is honestly the first fifty, sixty days we shadowed operators in the business. Literally sit behind people's desk in an almost creepy way and just observe what they do. One thing in this industry is that trucks run 24-7 all the time. And so people work nine to five, sometimes they have offshore teams, but things can happen on the road. Trucks can break down, appointments can delay, labor shortages happen, so when they arrive at a warehouse they might have to wait. And for all of those you need to communicate to everybody else who's impacted. account for the fact that trucks run all the time and bad things can happen, but people work in their shifts nine to five and then some offshore teams that take over, they create these email groups or email listservs. So customer X has a listserv and people are part of that and the offshore team is part of that. So every email about that customer is going to everybody. And that's done because they always want coverage And this is a safer way to say everybody would get coverage because of that. Well, good idea. But the flip side is that for a single truck, there are at least 30 to 40 emails that are traded. So now you can imagine if somebody's running 20 trucks, is responsible for it. Their inbox has 600 to 800 emails a day. More than half of those emails are just for either informing, or they don't even need to read, or have been already actioned on by somebody else. But there's no other way to tease that out. You kind of have to read everything in your inbox to say, should I action on it If they don't know if somebody has action on it, they also write in the email listserv, hey, did somebody take an action on it? That's another set of emails. So just the amount of emails that were going around in this industry is still very, very high. And the noise to signal ratio is also very high. Like it's just too much noise and less signal, right? So we're building and thinking through a lot of that stuff and saying, okay, how can Augie help with that? The core problem is that somebody needs to be able to attend to a truck at any point of time because things can go wrong. And the core problem is that people work nine to five. But the current solution, it may not be the right solution. There could be better solutions. So this is one area. It's very obvious inside, but I don't know if you, like 50 sessions and you see people's inboxes and really go into those inboxes and read every email, you get conviction behind it. One of my favorite parts is when I go to a company, even today when we go sell or talk to the economic buyer, the CEOs or whatnot, we ask for a couple of hours with their operators. And then we learn, sometimes there's a disconnect between what the operators are doing And if you build software for what the CEOs say they want, it won't get adopted. It just won't get adopted. You might get a contract, but you won't get adopted. So next year, well, you won't get the renewal. I find it extremely important that even when you're having a good welcome from the leadership to spend time with the people who are actually going to use your product.

Pablo Srugo (00:46:01) :
Have you launched publicly now? You just raised an $85 million Series A last month, so I assume things are moving pretty fast.

Harish Abbott (00:46:07) :
Yeah, we've launched. We are now live with over $35 billion of freight under management, companies collectively combined. Augie's doing amazing things in the wild. There's so much to build. We are still very early innings of this. Literally, maybe day one or day zero is what I call. But yeah, it's receiving good, positive feedback.

Pablo Srugo (00:46:28) :
How fast did you hit a million ARR?

Harish Abbott (00:46:30) :
Pretty fast.

Pablo Srugo (00:46:31) :
Like a few months? 

Harish Abbott (00:46:31) :
Yeah. Very very exciting time. You're delivering meaningful value, meaningful ROI. So customers are more than happy to adopt and pay you for it.

Pablo Srugo (00:46:43) :
Yes, I mean, that's one thing that's changed so much since before and after this Gen AI. Before, the hard thing was finding crazy value to deliver. Now, there's so much value to deliver. The crazy thing is staying ahead of everybody else that's trying to deliver similar value.

Harish Abbott (00:46:57) :
It's very competitive, too. The space is extremely competitive. I think for the buyer, it's a noisy market because it's very hard for them to tell the difference between a demo, because it's very easy to create a demo today. You can literally go and level-ball and create a demo in a day. or two days, and it looks pretty good, but then a production-ready product still takes a lot of time and a lot of iterations, and it's very hard for a buyer to tell the difference. So it's an interesting time, unlike anything I've seen so far.

Pablo Srugo (00:47:25) :
Let's stop it there, and I'll ask the three questions we always end on. The first one, choose whichever company you want answers for, but when did you feel personally like you'd found true product-market fit?

Harish Abbott (00:47:35) :
For Augment, I think when we had, let's say, Augie deliver an end-to-end workflow for maybe like 10 or 15 users, and the user's sort of reaction was quite positive, like these are end operators, their reaction was quite positive. I mean, we had a lot of signups before, like people wanting to work with us, but I think the conviction started to build when Augie was doing work, you know, real work,

Pablo Srugo (00:48:05) :
Deliverrended up being a massive success. Was there a time where you actually thought maybe it wouldn't work out?

Harish Abbott (00:48:10) :
Man, entrepreneurial journeys are like that all the time. Like there are days where like, oh my gosh, like, you know, will we survive tomorrow or will we survive next week or next quarter? And, you know, entrepreneurial journeys are all about these peaks of optimism and, you know, valleys of despair. So I don't see there's any like all of my friends go through the same. So I'm sure there are people and entrepreneurs who have it just going up, but I am not one of them. I think big part of these journeys are just things are never as good as they seem and never as bad as they seem. and just keeping that mind that the only thing that matters is customer delight. And that's in your control and there are things that are not in your control and you obsess about it and you go deliver for it. And that just cuts the noise out too. A lot part of entrepreneurial journey in my view is also like it's a mental journey. There's so many ups and downs and a lot going on. I actually don't think people talk about it as much that they should. We try to glamorize the big exits or really celebrity entrepreneurs and people who've done amazing things, which is great, but how do you execute through these very high highs and low lows and maintain a mental sanity all the way through? And I find, for me personally, where I get that is just focus on the customer. That's in your control. Almost anything else is not in your control. And so don't obsess about it.

Pablo Srugo (00:49:42) :
It's hard, especially, I think, like having gone through when I was young, like especially when you're young and you get like, for example, some crazy fundraise or some big validation, whatever it is, maybe it is a customer, but it's just one customer and you just kind of start to over index on how great it's gonna be when you're the next Mark Zuckerberg. You know what I mean? And then tomorrow some huge failure happens and it's tough to balance that unless like you said you have some grounding focal point that keeps you level-headed.

Harish Abbott (00:50:07) :
Yeah, and you kind of have to have a few my daughter's sort of diving coach. She's actually a two-time Olympian. You know, speak to her and her friends about, you know, if you think our journeys are tough, like these Olympians' journeys are maybe 10x tougher, right? They have to go perform. Sometimes they hit, sometimes they don't. I mean, sports are so competitive. Just to be at an Olympian is like top 0.1% or 0.001%. But I think it was very cool. Like, I learned a lot. It's like, hey, you know, you're like, If you have a little boat in a big ocean wave and you only have one floaty and you go up and down, but if you got three or four things you can hang yourself, whether it's faith, whether it's your family, whether it's your relationship, whether it's a sport, whether it's a hobby, then one is up, the other is down, it balances you out. That was her method and it taught me a lot. That's very important for us to have too.

Pablo Srugo (00:50:58) :
Well then let me ask just the last question, what would be some of your top piece of advice for an early stage founder trying to find product market fit?

Harish Abbott (00:51:04) :
Just obsess about the end user and get in their shoes, not the economic buyer. The economic buyer will follow when the end user loves you, But just any time you can spend with the end operator using your product, using their existing status quo, learn the nuances, the little things, and just be obsessed about that problem. And I think if you're obsessed about it and you build for it, I think good things happen. Obsessed about that, not about revenues, because if you're obsessed about that, revenue follows. But if you're obsessed about revenues, then I think bad things can happen. So just keep that the main thing is probably,

Pablo Srugo (00:51:43) :
Love it. Harish, thanks so much for jumping on the show, man. It's been awesome having you.

Harish Abbott (00:51:47) :
Yeah, thank you for having me.

Pablo Srugo (00:51:48) :
Wow, what an episode. You're probably in awe. You're in absolute shock. You're like, that helped me so much. So guess what? Now it's your turn to help someone else. Share the episode in the WhatsApp group you have with founders. Share it on that Slack channel. Send it to your founder friends and help them out. Trust me, they will love you for it.