LEDGE: Alright. Michael, thanks for joining us. Really cool to have you on today.
MICHAEL: Happy to be here, Ledge.
LEDGE: Can you just give your two or three minute background story, so the audience can get to know you a little bit?
MICHAEL: I've got a 30-year background in high tech. A lot in the startup world. This is my 10th startup. I've done work in all parts of the industry and spent a decade as a consultant. I advised companies as broad ranged as Apple and Amazon, Adobe, Sybase, Symantec – enterprise, mobile, web. I spent the last 10 years in the data communication space and right now I'm focused on the retail and ad technology space.
So kind of a broad range of experience over a long period of time.
LEDGE: It's the best kind of experience. I imagine you're able to bring the broad perspectives to the new space.
You and I talked a little off-mike about the revolution in ad tech. I'd love if you would dive into that. I think the audience would find that really interesting.
MICHAEL: You think about what's going on, and ad tech is about to go through the largest disruption in 20 years. It really has become crystal clear to people, even as recently as last week. People know that the largest amount of advertising flows through two networks. They flow through Google and it flows through Facebook. But what people don't know is the third largest ad network in United States is Amazon.
Amazon was scheduled, last year, to do $3 billion in advertising. They didn't. They did 3.4 billion last quarter. They did 10 billion last year. Ten billion. This has shocked the entire industry. They are now the third largest.
We're in this disruption and people really haven't figured it out. The best way to understand why is the money flowing to Amazon is to ask ourselves a simple question. Why do the best components of marketing, the best internet marketing, why does that happen as far away from the time and place of purchase online? Think about it. The worst marketing is done on the retailers' websites and the best marketing is done off the retailer's website. Why is that happening? Because the ad tech world is backwards. Amazon is showing you need to bring the advertising to where the shoppers are, which is while they're shopping. There's a lot of value in doing that, and then…
LEDGE: Just for perspective, this would be like I'm shopping on Amazon and I see the sponsored listings right there in the product. I see other places where they're selling their real estate on the product page. It's hyper targeted to, "Well, if you like this one, you might like this one." That kind of stuff?
MICHAEL: Yeah. You're starting to see that. I run a company called SYNQY, S-Y-N-Q-Y. What we’re doing is we're delivering actual product ads right alongside the product when the product shows up in retail.
What this allows us to do is tell the brand story that is being told on Facebook, or being told on a television ad, or being told on a microsite that the brand owns, which no one goes to – and actually bringing that story to the consumer while they're shopping, while they're receptive for the message.
We're actually pushing beyond what Amazon is doing, but delivering a truly interactive experience in a way that hasn't been done before. There's a tactical underpinning to do that, which we can talk about in a few minutes.
I'll give you an example. One of our largest customers is Albertsons’ Safeway. It's the second-largest grocery chain in the country. They have a personalization engine called 'Just for U'. In that personalization engine, it determines what coupons you should see based upon past purchase history and so forth, but all of those coupons are static images and we allow interactive content. This could be recipes, other things that are helpful to the consumer, to be available when people are looking at coupons, as an example.
LEDGE: What's the technology look like to make something like that happen? I come, for example, from the early online days of newspapers when classifieds were first getting disrupted and it was magical to have a moving ad made with Flash.
Fast-forward to here where you're talking about taking on this new behemoth. I happen to know that Albertsons is actually one of the companies that dropped AWS, because they were so worried about the footprint of Amazon competing onto their retail turf. Talk about serving that, and what's that look like to get that done in a way that can compete with such a monster.
MICHAEL: Let's take a step back and let's talk about how the web works from a marketing point of view today.
You're a manufacturer and you sell products through other people. We call them resellers or retailers or other people like that. They have a website, and the problem is you don't manage their website and they don't manage yours. So we have this conflict, a tactical conflict of interest because I want people to come to my site and they want people to come to their site. I can't express what I want to express through code on their site very easily because they don't give me a way to manage my content on their site. Are you with me from a tactical point of view of the tactical that?
LEDGE: Yeah, absolutely.
MICHAEL: The whole web is predicated on this issue. All of the ad tech says, "I'm going to take people from where they are and bring them to domains I can control. In doing so, I will be able to communicate my value to them. But I have now them taken somewhere. I've taken them on a journey." I call that a marketing transportation tax. That we're spending money to ad companies, we're spending money to Google, to move the shopper from where they are to where the marketing is.
We said, let's think about this differently. Why don't we move the marketing to where the shopper is?
Now, what that requires us to do is some different ways to think about the problem. You're familiar, I'm sure, with idea of web content management. Think about us as remote content management. We're managing our content as it is communicated across sites we don't manage. Our ability to control that experience in a way that meets our requirements but does not harm the host site we're running on, that's the answer to what we…
LEDGE: Sure. There must be some technology there but I imagine you try to limit the footprint of what they have to do on their site.
In some simple examples, we can go on a retailer's site and if a certain product comes up we can go grab an advertisement related to the identifier of that product. We have a matchmaking technology that we know have a patent on. It's our ability to identify the content needed and deliver it in real time. Real time is post page delivery, under 20 milliseconds.
The second piece is we can take additional elements, such as the geography of the shopper or the geography of the grocer itself... We work all over the country, but they have certain products that get marketed differently in different regions, or different offers based upon the geography of the store locator or where the consumer is.
Let's say you have a coupon that has restrictions in the State of New York. Well, how do you handle that? We can identify elements and dynamically deliver a different piece of content and personalize the content, based upon those kinds of attributes.
I can't get into too much details, but what we're implementing now is the ability to observe behavior on the page. Like, somebody wants to configure a product and they have to talk to the brand and get configuration data. Then, actually reconcile between the two and update the page, post-load and insert content in that ends up allowing us to update the order. All in real time, all without any human intervention.
LEDGE: It's impressive. I know you talked about, while you're not an engineer, but what problems do you know that your engineering team had to solve there? Ultimately, the vision has to flow downstream into that actual implementation. I wonder what feedback you were getting back from the engineering product groups.
MICHAEL: Excellent question. My toes got really, really big, because we stubbed our toes on a bunch of tactical problems. We have no awareness ahead of time what platform we're going to run on, what operating environment the website is running on, what version of anything in the stack that's happening there – whether there are any areas of conflict, people who've implemented other componentry incorrectly. Even if we comply with how it should be done, we've discovered that the developers around don't actually comply. They break the rules because of stupidity or cleverness, or anything in between.
We basically created our own namespace. This allows us to operate more freely and clearly. That was something that we had to do. We needed the ability to do a little bit of machine learning about the constructs and norms of these types of systems.
Our system is what you might call a lazy loader, where we can't be a blocking script because e-commerce has to flow very quickly and you can't be a blocking script on such things. How to deal with issues around infinite scrolls, where we're real-time sniffing but the site's changing in real time as well. How are we staying current to those. Nested filters, and how do you reveal content over time? There's quite a bit of tactical challenges to create an incredibly reliable solution.
We have an internal mantra which we've had since day one. We say we have to be a great host and we have to be a great guest. What we mean by that is, we host the brand's presence on the web. We work with some of the largest companies in the world like Unilever and Pepsi and Dannon and Nestle – all the big companies, work with Canon and SiriusXM and so forth. They expect us to make sure that their brand always looks perfect. We have to host that brand in a way that doesn't distort the brand in any way, on any kind of site, regardless of device, OS and everything in between.
But we also have to be a great guest because our technology runs on somebody else's website. Even if they behave badly, we have to behave in a way that is non-damaging to any of the consumer experiences that happen on their site. That took us two or three years to sort all of that stuff out.
LEDGE: Let me ask you this. You have a journeyman career. You've done a ton of different things. You, no question from a leadership and business perspective, have brought all that to bear here.
What's different about this experience and the team that you had to put together to just have this problem-solving, bleeding edge disposition. What were the difficult areas there and how do you know you have the right people in the right seats? There's just a lot going on when you jump out and try to build a bleeding edge organization to solve this kind of problem.
MICHAEL: It's an excellent question. Most important experience – and I've learned this the hard way, and I don't want to express where I learned it – but what I've learned is culture is king. One bad apple, you can wreck the whole company. Really wreck the whole company. I've seen it in multiple occasions over the course of my career.
So, I have become extremely focused on picking the kind of people who are smart – obviously – super competent, but don't necessarily have the sharpest elbows. I want conflict and the ability to talk freely and debate things, but I found that anybody who's oriented to their own outcome instead of the team's outcome, who is focused on a personal win over the team's win, not a person I put on my team.
We're super, super, super sensitive now to creating the kind of culture where we can disagree, and we can disagree fervently, but we're all doing it not to be right, we're all doing it to be better.
Our team has actually been together now for six years, and our team culture is great because of that.
In terms of the challenges that we face, we operate on a two-sides sales model. We have to both simultaneously get retailers who are our publishers, and then the brands who advertise through those retailers. That is a harder sales process. We've had to learn a bunch of things about how to really accelerate the sales automation and the sales processes around that. That's been a challenging piece – a chicken-and-the-egg kind of piece.
LEDGE: No doubt. Have you built and used engineering resources to build any of those tools or do you use a collection of off-the-shelf, well-known solutions for sales automation, process automation, CRM? I can imagine you have all kinds of support mechanisms and customer success mechanisms. Did you build that stuff because it didn’t exist, or did you…?
MICHAEL: It's an excellent question. We're ultimately building it. We looked around and we've come to realize, in the last nine/ten months, that we can't get there with anybody. We have a very unique business proposition. Actually, we see it as a real essential element of our business.
If you think about it – at least this is my perspective – the Google innovation around page ranking was obviously an important innovation and I don't dispute the importance of the fundamental search capabilities of Google. But the most economically advantageous thing they did is they created an automated ad-buying process. There were no sales people for like a decade. None. If you wanted to run an ad on Google you had to use their software. It was a software investment.
We've come to the realization we need to do the same thing. So we've actually built out some pretty sophisticated sales automation tools where we are dynamically acquiring potential ads, basically ingesting them into our platform, and then presenting the ad as a completed solution for target buyers. This is a new thing that we just started to roll out in the last 30 or 40 days. We think that, as we build out our solution, our ability to eliminate a lot of the complexity of web advertising is going to be paramount.
Most of your listeners probably are unfamiliar with how the ad industry works, but indulge me 90 seconds to do a quick history lesson. The television business was a very, very beautiful business for advertising agencies. They got 15% of all of the spending that went on television. You would get paid to create a commercial or you would do the commercial as part of your deal. Maybe you spend a half a million dollars on the commercial and you bought $50 million worth of media, you got 15% of it. In the old days, you had three networks that you had to get your materials to. It was really easy and very, very, very profitable.
Now, with internet advertising you have so much micro-targeting and you have so much personalization and data integration and data optimization and things that you're trying to do to make your ads effective, the cost of sales is like in the 30% range. You're make 15% margin on a product that cost you 30 cents to do, that's not a business that you're going to make money at. It's very complicated.
We have come to recognize that we have to build in not only the facility to deliver this interactive content anywhere it needs to go on the web, but to be able to do it in a way that people can bring that content in and manage that content almost like a network operating center, if you think about it from that point of view. The sales process, from our perspective, is just a few steps in the UI to an overall optimized workflow, if that makes sense.
LEDGE: Yeah, absolutely does. The first thing that my head goes to is just the mandate now of everyone, retail or otherwise, it just has to be an omni-channel marketer out of the gate. Just how difficult and what an administrative and operational lift that is to just say, "Geez, I need to touch my own site. I need to touch Facebook, Instagram, Twitter, Medium," wherever it is, Pinterest could be a big deal for retail now.
Just the daunting nature of the staff overhead to make that footprint just to drive that single sale, like you said, it may be a low-margin proposition. The mandate for scale is just simply outrageous now because of all these massive players.
Last question there. You're able to and I think you brought to bear, something that is pretty big and massive scale. You're dealing with big clients. Advice for the small fries out there that are trying to do some disruptive technology and do not have the wherewithal to do that yet. Just wrap around your full experience there – 10 startups, doing some really disruptive stuff. What do you recommend for people that are starting out, bootstrapped or small seed round?
MICHAEL: Well, understand, we've raised all of our money from Angels, so we're a very thrifty and scrappy firm. We have great, great customers and we're delivering tremendous value to our customers.
One lesson, do a one-sided sale. It simplifies everything. This is why Amazon is doing so well, because they don't have to… They own one side of it, which is Amazon. That's a lesson that I've learned.
Second lesson I've learned is one… As a consultant I helped about 70 startups, so I have a very broad range of experience with companies literally from infancy to, obviously, companies the size of Apple. What I've learned is timing is everything.
With the case of Steve Jobs, Steve Jobs created NeXT in the 1980s. It failed, really, as a company. NeXT is OS X. OS ten is NeXT. He was too early. One of the reasons he failed is he forced everybody to optical drives and they were too slow, they weren't ready. One of the things I've learned is, sometimes the timing is the hardest thing. You’ve got to know when is it really ready, when is the market ready, and what are the components of the market that are ready.
I would say the next thing that I would focus on is what is the friction, and how are you intending to address the friction? It is not usually lack of interest that kills a project, but the friction that kills the project. You worked on your beautiful, elegant piece of software but you don't realize they also have to work on a piece of software and you're number 10 in their priority list.
I see this a lot with startups who have a cool component.
Again, without naming names, I had a customer, this is 10 or 12 years ago, and they had the most brilliant piece of software. It was astonishing. It came out of spy technology. But they had their own language so you couldn't use C++ or something, you had to use their language. Well, that's friction. That requires a developer to make two purchase decisions. One purchase decision is, is your software going to solve my problem? Then, is it worth my time to learn your new language to program in that environment? The answer, ultimately, was no. The company didn't make it, even though the core technology was astonishing.
That was almost 20 years ago, I still haven't seen anything like it. They misunderstood that people have to use this stuff and they have to understand the entire end-to-end process that's needed for it to be adopted and to work. That's something that's still an open issue for a lot of people, and they don't think those things through. By the way, big companies make the same mistakes. I saw that problem in places like Adobe and others. It's not isolated to startups.
LEDGE: Yeah. I love that insight. That’s fantastic. I do see that a lot, with misunderstanding the implementation and even adoption hurdles. It's just like, you may have the most amazing thing but you've got to get through those largely human problems of just determining how to put your thing on top of the priority stack. I love that.
Michael, thanks for spending time. Fantastic advice here. I know the audience is going to love it.
MICHAEL: Well, I appreciate the time and…