In this episode of Product Hacker, we’re joined by Brian O’Neill, the CTO of Monetate, a cloud commerce platform service provider based in Conshohocken, PA. O’Neill will describe how Monetate is helping marketers accelerate their processes by putting AI to work.
Kurt Schiller [00:00:03] Hello and welcome to Product Hacker, a podcast about building digital products. Each month we discuss product strategy, design, and development with experts and practitioners from every part of the product life cycle. I’m your host, Kurt Schiller.
Kurt Schiller [00:00:16] Product Hacker is brought to you by Arcweb Technologies. We help companies solve their toughest business challenges through technology, consulting, experience, design and custom software development. Learn more at Arcweb.co.
Learn how our team used smart UI design to develop an error-reducing, cost-cutting pharmaceutical dashboard.
Kurt Schiller [00:00:29] If you’re in the Philadelphia area, make sure to check out our Healthcare Design Hackathon coming up on May 9th, part of Philly Tech Week presented by Comcast. We’re teaming up with the ShameOnUX Meetup to give design experts and newcomers alike the chance to pitch design prototypes to a panel of expert healthcare judges. Get more info at Arcweb.co/ptw.
Kurt Schiller [00:00:53] For years now, we’ve been hearing about the huge opportunities for automation and A.I. in customer experience. But the proof, of course, is in the pudding. And to date, that proof just hasn’t been there. Luckily, that’s finally starting to change with massive growth in the use of personalization, automation and more. But how do you sell A.I. and automation and more importantly, how do you introduce it to an existing product line?
Kurt Schiller [00:01:16] Today, we’re talking with Brian O’Niell of Monetate, a company that’s grown from marketing analytics to full-on AI-enabled customer experience personalization. As Monetate’s CTO, Brian has overseen the company’s technical strategy and growth and had a front-row seat to major shifts in technology positioning and value proposition.
Kurt Schiller [00:01:34] Brian, welcome to the show.
Brian O’Neill [00:01:35] Thanks for having me, Kurt.
Kurt Schiller [00:01:36] So to get us started, could you give us a quick overview of Monetate and what it is you do for customers?
Brian O’Neill [00:01:41] Sure. Monetate focuses on improving the consumer experience. So right now, if you think about the Internet, it’s really a one size fits all. Meanwhile, everybody has different interests, wants, needs, passions, values.
Brian O’Neill [00:01:55] If you think about your consumer experience right now, you traverse the Internet and it’s really a haystack that you need to go find what you’re looking for. We want to flip that on its head. The Internet should understand you put content in front of you, services, products, you know what’s meaningful for you in a push fashion, right. And personalize that whole thing.
Kurt Schiller [00:02:14] So you’re really focused on personalization now, but that wasn’t how Monetate actually started, right?
Brian O’Neill [00:02:48] That was a real game-changer for a marketer because the previous sort of way to do that would be, you know, submit a ticket to the I.T. department within six months of your campaign life. So we call that marketer agility. So from there was an easy transition to an a/b testing company. So a marketer has an idea for a change they want to make, they’re not sure if it’s a good idea. So they take that thing live, but only for a percentage of the audience. They wait a little bit, see if it’s good, if it is, roll it out to the whole– the whole site.
Brian O’Neill [00:03:16] So that was it. But that’s not personalization, right. That’s just A/B testing tool. But that got you, you know, kind of hockey stick growth for the company. Meanwhile, David Brussin and David Bookspan, the founders of the company always had this personalization. So I joined about four years ago to sort of make that vision a reality where, you know, we’ve rolled out we’ll talk more about it, but we rolled out the artificial intelligence components, opened up the system through API so we could act in more places than just the web because we realized that to achieve our goal of personalization, it meant impacting the consumer-brand relationship and conversation, wherever that might happen. So on mobile apps, you know, everywhere, even an in-store, we have some use cases where we put information from a store associate so they can help personalize your in-store experience.
Kurt Schiller [00:04:03] So going back to that shift from testing to personalization, would you talk us through that pivot? Because I’m sure there was a ton of changes that had to happen from who you were selling to, what features were important, what mattered to people.
Brian O’Neill [00:04:14] And honestly, that’s still happening today. So if you take a look at basically the two things that we sell, they appeal to very different people. So if you look at the A/B testing tool, you’re really– you’re a marketer, you might be in the e-commerce department at a brand, your day–to-day is going into the UI and figuring out what’s going to improve that click-through rate, the conversion rate. That’s how you’re measured. That’s, you know, everything evolves around that. Meanwhile, if you take a step up into the C-suite or right above that, they realize that it’s not click-through rate, that it’s really important. It’s lifetime customer value. Right. It’s improving, you know, the brand image in the consumer minds so that when they come into market, they think of you first. Brian O’Neill [00:04:54] So now we’ve got these two levels. You’ve got the marketer who needs to get their job done, roll up these campaigns and is assessed on click through rates. Meanwhile, the company itself wants to impact that lifetime customer value. So these days, you know, the changes in the product itself is instead of focusing on focusing on session-based metrics, which is that click-through rate, we’re taking up a level. Take a longitudinal perspective of the consumer and figure out how you can impact that lifetime. So this is more about what are the behaviors of a consumer when they’re dedicated to your brand over a long period of time.
Kurt Schiller [00:05:31] In terms of how you’re testing versus the personalization products were positioned, did you have to go through any changes in terms of the technical aptitude, I guess, of you were selling to?
Brian O’Neill [00:05:43] Yes, absolutely. So anyways like I said, it’s ongoing, so I’ll take you through the buying cycle that we have. So for that personalization platform, you know, this is not. Just your website anymore? Right. This spans channels, it’s e-mail, mobile app, like I said. So when the CEO has this goal of personalization, improving that experience, the only person right now that crosses all of those silos is the I.T. department, right? So they get their CIO or CTO involved. And now it becomes a technical RFP, right. So, you know, there’s one hundred and fifty questions and a big RFP. And a third of those are our technical, right.
Brian O’Neill [00:06:19] Because they see it now as infrastructure, almost as middleware to achieve this vision of personalization. So that comes with all of the you know, all the typical things like SLAs, you know, we try and keep our response times down under 20 milliseconds, which is absurd sometimes. Literally the speed of light starts to play into that.
Kurt Schiller [00:06:37] I mean, that’s gonna be a big change from the more marketing departments are probably especially now or more aware of the technical aspect of things as they’ve had to become that. But several years ago, that wouldn’t necessarily have been the case. I’m sure they were just looking for like a drop-in, like you said.
Brian O’Neill [00:06:54] Yeah, that’s exactly it. I mean, if you take a look at our UI, I think a lot of the success of the company was about user experience for that marketer. They could log in– it’s literally– . So we overlay the client’s site and they can click and replace a banner, you know, really, really simple. And again, that fueled that trajectory. Now we have a more focus on all the technical aspects. The API is different forms of data can we take in?
Brian O’Neill [00:07:18] Because if you take a look, if you think about the I.T. department, they’ve probably spent probably millions trying to understand their customers and pull together a customer profile. They’ll have their own business, intelligence and data science, people trying to figure out what personas they have in their consumer base. We can then ingest all of that information. So that persona has a liklihood to purchase different categories of items. And then we’d make that live because we have this ability to action on site.
Kurt Schiller [00:07:46] So Monetate was really, sounds like, coming up as personalization itself was just becoming a bigger and bigger thing. And it sounds like your founders were lucky enough to have been out ahead of that. Or maybe some were smart enough to be that. How does monetary approach the challenge of staying technologically competitive as more and more people jumped more and more into that space?
Brian O’Neill [00:08:05] Yeah, exactly. Yeah. So we do we have the advantage that we’ve been around for a decade now, we’re a 10-year-old company, David Brussin and David Bookspan, absolutely visionary. Right. So with the thought that they were eventually gonna get to personalization, we actually store everything over all time. So, you know, I don’t know, we’re probably at a petabyte, the three-quarters of a petabyte at this point of data that gives us the real advantages. When we’re researching and algorithms to deploy and artificial intelligence, we can actually run retrospectively to see how would they have done.
Brian O’Neill [00:08:33] So that’s a huge advantage, to, you know, compared to where we started, that’s just trying to get off the ground and guessing that a certain algorithm will work. Also with that, you know, we’ve learned to run at scale. You know, we handle the, you know, the biggest shoe companies in Europe that have these, you know, large releases of shoes. You know, we handle that scale.
Brian O’Neill [00:08:52] So because we’re deployed globally and a customer running at these, you know, insane SLAs, that gives us an advantage. And we can then roll out new features or feature after feature on top of that infrastructure because we’ve already solved that hard problem.
Kurt Schiller [00:09:06] I think that the data challenge that you mentioned is so interesting thinking about it as a marketer, because so often you’re really just trying to intuit, “Well, I know that things have changed. I’m trying to look back at past performance”. But so often you are almost starting from ground zero with a new marketing methods, that’s gotta be just as a huge opportunity for you guys. You can look back and compare and kind of intuit where things are going to work.
Brian O’Neill [00:09:30] Yeah, exactly. And then I’d even add to that a little bit and maybe go back to your previous question about selling to two different levels.
Brian O’Neill [00:09:38] So we rolled out this artificial intelligence. So when you come to a website, it takes everything into account. So where you are geographically, the whether that’s outside, you know, census information, all of this information feeds into the engine. We present an experience in front of you. And then we watch to see, did it impact your goal metric? Right. Whatever that way is, conversion rate click through. So it learns and it balances this exploitation of what it thinks the right answer is versus exploring all the other options that are out there. But if you think that the marketer this point. Right. It’s a little hands-off, right. They’re not used to doing that, they’re accustomed to creating rules that they know or into it would work. So what we had to do is build a second level of inspection or insights. So we actually take all the data that we’re collecting, run another algorithm, sort of reverse engineer what the machine is figuring out, which gives the market this warm, fuzzy feeling that, “Hey, this experience or this category of dresses like black dresses are selling well to people that use iPhones in New York. Right. And it’ll put that in front of the market”. We call it the Harry Potter sorting hat. You never know what’s going in the sorting hat’s mind. But by watching how it’s categorizing people, you can sort of figure out the rules that it’s using.
Kurt Schiller [00:10:53] You’re basically using AI to assign people to personas in real-time.
Brian O’Neill [00:10:56] Exactly, that’s exactly it.
Kurt Schiller [00:10:58] That’s really cool. I want to talk about that kind of A.I. machine learning angle in general, since it’s become such a hype oriented topic in the last couple of years. I get the impression you guys are really legitimately using it. And I just know it sounds like you guys were maybe even a little bit ahead of the curve on moving over to that. How did that process come about?
Brian O’Neill [00:11:17] So a couple of things there. So eventually we see programmatic personalization. So, you know, the machine even generating the ideas to roll out to the consumers. But along that trajectory, we’re going to ease, ease marketers into it. So maybe its methodology and just walk them through a workflow to roll out a new experience rather than the machine doing that entire thing. But we did actually look at the market and said, OK, we’re going to push A.I. We’re going to develop it, get it live and actually, to your point, we actually got feedback from Forrester that– we showed him our interface and showed the A.I. working like, wow, you’re not just talking about actually there’s so many people talking about this, but this is actually running
Brian O’Neill [00:11:57] We’re like, yeah, of course it’s running, but along kind of a funny story along the way. I was so confident, David Brussin was so confident in AI, we’ll just run this.
Brian O’Neill [00:12:06] Why wouldn’t a marketer use this on every experience they ever went live with? Right. Just make it an option. Let the machine figure out the rest. But that isn’t what happened. I mean, I’d actually get in arguments with our V.P. of product at the time because I was like, “We don’t have a control group and demonstrate that this thing’s working. The math just works”. That wasn’t what happened at all. People were hesitant to use it. We actually had to have a control group and demonstrate that this is actually working better than you could have done yourself.
Brian O’Neill [00:12:33] So all those things along there sort of put a stake in the ground, and said, we’re gonna go AI. Get it live and working and then we’ll bring the marketers with us.
Kurt Schiller [00:12:41] What was the customer reaction? Because it sounds like having come a bit before the hype, people probably weren’t clamoring for it.
Brian O’Neill [00:12:47] Yeah. No, no. This is that you know, is it Geoffrey Moore with the crossing the chasm, the adoption curve. So this was the– or it where our testing tool was sort of– on the other side of that curve where it’s, you know, the laggards it’s just an accepted thing, everybody needs it. AI was on the other side of that chasm, right.
Brian O’Neill [00:13:04] It is the people that share our vision and want to get there. But, you know, we didn’t have the results at that point. So it’s more of a you know, the passionate people that have the faith that this is the thing that’s gonna work.
Brian O’Neill [00:13:18] But now, you know, it’s slipping and we always try and figure out where we are on the chasm at this point because we have evidence of it working. We can show you the ROI that this thing produces.
Kurt Schiller [00:13:26] Do you feel like the kind of machine learning industry, I guess is catching up to where you guys are in terms of level of maturity? Is it something where it’s becoming something that can just be dropped in or is it still a ton of hype out there?
Brian O’Neill [00:13:41] I think so. There is still a ton of hype out there, but there are just like table stakes as well. You know, we do product recommendations. Like the Netflix like collaborative filtering algorithms, their called, which you know “You view this, you might also want this” like that’s table stakes. Where we’re ahead is our trajectory.
Brian O’Neill [00:13:56] If you talk to some of these marketers that are more forward-leaning, more visionary, they’ll look at our AI and say, of course, if I put a 20% percent off coupon in front of somebody, they’re gonna click more. That’s what your machine is gonna tell me, that’s not very useful. Instead, what they want to do, what marketers want to do is change that impression in the consumer’s mind. So, you know, maybe it’s a sustainable materials, you know, worth their shoe company. Maybe it’s breathable shoes. You know, those kinds of messages get that across. That’s what’s going to move the dial on the lifetime customer value.
Kurt Schiller [00:14:28] And that’s so interesting because it sounds like once you know for sure what’s going to work, you can start taking out the stuff that isn’t working, which is always like that’s always the marketing conundrum is like I’ve got I’ve got these 20 things, these 20 positives. But if I give someone 20 positives, they’re not going to go through all of them. I want I really want to just give him or her the one thing that’s gonna work.
Brian O’Neill [00:14:50] Yeah, exactly. Yeah. David Brussin does this analogy where, you know, you’ve got flavors of ice cream and the traditional sort of market or route was let’s figure out if chocolate or vanilla is better, then go live with that message, right.
Brian O’Neill [00:15:03] With our system, you know, just hey roll out strawbery, roll out, you know, rum raisin, right down the line and let the machine figure out which ice cream goes with which person. So, but when we walk into some scenarios, they have so many rules of, you know, they build up over the years on the marketer campaigns and what recommendations to show where. I don’t how a marketer survives and we can reason about all these things that they have live. Right. So in a way, this greatly simplifies everything.
Kurt Schiller [00:15:30] How do you go about that process? I guess onboarding, I guess we’ll call institutional knowledge that your customers have existing about their own customers.
Brian O’Neill [00:15:39] Yep, yeah. So the phrase I use is cyborg marketing. So there is an AI engine there. Right.
Brian O’Neill [00:15:45] But there’s still that aspect of guardrails and things that the marketer does know about there. You know, they’ve done the research. They’ve been living with their brand. You have to incorporate both of those. The market will come in. They can set up so the guide rails for the AI and then they let the AI take over within.
Kurt Schiller [00:16:02] That’s part of that shift again from testing into personalization, then even kind of I guess on my almost automation, you’re really going from a tool into increasingly a service. How has that impacted how you guys approach your products?
Brian O’Neill [00:16:17] A lot of that is the positioning aspect. And we’ve even changed– you know, we used to use the word tool. “Buy up our tool to help you do things”. And now we’ve changed it entirely to the platform, right. And that that is pervasive throughout our organization and influences everybody. So when you think of a sales cycle for tool versus a solution, right, that does personalization, everything changes. So the marketing position, the way that you go in and have to ask questions during the sales process. Right. And discover what they’re actually trying to get at and even help them get to it, because right now, personalization isn’t well-defined, right. Some people have taken the personalization word just to mean better product recommendations, you know. But we’re talking about, you know, the content, the services you deliver, the inclusive of the product recommendations. But the holistic view of personalization isn’t well-defined so with the help all of our to prospects understand that.
Kurt Schiller [00:17:07] I bet the winds are a lot more meaningful to in terms of when a service is succeeding versus a tool, because, you know, as a marketer, I’ve got 30 different tools at my disposal. Someone comes along with here’s a thirty-first tool.
Brian O’Neill [00:17:21] Right.
Kurt Schiller [00:17:22] I’m really hesitant to even bother trying to on-board it because the reality is I’m probably not going to have the time for it was you’re saying no, you don’t have to take the time. It’s a platform for you. It’s not just a tool that you’re putting in a box.
Brian O’Neill [00:17:33] Exactly. You know, if you look at these solutions cells, they’re really strategic partnerships.
Kurt Schiller [00:17:38] Right.
Brian O’Neill [00:17:38] So, you know, the client, the prospect or client is agreeing that we share a vision and that we’ll get there over time, getting quick wins along the way.
Kurt Schiller [00:17:47] I bet once people have pulled the trigger there, they stick with you guys.
Brian O’Neill [00:17:50] Right. Yes.
Kurt Schiller [00:17:50] Yeah. So I want to talk a little bit more about kind of the marketing aspect of your customer, because even though it sounds like the marketers themselves have become less your buyers, sounds like they’re still your primary users, right? For the most part. So marketing kind of as an industry is– seems like one of the first informational industries that’s really seen widespread task automation in machine learning adoption. With tools like Monetate’s being more and more intrinsic to what marketers do. How do you see the role of marketing changing? Is it going to be like a puppet master inside of all these automated systems kind of pulling the strings?
[00:18:25] I actually think so, yes. But it also depends where you are on this spectrum of sort of upmarket enterprise versus down market. If you just contrast them for a second. So, you know, the enterprise, as we discussed earlier, to achieve personalization, you are literally a puppet master across these different silos, the different channels within your organization. So the enterprises are actually putting together Tiger teams, you know, representative from each silo putting together. And that’s how they’re doing sort of wrangling the beast that is these large enterprises to get personalization live. If you go downmarket, though, even go as low as Shopify or e-commerce, one of those that these people don’t have time to focus on marketing because they’re doing so many hats inside their job. The AI benefits them as well because they can go in, not have to spend time constructing all of these rules and everything and just let the machine take over with some guidance, right. So this is the whole encode the methodology into the tool and let them go live with things like abandoned cart emails, triggered e-mails, that kind of thing. So we can help both sides.
Kurt Schiller [00:19:25] Again, speaking as a marketer, with that level of personalization and even automation, how do marketers not lose sight of who their customers are? What can they get back out of the platform to help them better understand their customers?
Brian O’Neill [00:19:40] Well, that’s a funny thing that we didn’t talk about this ahead of time either. So we’re actually going live with a segmentation capability that allows the marketer to slice and dice how everything’s working. So again, we’re sitting on a mound of data. So to let them explore that data in real-time, slice it and dice it to see what things are working for who is part of the creative process that marketers still need to go. So they can then, you know, take a look at “Who is my underserved segment here that I should be messaging in a different way”.
Kurt Schiller [00:20:10] So it’s even really edging back into that, the analytics and reporting area. That’s really cool.
Brian O’Neill [00:20:15] That’s exactly it.
Kurt Schiller [00:20:16] So last question, I feel like we’re kind of getting into the habit of always making this the last question? What does success look like for Monetate either as a technology or for your customers, I guess.
Brian O’Neill [00:20:29] Yeah that’s a great question. So the reason I joined Monetate honestly is because I believe strongly in this, improving the consumer experience. I really do. And historically, we’ve focused a lot on the retail. It’s our best sector right now. But if you think about personalization as a whole, it has a lot of pertinent areas. So my previous company was in the healthcare space. I actually thought that, you know, while we could leverage monetize its capabilities to keep diabetics on the compliant with their regiment. And then, you know, that would be tremendously impactful for that market. So success probably for me, but also at I think Monetate is to realize this greater vision of personalization and delivering what matters to you as a person, as an individual. Most and not make it this you know, crawl through a haystack to maybe stumble upon what you’re interested.
Kurt Schiller [00:21:25] And that’s really cool because that’s really the argument that marketers always use is not I’m not trying to persuade you to pick my thing. I’m trying to help you find the best thing.
Brian O’Neill [00:21:34] Exactly. That’s exactly it.
Kurt Schiller [00:21:36] Well, thanks so much for being on the show today, Brian. Really appreciate it.
Brian O’Neill [00:21:39] Thanks for having me.
Kurt Schiller [00:21:41] Product Hacker is brought to you by Arcweb Technologies, a digital design and development firm in Old City, Philadelphia. To learn more, visit Arcweb.co. This show is hosted by Kurt Schiller and produced by Martin R. Schneider. As always, thanks for listening. And don’t forget to like and subscribe.