Sacra Logo Sign In

Andrew Yates, CEO of Promoted.ai, on driving marketplace ARPU with personalization

Jan-Erik Asplund
None

Background

Andrew Yates is the CEO and co-founder of Promoted.ai. We talked with Andrew because Promoted.ai sits at the middle of two major trends: (1) the rise of vertical marketplaces serving unique supply and demand, and (2) the decline in the performance of social media marketing.

Questions

  1. What’s your vision for marketplaces in general, and why did you choose to bet on them?
  2. The team has a lot of experience working on commercial media and ad networks at different companies. I'd love to hear a little more about that experience and how you translate that into running marketplaces.
  3. Let’s dig in on the landscape of paid marketing in general and how you’ve seen that shift, particularly in light of the iOS privacy update.
  4. I’d love to hear your thoughts on different kinds of marketplaces. Are there differences in how well they can be optimized with data science? For instance, an ultrafast startup like JOKR that owns its own supply versus one that connects supply.
  5. When you talk about different stakeholders involved in the process who might care about what is being shown, does that lead into this question about marketplaces launching ad platforms? How do you -- and how should founders, in general -- think about that as an option for marketplaces?
  6. Across marketplaces that you see, do you have a sense of what the breakdown looks like between what revenue comes from just transactions and what comes from advertising? In a typical marketplace and also a mature one like Uber Eats?
  7. Could you speak generally about what it looks like when everyone agrees that it is value add? What conditions create that?
  8. Going back to JOKR and other grocery services, do you think Promoted and similar tools can help optimize on the supply side, in terms of understanding what types of inventory you should be purchasing? In particular, understanding better what folks are buying, what drives conversion, what drives bigger AOV.
  9. What is Promoted’s take on super-apps? It’s important to refer customers across different verticals and different types of products to get them to buy more valuable stuff more often, and it seems like Promoted can help drive that kind of behavior.
  10. Could you describe Promoted as an ad network?
  11. Why doesn’t Facebook or Google do this?
  12. You mentioned recommendation tools. Do you think that the best engineering teams will build this recommendation system in-house? Does that factor into your thinking?
  13. I'd love to hear about your philosophy on hiring for Promoted, especially given that salaries in this space are so high and you've talked about giving out a lot of more equity than normal for early engineering hires.
  14. Is there anything else that we haven’t talked about that you think is important to mention?

Interview

What’s your vision for marketplaces in general, and why did you choose to bet on them?

We're going after where the value is, and we see the value in marketplaces. Our specialty is commercial media. You know commercial media in terms of Google, Facebook, Amazon -- it's extremely valuable to show people the right thing at the right time. That's what we've all built our careers around. Marketplaces are where this is both very valuable and very difficult. Since we're selling a technology solution, the difficulty needs to be there as well. Marketplaces are two-sided. They have critical, frequent inventory challenges, whereas for a general ecommerce site, maybe there are only ten items for sale. It's not as complicated, although it still benefits, and general ecommerce is something that we want to expand into in the future. But we believe that we can solve the most challenging and the most valuable problems first, and then as we expand, we can gradually go after marginally less valuable types of problems using the technology that we've already built. Then you’re building a subset of that.

The effect of this from a strategic standpoint is that it allows us to build really fantastic solutions. Then as we expand out to other types of ecommerce or down market, we have an order of magnitude better solution. We can also sell to the best companies and start our network in that way. That's another part of it: for marketplaces, it's very much a winner-take-all network effect type of business, even though you may have different winners in different verticals. One of the things that we've seen is that, to make this business work, you need to be able to sell into the winners and not just into the long tail because it's just not evenly balanced.

The team has a lot of experience working on commercial media and ad networks at different companies. I'd love to hear a little more about that experience and how you translate that into running marketplaces.

This is a business that I've always wanted to start. I came to San Francisco after graduate school. Before I was studying bioinformatics, doing cancer research, and I made a decision that, no, I wasn't going to continue down that path. Instead I was going to join industry. Early on, it was obvious that the business of the Valley is what we call ads -- this general sense of commercial media, so Facebook, Google. I knew right away that was the industry I needed to master to be successful in technology.

The other thing that was quickly apparent to me was that trying whatever was on top of mind wasn't going to work in this industry. This is something I think people learned the hard way maybe five, eight years ago. Without deep experience about why top companies are the most successful, you're competing against Amazon, Facebook and Google, and that ended poorly for everyone who did so, in addition to people who were trying to build tools on top of what Google or Facebook were doing.

So I wanted to acquire deep understanding of how the winners worked and then go start a company. That's precisely what I did. Originally, I joined a startup when I first came to San Francisco. I saw very quickly that Facebook was printing money. This was at the time mobile app install ads were super hot. I joined Facebook and learned how they did their machine learning, ads auction, and that part. And then I joined Pinterest and ran their marketplace engineering team.

The trigger point for me was a head of BD at Pinterest who said, "Hey, Pinterest is trying to acquire some startups in this space to accelerate our ads business. There are no more good startups in this space that are interesting to a level of Pinterest.” That was really interesting to me. Before, it was like everyone was doing an ad tech startup if you didn't know what else to work on but wanted to be reasonably successful, but the whole world had changed. At the same time, we were responsible for hiring at Pinterest, and the competition was insane for discovery engineers, search engineers, ads engineers -- especially ads engineers. It wasn't just the usual suspects -- the usual suspects being FAANG -- but also, it was marketplaces. At that time, it was Coupang, Grab, DoorDash, Airbnb. It was starting to be a million dollars total annual compensation for a staff engineer, and up and up and up.

I saw this as an opportunity. Not only is this going to be incredibly difficult and expensive to build, and everyone is going to go try to go build it, competing for the same talent, but also, having done this a couple of times in industry, your v0 is not going to be good and neither is your v1. It may never be good. I think that's something that they say some other social media networks are struggling with, where they're competing against Google, Facebook, or Amazon, especially for the advertiser side but also the creator’s tools side, and they're just never able to pass that bar. They're always much worse, because they did some expedient things in the beginning and were never able to advance past that. So we saw this opportunity of, “Let's do it right to start with, and then sell to all of these companies as they're growing.”

Then, of course, COVID took off, and this trend became accelerated. But we had seen this opportunity back three or four years ago: just the demand for search discovery commercial media optimization, which we call ads -- this general idea of revenue optimization or “How do you make people successful when your only interface with your customer is a mobile screen? How do you make that successful in ecommerce retail and all kinds of marketplace businesses?”

Let’s dig in on the landscape of paid marketing in general and how you’ve seen that shift, particularly in light of the iOS privacy update.

One interesting thing about ads is that if you say the word “ad,” people stop thinking -- they shut down. People really take it for granted these days that this is how you run ads: you have a tracker pixel and you spend on Facebook, you spend on Google.

In general, one interesting thing about the iOS privacy update is that nothing really changed. It's just the same but worse, in terms of high-level strategy. Again, I think that is compounding and realizing this market distortion where you have more and more marketplaces -- new apps that are desperate for performance growth -- and, at the same time, the existing channels are becoming more and more inefficient and there are fewer and fewer of them. The bigger ones just continue getting bigger, and the smaller ones continue not to really take off. So people are running out of ideas.

You've seen some of this pressure being relieved in terms of, “Why don't we just give away free money?” You saw that with Coinbase. It's like, “Just download our app and we give you free money.” We’re getting to the extreme where, when performance media is that expensive, why don't we just pay users to just install our app? There are a variety of ways to express this, like coupons or discounts.

With the iOS update, it made it more difficult to follow the playbook of traditional performance media and growth strategies. But it didn't really change anyone's playbook that much. It just made it even more expensive and more painful.

I’d love to hear your thoughts on different kinds of marketplaces. Are there differences in how well they can be optimized with data science? For instance, an ultrafast startup like JOKR that owns its own supply versus one that connects supply.

If there's nothing to decide, then there's nothing for machine learning to do. If you only have three items and they're a fixed price, it doesn't matter how much machine learning you have or how much measurement you have. There's just nothing to be done.

At the same time, even in the case where you have a relatively smaller fixed inventory, as you become more and more complicated, it requires much more machine learning and marketplace matching. One famous example is Uber, where they only had one product -- well, I guess they had a few different variations of the same product, but it was one product -- which is, you press a button and then someone shows up and picks you up. It wasn't like a catalog of different items. Nevertheless, they had a tremendous investment in data science and machine learning. It's not just “the item” -- it's “this item, delivered to you right now, at a price that makes sense for you with long-term economic effects.”

Plus, there are the other stakeholders as well. JOKR may own their inventory, but they don’t manufacture the products. It’s not like they have their own snack factory. The consumer packaged good companies care a lot about which goods are being shown. So now you have stakeholders. Plus, you also have other dynamics. Even within a fixed inventory, what's available? What are you trying to push? What can you deliver? What are the efficiencies around that?

As you become more and more sophisticated, you realize it's not as simple as, “Okay, there are twelve SKUs, and that's it.” The product is the fulfillment of that product plus the long-term relationship. That's where data science can become really important and critical, in fact, especially when you're running at close margins. At volume, it can be the difference between a profitable unit and an unprofitable unit, and that makes a huge difference in how your business is modeled.

When you talk about different stakeholders involved in the process who might care about what is being shown, does that lead into this question about marketplaces launching ad platforms? How do you -- and how should founders, in general -- think about that as an option for marketplaces?

I touched on this idea before that people shut down thinking when you say the word “ad”: “Banner or ad, I hate it. People hate it, but it makes money. I don’t want to think about it.” We have a much more expansive thinking on this, which is that it's just revenue optimization, margin optimization, overall marketplace optimization. How do you match people and come up with a value for it?

Effectively, what ads are trying to optimize the margin or take rate in the marketplace. As you get more volume, then you can demand a higher take. For newer merchants or for diversity or other user experience considerations, you can have a lower take or even negative take in the sense that, if you're doing promotions, you're trying to subsidize growth. That's how we model it. Our insight was to deliver everything like an ad, in the sense of thinking about all the sophistication around spending on Facebook or Google around measurement, incrementality, A/B testing -- you've got creatives, your rotation, multi-touch models. In contrast, look at how the commercial media is run on, let's say, your own marketplace. None of that. Maybe you have a search quality score that's heuristic, like “Okay, this is good enough.” No measurement, no concept of incrementality, no concept of margin optimization, and it's just done ad hoc.

Our belief, our philosophy and our big insight was we took the technology that has been developed and refined at the world's greatest and most successful companies in commercial media -- in particular, Google and Facebook -- and then we're applying it to a new domain of marketplace optimization in general. We can do subsets for just running your search and feeder promotions; we can remove the ad auction, so it's not paid anymore; or for promotions, it's, “'Where does the budget come from, to maybe have an internal currency?” Do you just do ads or organic, in terms of your ordinary search ranking, or do you do some kind of promotion or recommendations thing? In our minds, this is all the same technology. It's just different configurations around what your objectives are as a marketplace in terms of maximizing long-term growth, maximizing margins in a fair way, and sustainably buying growth in a way that you can reverse later. That's how we think of ads -- as just one facet of this overall question of, “How do you manage an economy where the interface is an app?”

Across marketplaces that you see, do you have a sense of what the breakdown looks like between what revenue comes from just transactions and what comes from advertising? In a typical marketplace and also a mature one like Uber Eats?

This gets into the idea of: do you think of ads as an entirely separate line of business, or do you consider ads more as another tool in margin optimization or fair ways of growing your marketplace? If you're going to take higher and higher take rates from your vendors, to do it sustainably, you need to provide some value in exchange. That's fundamentally what ads are. What we've seen is that you don't really start trying to pull out more and more margins from your top vendors until you have top vendors and they have a reason to stay.

Otherwise, first of all, you're not going to get much volume from it. Second of all, they’ll balk at it, and you need them more than they need you. This dynamic means that, at the low level, you definitely should not be running ads. I'm just coming out straight to say, “Look, if you're just trying to get started, unless you just want some low-burn sustainable boutique app kind of thing, you're probably most interested in increasing total growth of both supply and demand, and your best lever is going to be giving out some discounts.” But as you grow -- and I think Amazon is the pinnacle of this now -- ads business becomes all of your profit. You go from “you shouldn't be doing it at all, it's a red flag” to “it's a tiny fraction, but it's an interruption” to “it's all that matters.” That happens really, really quickly. And the technology investments to be able to go from “this is just a quick hack someone can just throw in over a day, and we probably shouldn't have done it anyway” to “oh my gosh, this drives 100% of our profits and everything else is just sustaining the media business” -- like on Amazon -- is very, very challenging.

Back to your question more directly -- for small marketplaces, ads are either zero or a tiny fraction or a much larger fraction than they probably should be. For marketplaces as they grow -- let's say Uber Eats -- as they become scaled and now it's about revenue optimization, not just growth at a loss, then it becomes almost their entire business. That's where the profit comes from, because it's the revenue optimization side. And it's critical that it works well with the rest of the product. But again, it’s very challenging to build in a way that everyone agrees is a good part of the product, as opposed to just feeling like value extraction, which I think is a symptom of how this technology is done poorly.

Could you speak generally about what it looks like when everyone agrees that it is value add? What conditions create that?

Well, it looks like a trillion-dollar company. Look at the difference between Facebook and really everyone who's trying to copy Facebook's model -- let's say it’s orders of magnitude and more of value. Even though when was the last time you open Facebook News Feed? But Facebook is still worth a trillion dollars. The difference looks like a positive feedback loop versus always having to keep throwing more fuel into the fire to keep things running.

At Facebook, I worked on the systems in the very backend in terms of maximizing value where it just became a data problem, where you just keep doing a better and better job of predicting what people are most likely to be interested in and delivering that effectively. It's like 1%, 1%, 1%, 1%. Every time you did that, the overall experience plus the revenue that you could extract both increased. That's what it looks like when it's a positive feedback loop.

When it's not that, you get the traditional effect where every promotion and every ad you show is cannibalizing your user engagement and eventually your daily active users or monthly active users. You start losing users. I don't mean to call out any particular social networks, but has this happened? Maybe. That's the typical pattern. Groupon is another example where this went awry, though a little bit different dynamics -- this is more of the discount type of promotions as opposed to paid promotions. It’s another example of an unbalanced marketplace where it wasn't sustainable.

You can see it in both directions. Doing it poorly means that you don't get that exponential growth anymore, and trying to extract more and more revenue means that it comes from your future revenue. Doing it well means that not only does your revenue go up, but your growth goes up as well in the future.

Going back to JOKR and other grocery services, do you think Promoted and similar tools can help optimize on the supply side, in terms of understanding what types of inventory you should be purchasing? In particular, understanding better what folks are buying, what drives conversion, what drives bigger AOV.

From our technology perspective, that's not something we're focused on. It's something that is revealed in that, if you have a limited supply of a specific item and you can project that into the future, you have three options, and we provide two of them.

One is you can increase the price to lower demand. Two is you can lower the availability. You can keep the price the same, but just not make it as prominent: you basically push it to the back of your store, so to speak, but in app, it would just be less prominently featured. Because fewer people see it, you're going to get lower demand just because people aren't reminded about it as much. Then the third piece is you increase your supply, so you need to project what supply you'll need in the future and order for it.

At least from our product’s perspective, we would help you with the first two. The fact that you're doing those first two would be highly informative for ordering more supply based on projections -- well, perhaps it would be more efficient if you simply had more supply to meet projected demand, and then you wouldn’t have to do either of those first two things to artificially suppress demand.

What is Promoted’s take on super-apps? It’s important to refer customers across different verticals and different types of products to get them to buy more valuable stuff more often, and it seems like Promoted can help drive that kind of behavior.

Well, that is our target use case -- to help optimize those super-apps. The idea of a super-app is you have one screen, you control that screen. Well, okay, how do you control that screen? It becomes very, very complicated very quickly. It's almost like, not only do you have the traditional third-party sellers and your users and optimizing conversion rates, but you also have competing product teams, competing regions, different regional managers. How do you make this work in a sustainable transparent way that can be managed at all?

And then, of course, the user needs to have a good experience. It has to not only work, but also not look like a hodgepodge of a bunch of committee decisions. I think people have seen that -- where maybe technically, everyone got in on the app, but it's not the great user experience anymore. You're not optimizing for the entire package anymore.

First, our perspective is we specialize in optimizing the super-apps where you have an internal currency for all of your product teams and merchandising teams. Within those merchandising teams, you have how you decide what to show -- for example, products or cross-promotions. They only have one screen, and you only have so much attention. How do you make everything optimized within that in a way that you can report, control, and optimize?

The second piece of this in our grander vision here is, why be limited to one app? Why not do it across apps? Why does everything have to be in a super-app? I think the existence of a super-app -- as opposed to multiple thin apps that are very efficient at what they do -- is a symptom of that it's really, really difficult to do this type of cross-promotion or to get apps to work effectively with each other, even though it may be a better user experience to have just one app, one function.

Our vision is to build distributed AdWords and take the technology that we use for optimizing within a single app and then start allowing things to be shown on searches on other apps. That’s also where you get a lot of interesting dynamics in terms of supply and demand of attention. Right now, we're focused on marketplaces where it's super, super critical to get all of the things for sale optimized, and there's a tremendous amount of demand for attention. They don't have a surplus of supply of attention. They just have a need for more and more attention. But then you have other types of media networks like, say, social media, Reddit, Twitter, and TikTok, where they have tons of attention, just hours of your time. They're looking for buyers for that attention, but it needs to be effective.

Our vision is to match the demand with supply of attention and do it in a way that people love. Unlike every other ad network, we start with the very best, most actionable media that people already love to see and want to buy right now, which is on marketplaces like Snackpass, Hipcamp, Gopuff, Outschool, these listings. Then, we find placements.

When people see an advertisement, they don't want to be interrupted. But there's a threshold of, “Well, hey, this was actually something that I want, and I could get it right now, and that was OK, I’m OK with being interrupted about that.., That’s not the experience people get today because Promoted doesn’t exist today. When I'm on Twitter or Reddit, I don't want to see an ad “about” Starbucks -- like, the Starbucks brand exists, but you don’t need to remind me about Starbucks or to download the app. I don’t care. I want to see, “Push this button and a Gopuff guy shows up at my door in 10 minutes with a hot coffee.” Maybe that guy even helps me install the app and collect payment info at my door to get a free snack. Now that’s performance. That's the sort of cross-app success that you can see within these super-apps. We want to take that across all apps.

Also, thinking in terms of going after growth budgets that are all being dumped onto Facebook or Google for lack of better options -- wouldn't it be so much better if those sorts of marketing budgets were going to vendors within your super-app to help them get more direct exposure that is directly attributed to succeeding within your app for whatever that conversion means for you? Versus just general brand awareness or “Here's free money, download my app and then once you’re in the app, maybe you’ll figure it out.”

Could you describe Promoted as an ad network?

That's what we want to build. How do you build an ad network when you have no media and no demand? How to get started? Now, there's a previous strategy, which is, “Well, we'll go convince some bunch of brand advertisers, and then we'll go partner with a bunch of publishers and try to match it.” But that's been done a lot, and it's always failed, except there are already top companies to do that, you know, Google and Facebook. Are you going to directly compete with them? That’s' insane. It doesn't make any sense.

Our strategy is instead -- going back to what I was saying about commercial media that people actually want to see -- I really, really want to see the commercial media that's successful on super-apps. I want a coffee. how me a beautiful vacation holiday that I can book right now.. it's a high-quality experience. Start there, and then start expanding from there. Start with the things that people really, really want and are desperate to find, but can't. Make it work within that app. Get the best possible marketplaces in the first place, and then build your network from there by networking together these marketplaces.

Then this becomes your book of media, so to speak, for going to other parts of a network where it would be publishers or social media apps that are just desperate for performance media that people are interested in seeing. They don't have a problem with getting brand ads from big advertisers and getting experimental budgets. Their problem is, well, how are you going to scale showing a coffee from down the street as an ad where you only spend like maybe $5 per ad, but you have millions of them? They have to be automated. But Promoted can do this because we already are doing it within each app. We just have to do the cross-promotions part.

That's our vision: start with all of the demand for the highest quality commercial attention, and then expand from there. We’re still in the collecting the highest quality demand phase of the ad network vision.

Why doesn’t Facebook or Google do this?

Well, one answer is they do do it, within their own network. When you check that box to take your News Feed ad and go onto Instagram or Audience Network or Messenger or whatever, Facebook makes it look so easy. No. It's really hard. Facebook already has of, here's just a thing, and Facebook will figure out where it goes for you. Google, of course, has AdSense network and their own ad networks.

But I think the biggest difference between what our strategy and what Facebook, Google, and Amazon are doing is that these three are centralized. For Google, it's search: everything comes through search. For Facebook, it's user identity: everything is tied together by user identity. For Amazon, it's fulfillment. You go onto Amazon’s app, and then they own the fulfillment. That's why I use Amazon, because I know it will show up at my door. Other vendors? No.

For us, we are not centralized. We don't have some centralized identity. We don't have a centralized app. It's entirely decentralized. The advantage here is that apps can specialize in their own domain, their own vertical. If Facebook or Google or Amazon were directly to compete with us, then that would mean giving up the thing that makes them the most successful and the keystone of their entire strategy. They technically could do it. But culturally or economically, it wouldn't make any sense.

But in terms of technology offerings, TikTok provides their algorithm for a service. Amazon has their AWS Personalize, which they say is what they're using on their backend. I'm sure it's some version of it. They already are providing some of the basic tools. But would they build a competing media network that, one, directly competes with their successful model and, two, doesn't leverage any of their strategic assets in terms of that centralization? Nothing prevents them from doing that, but it doesn't make much sense for them to do that.

You mentioned recommendation tools. Do you think that the best engineering teams will build this recommendation system in-house? Does that factor into your thinking?

Absolutely. The only competitor we care about is “build in-house.” Of course, we would care quite a bit about Google or Facebook making it a very direct competition, but we'd actually be pretty excited about it, because it would help some of our go-to-market. And given what I just said, well, I don't really think that's going to really be an important part of their strategy.

The funny thing about the way you described that is it takes for granted that you have a top engineering team that is going to build it, and it works. That's how people describe it. It's like, “Oh, well, won't everyone just build it themselves?” Let's look into that a little bit. This goes back a little bit to some of our origin story in hiring engineers and building these teams at Facebook, Google and Pinterest. And everyone on the team, by the way, has experience with building this at other similar companies like Grab, as an example.

First, you have to hire these people. How much does that cost? Well, have you tried to hire senior staff engineers in machine learning and real-time infrastructure lately? It's going to be at minimum $500k or millions of dollars per year per engineer. And that's just to get them with a company laptop, and everyone's remote -- so not even in the door. They don't even show up at your office. You're like, “Okay, now you’re on payroll,” and you send them a laptop. Now they have to build the thing. Okay, so how many of these do you need? Well, look at the industry. Is it one? Is it five? Is it twelve? Is it 50? Is it 100? Is it 1000? Is it 10,000? The number of engineers keeps growing, and there are companies employing this number of people on every point on this curve, and it keeps going. When your company is really serious, the number of engineers keeps growing, at least super linearly.

Then you think about what are these engineers working on? Well, let's say you're just getting started. Okay, we're going to hire a top team. We're going to hire a VP out of Facebook. They're going to put out a job listing. We're going to say, “Look, we just raised at a $2 billion valuation. We're super hot. We'll pay you a $200k salary, and our stock options are going to be super awesome, and you can make up to a million dollars a year.” First of all, you're going to be competing against Google and Facebook who will pay you in cash. Where's the growth opportunity for you worth that much money? Second of all, after hiring all of the people and getting them to work together, how much time do they have to go build a really fantastic system? Have they worked with each other before? No. Now you have one quarter with a bunch of new people who don't know each other, and you’ve got to slap it together.

We've seen it over and over again. You cut all these corners and you get something out the door. And voila! You're showing a banner ad at the top of your app, and you're working with some big brands. Look, congratulations! But let's look at Coupang. Coupang has several hundred engineers on ads -- just ads -- and several hundred engineers on just search and discovery, and they have all the supporting infrastructure, data, client logging, and then, of course, all the connections back to fulfillment house. It gets super complicated. It's like, “Well, is it really a reasonable prototype that you showed a banner at the top of search, or you pinned the top result and you have a simple crud app about now you can control who's at top?” I mean, I guess technically that's on the same path -- in the sense that if I jump up, that's closer to the moon, depending on where the moon is in the sky.

That's what we've seen with “build in-house,” and we've seen it over and over again. If you've already gone to SoftBank and raised a billion dollars to go build this, and you're dead set on being a technology leader or the next Uber, that’s awesome. That's probably not a customer that is going to be great for us today, although we have a lot of specialty tools that maybe you should look at and that maybe you haven't already built. But for companies like JOKR, how much time do you have, and how many businesses can you be the best in in that amount of time? “You're going to be the world leading media technology business and the real-time delivery business, and you need to do it in six months.” That doesn't sound like a successful strategy.

But I don't think people have really thought through this carefully, because, as you described it, “We'll just build in-house. How hard can it be?” It's deceptive, because it's not that hard to hire a few engineers. It's not that hard to convince yourself that, “Hey, look, someone worked at Facebook before. We're going to rebuild Facebook to get started.” But then two or three years in, your marketplace isn't growing in any sustainable way, you have this huge mess, and you just keep doubling the size of the engineering team. It's not really progressing in any sort of way.

I think people will start realizing that they should focus on being an expert in one business versus two, in terms of where you're going to invest your top talent and resources. Is it going to be logistics? Is it going to be fulfillment? Or, is it going to be the commercial media optimization in your app? Which of these do you want to specialize in? If it's all of them, why wouldn't you just be outcompeted by someone who focuses on one and outsources the other pieces?

I'd love to hear about your philosophy on hiring for Promoted, especially given that salaries in this space are so high and you've talked about giving out a lot of more equity than normal for early engineering hires.

Part of the way the company is structured requires that everybody be highly experienced, because everybody is remote. Getting your head around all the different concepts is extremely expensive, if not impossible, and that's compounded by people being remote. Part of the philosophy there is it's required that we hire in this way to operate the way that we do.

Then, in terms of how we think about hiring engineers, this dynamic of “build in-house” that I just described is no fun for anybody. It sounds like it's no fun for management and leadership. It's also not fun for really talented engineers. No one wants to be told that they're building the world's greatest super-app experience, but actually they build the worst piece of crap that can get out the door, and then they have to fix that worst piece of crap just to keep the lights on over time. That's not a fun experience. The more talented and more experienced you are -- especially if you've come from companies that do it right, including Facebook, Google, or Amazon -- the more you know what good looks like and what trying to copy good in an expedient sense looks like and what it feels like to work in it. It feels bad.

Our pitch is: first, come build what you know is right with us, and you own it. You know how valuable it is to do it right, because you've seen it done right. If someone's going to pay you a million dollars a year to sit around and refresh hive jobs or whatever it is, like do a bunch of A/B experiments on twaddling some parameter, that's worth a hell of a lot more to them than a million dollars per year. The vision is, “Look, from an experience perspective, build the thing that you know is right with people who also know that it's right. Let's do it the right way, and have the company structured in such a way that we benefit from that.” Because we own that system and then we're reselling it out to other people, the quality matters tremendously, versus just the internal dynamics of, “Hey, you’ve got a product manager and they need to get it out there this quarter, and then it’s someone else's problem in the next quarter.”

The other aspect is the value of it. It makes sense when the expected value, if successful -- even if it's maybe a challenge -- is billions or trillions of dollars, and that is the market size for performance media. We talked a little bit about some of the dynamics here of marketplaces just craving performance growth that scales, and there are just no new ideas. Or the existing ideas are getting worse and worse, and there are still no new ideas, so there's more and more pressure. People who are building these systems are smart people, and they recognize that these market pressures exist. There’s a real possibility that this large equity stake that we give to our engineers -- almost like a partnership -- will be worth it. Not only in terms of being proud of the system you've built, but monetarily as well.

Is there anything else that we haven’t talked about that you think is important to mention?

I want to push back more on the ad idea. This is something we've seen over and over again, especially when the venture market was not super hot, like right when COVID started. You say the word “ad,” and people are just like, “Ugh, I'll never do that. No, never.” Or they would even be doing it internally, and they would lie about it because they're embarrassed about it or something like that. Of course, they're building their ads team and hiring backend ads engineers in the meanwhile.

It's really created this fascinating blind spot, culturally, for everybody. On one side, people are super excited about crypto, and it's interesting, but then you look at -- well, how profitable are companies like Facebook, Google, Amazon, and Microsoft for that matter? It's just commercial media. It's deciding what people see at the right time. It's almost like people hear the word “ad,” and they stop thinking about what have all of these engineers been doing at these companies for fifteen, twenty years? There's a lot of really fascinating technology, economic techniques, philosophy around how these world-dominating companies work that people just don't want to think about. It’s just a blind spot where you say “ad,” and people immediately think, “Banner ad -- I hate ads. That's lame. That sounds like a cheap, easy way to make money. I'm not interested in it. I'm interested in other things.”

For us, because we've always been on the winning side of all of these ad network battles over the past -- we've always been in the in-house of Facebook and Google, where things were always growing, and it was great -- we've had the time to take a step back and think more about the industry in an economic optimization way: What's really happening here? How does the attention economy work? What are the pieces of the attention economy in terms of theory or technology? What information needs to flow to where? Why do people pay the prices that they do for attention? Why is that sustainable or not sustainable? And how does that fit in with things that are similar but that seem totally different in terms of an organization? Like, today, you'll have your search team and then you build an ads team, and these are totally different teams. Then you have a sales team that goes and does an ads business. 

What's the unifying theory behind all of this? At the end of the day, all you have is a screen. You’ve got a small screen. And you're just showing some images at the top and some text and deciding how much that's worth in dollars. No matter if it's some sort of promotion or your search feed or ads. We think of it as: build the fundamental theoretical attention economy infrastructure, and then you can configure that however you need to accomplish whatever the objectives are of marketplaces or ecommerce apps -- or maybe someday even publishers or media channels, like video apps or social media.

I just keep wanting to say, stop shutting down when you hear the word “ad,” because a lot of things are happening, and it's not all the scummy “hey, stop stealing my privacy.” I think another aspect is people get distracted by the politics of the 2016 election where, suddenly, the darlings of technology were used by the wrong people and now everyone had very different opinions about how that technology was. You say, “How do I think of attention economy from a politician’s perspective?” And all of a sudden, your mind goes to all these dark places that are not very interesting in terms of actually answering the question. It's more just a reaction to the emotional load of that or tracking identity and privacy issues. Again, these are critical matters. I don't like being tracked. I have very strong opinions about keeping private materials private. But again, people get stuck in that rabbit hole of “Ads -- oh, privacy -- oh, I hate when people steal my data. Yuck.” And they aren't thinking about why $1 trillion companies are valued at $1 trillion.

Disclaimers

This transcript is for information purposes only and does not constitute advice of any type or trade recommendation and should not form the basis of any investment decision. Sacra accepts no liability for the transcript or for any errors, omissions or inaccuracies in respect of it. The views of the experts expressed in the transcript are those of the experts and they are not endorsed by, nor do they represent the opinion of Sacra. Sacra reserves all copyright, intellectual property rights in the transcript. Any modification, copying, displaying, distributing, transmitting, publishing, licensing, creating derivative works from, or selling any transcript is strictly prohibited.

Read more from

Read more from

Read more from

Ratnesh Verma, CEO of Pidge, on on-demand delivery logistics in India

lightningbolt_icon Unlocked Report
Continue Reading
None