Ensighten | Interview with its CEO & Founder – Josh Manion
In San Jose (CA), we meet CEO & Founder of Ensighten, Josh Manion. Josh talks about his story how he came up with the idea and founded Ensighten, how the current business model works, as well as he provides some advice for young entrepreneurs.
Martin: Hi, today we are in San Jose in the Ensighten office. Hi, Josh. Who are you and what are you doing?
Josh: Well, I am Josh Manion the CEO and founder of Ensighten. We are a digital marketing company that is really trying to transform the way that large enterprises collect, own and act all the digital data that they have about their customers. So anything from their website, to their mobile apps, to their display ads, social media. We have been at it for just over five years now.
BUSINESS MODEL OF ENSIGHTEN
Martin: What did you do before and what made you come up with the business idea of Ensighten?
Josh: If you think about what I was doing for, I guess, seven to eight years before starting Ensighten – I was running the digital analytics agency. My wife and I had started this company and it was based in Chicago. We generally work with big enterprise customers, so sort of a similar customer overlapped what we have now. And to me, there was just this growing opportunity that existed and I’ll sort of describe that dynamics that revealed it to me, which was, we would watch these large enterprises go from working with anywhere from when we first started that digital analytics agency, probably two or three different marketing technology, ad tech vendors, and this was sort of like 2002 time period, they’d only have a small handful of those. And then by the time when we started Ensighten, there were certainly dozens and dozens of those per enterprises and in some cases hundreds depending on the size and business model of a particular enterprise they we were working with. And what I observed was that the enterprise was getting value, they were creating a positive ROI of all the initiatives that they were pursuing with these vendors. But what they were not doing was keeping up with the potential value of what they were creating. So if you think of it like moving from three vendors to 30 vendors, you would have this exponential explosion and sort of the potential value of what you could get out of those solutions, but instead they would just have a tiny little linear increase of “Ok, what we’re getting more out but nowhere near what the potential value was.”
To me, it seemed like there were two big issues that existed that were stopping them:
The first was this concept of marketing agility. So if a marketer wants to do something but maybe it’s to collect a piece of data, maybe it’s launching a campaign, maybe it’s to try out a new vendor, whatever it might be, they literally couldn’t do it. They would sit there and they would go to their IT department and they would say “Hey, can you put this little piece of code on all of my different pages,” and IT department said, “But you have hundreds of thousands of these little different pages that’s going to take months to do”. And the marketer would respond with something like, “But that doesn’t work because the campaigns launching in three days.” And so they would respond with, “Well, you’re not going to get that data done,” or whatever the scenario would be. Essentially they were continually bottlenecked by their own processes and systems that didn’t give the marketers the ability to keep up with what the customer was expecting.
The second issue was the data itself. So if you think about working with, I don’t know if you guys are familiar with LUMAscape, it is like this eye chart that has 2,000 vendors on it. It literally maps the growth of the digital marketing ecosystem and sort of the ad tech ecosystem over the years. So they put everybody in these tiny little categories and shrink those down. It’s completely overwhelming, but what you find that’s common about those thousands of vendors that are on there is that they own all of the data that those enterprises are collecting. So the enterprise would essentially engage in this model that would say, “We reached out to you as a vendor, let me pay you to put your code on my site then you would collect my data, but you would collect it in a format that the first party to you and third-party to me which makes it virtually impossible for me to join it with any other data. And if I ever need that data, you’re probably going to rent or sell it back to me.” And it was really rather crazy if you think about it from just with a beginner’s mind. But that was what everyone was used to, that was how I had servers working so that was just sort of the standard that expanded.
So that phenomenon of creating all of these different silos so if I stick with that example or that you have thirty different vendors on your site, those 30 vendors now have created at least, and sometimes more than 30 different silos of data. So your ad server will have one view of who that customer is; your website analytics tool have another; your mobile analytics tool will probably have another; your personalization engine will have another. And what you create is this concept of never knowing who your customer is, or you are always treating them as if you either don’t know them or you know different random facts about them. So you provide very choppy sort of user experiences as they I guess progress through this journey of either learning about your company, becoming a customer, becoming a repeat customer, optimizing lifetime value with a customer, etc.
So the opportunity in my mind was to solve that by essentially giving the enterprise one consistent view of who that customer is, that was first party to them, that was completely owned by them and that would be actionable for all of those vendors to seamlessly plug into. And so when you really boil down what we’re doing in the space are the two things that disrupt the space most:
- We are taking that action ability or that agility platform that was first described, we are creating an environment where the marketer can just drag-and-drop something and collect any piece of data, try out a new vendor but without any involvement from there IT Department, their developers, etc. so it can literally happen in minutes now instead of weeks or months, whatever it used to take;
- And then the second piece is we’re allowing them to create this data profile that is truly you know encompasses the entire journey of the visitor. And when you combine those things together what you do is you create a network effect for the enterprise themselves that says, “Well, if I work with 30 vendors now using a platform like Ensighten actually makes all 30 of those vendors more powerful because they’re able to access a unified profile so they’re all able to make a slightly smarter decision on each interaction.” And we do other things; we make them faster, make them easier to manage and there’s less overhead and sort of having a number of vendors that exist on the site.
So if you are an enterprise you have to have a very strong incentive, start working with vendor number 31 because the original thirty get better when you bring the number 31 or 32, or so on. And so that really just sort of stems from that consulting work that we’re doing at those digital analytics agency for the seven years prior. When I saw that I was shocked that no one else was doing, it was sort of like it came to me and started doing a bunch of research and they were there are people kind of poking around the edges of it but there was no one was really taking that vision to the market the way that I wanted to. And so I spun out of that company and you know the rest is history.
Martin: Josh, how does it work? Imagine, I install this kind of simplified the tag manager in my company. So I have 30 vendors who I’m plugging into. How do you join the data then from those vendors?
Josh: So it starts with the implementation of the tag management system itself. On the back of the card is the implementation guide itself, so single line of code and one of the core requirements that we had from the onset was the implementation of our platform has to eliminate the complexity that is traditionally existed where it would be, “I want to implement a web analytics tool” and it would take months of complexity to configure it customize it, capture all the events and custom variables that you want to capture. And so that single line of code never changes for the client, once they put it there they’re able to take any net new vendor and essentially just take them out of our library, we have over a thousand vendors in our in our app library, and literally just drag them onto the site put in their user ID and password and configure them.
Now, for vendors that are already on their site, this is actually the hardest part of the… I usually describe it this way; you’re implemented the moment you put the line of code on the site and most of our customers start doing net new things immediately. United Airlines was an example of this, they first started working with us right after they merge with Continental, they had a list of 12 things that they wanted to deploy on the new combined website. Ensighten came up as a part of that new platform of the combined company and they realized that once they put the line of code on the site, those twelve new vendors they were done it in less than a month and they plan the entire year to do it. And that was a great example of sort of the agility that they can sort of generate from being able to just do any forward-looking things immediately.
Now the harder part of this is generally when you have an existing site and want to take, let’s say, you’re taking 10 vendors off that site and you want to redeploy them, the redeployment is actually very easy. The hardest part of this process, not surprisingly given sort of the legacy and sort of the challenge of trying to get away from, is you need to have your IT department to go in and actually remove them. And so taking the code off the pages and many of our customers actually will delay that because it’s a traditional IT project and so they say, “Well you know it next time we refresh the site I’ll get it off but just start managing that thing.” And so we can manage the code is already there, it won’t be quite as performance optimized as it would be if it was flowing into our system but it allows the customer to get immediate value out of it.
Martin: Josh, you actually said that you change the paradigm from having a third-party data to really owning the data as a customer. Imagine, I’ve implemented the tag manager I have something like Google analytics on there, are you able to push me in real time raw data of what the customers doing on my website?
Josh: Absolutely and thanks for reminding because I didn’t answer your full question before, which was: if you think about how we are changing the data paradigm, now when a customer who wants to own that data asset what they’re effectively doing is they’re creating their own copy specific to that individual on that – if you stick with United Airlines example, Microsoft or Sony or American Express or anywhere customers, what you end up with is they will take and start creating their own first-party, key to a first-party cookie that is only accessible to them and they will start collecting the superset of all the data that goes. So in your example the data that goes to Google Analytics they’ll also choose to collect that. If they’re using voice of customer solution like opinion lab they’ll also collect that data, if they’re using a personalization engine they’ll collect whether or not test A versus test B was working for this particular visitor or what segment they are in, etc.
Absolutely we can provide that data in real-time to them and a number of our customers like actually there was just an article and ad exchanger about Coke and how they use our platform to do exactly that use case whether collecting all of that data. And their model has been to not work with a traditional web analytics tool but to feed all of that data as a real-time stream into their enterprise data warehouse where they apply all the sophisticated BI tools that they have against that data for the modeling and data science teams that they have there.
But the important element here is that— and sometimes I show this was the first time your prospect of actually kind of short chart that shows like – imagine the customer views an ad. In that world the ad server once you double click will have a data point about that customers. You viewed it out maybe you clicked on maybe you didn’t. In our world, Ensighten would also have a copy of that or the customer take a Coke would have that same piece of data. Now something happens on the website, now double-click won’t have any view of that, this would be let’s say web trans, whatever the web analytics tool was. Well web trans will have one data piece, while Ensighten will have another data piece, so now we have two data pieces, and you have two systems with one data pieces and you can sort of extend that down through, “Ok now I’m interactive the mobile app, I’m on a social media platform I’m doing all of these things.” And it starts to become very clear that Ensighten has the master dataset where all of these other data sets are just as little fragments of, “ok, this is one channels interaction, this is one media type or this is only things on my own properties or these are only things on purchased or paid media”.
Martin: Do you also provide analytic tools based on your Ensighten stuff or is it only that you are collecting other data and then pumping the data to your customer?
Josh: So, we actually do both. So, philosophically we are very committed to the idea that our platform and the data needs to be open so our customers can connect any third-party vendor they want to that and they can extract that at any level of granularity whether it was like the Coke example of a real-time stream of that data coming out or they want to batch exports, whatever might be. So that’s philosophically very important but we very quickly kind of learned that our customers, of course now have this really rich dataset, they want to have Ensighten and they want to be able to immediately explore without taking it out of the platform and sort of taking on additional work.
So we actually did an acquisition in 2014 where we acquired a omnichannel data analytics platform and mainly because of this specific use case. We had this tremendously differentiated dataset and we wanted to rapidly accelerate our customers’ ability to explore and model that data and do omnichannel data analysis and segmentation of their customers. So we have a solution on that side as well.
Martin: Josh, you’re actually supporting thousands of vendors for the platform. When you started out, you didn’t support that many, but what the first three or six months look like? How did you start billing the product and finding customers?
Josh: Vendor support first. We actually started with four key requirements with the system.
- One was this one line of code that I referenced earlier.
- Second was that we had to support every vendor completely natively whether they existed today or they would exist in the future with no modification.
- Number three was we had to work in every digital touch-point, so it wasn’t just the web tool, it we had to work in mobile and had to work alternative things, had to work with any of these emerging devices; ATM machines, whatever it could be.
- And the number four was it had to make all of those experiences faster. It had to be a better customer experience, if you uploaded that code here it would load faster, drive a better customer experience.
So the way we got that was we were not forcing any vendor to modify themselves at all, we were able to just sort of natively support and take— even though we didn’t have fancy apps like we do now at our app marketplace, partnerships with all of these guys. We were able to literally create an environment where someone could cut and paste the solutions and say, “I want to run this version of this vendor”, cut and paste, boom, it’s done and so very easy.
Now the question about how did we get started, I think we had a little bit of an advantage because coming from that digital analytics agency I knew a lot of enterprise customers and I had at least enough credibility with those customers and sort of a reputation in the space when I would call someone up at one of these big enterprises they wouldn’t just slap the phone down. Because I used to have these conversations that would go something like this and say “If I could replace all the tags on your site with a single line of code, it would make your site faster you would never have to talk to your IT department again would you be interested?” And they would be like, “Of course but you don’t have something that does that.” And I’d be like, “But no I do”. And they’re like “Well you never lied to me before.” So “Come on in and I’ll at least look at demo.”
And that was literally how we got our first handfull of customers which were people like Nestle Purina, Paramount investments, Sony Electronics, etc. So we were fortunate and being able to acquire those first handful customers in our first few months of operation. And we took a slightly different sort of path than a lot of others Silicon Valley startups. That wasn’t go, come up with the idea, go up and down in Sand Hill road, raise a bunch of money and then spend a year building product and then go to market. We did it completely backwards.
So we went to recognize the need and the problem and started bootstrapping the company based on the success of the prior company which was a luxury and then as soon as we had an initial version of the product we were also fortunate that we were able to pretty easily attract people and say, “Would you help us beta test this and validate whether it works? Does it really solve the problem we say it does and what else would you like to see in it?” And so when the company started which was officially incorporated on the last day of 2009, by February of 2010 we were literally closing our first paying customers which were generally people coming out of that beta program that we were working with.
Martin: As you have a lot of B2B customers and especially bigger ones and, from my understanding is they having much longer sales cycles, how did you solve this kind of problem? Because when you are starting out you are kind of cash constraint…
Josh: That was a challenging thing for us for sure because we had this initial excitement when we launched the product when we closed our first half dozen customers in the first few months. And myself and other early folks here thought okay well that was just going to take off if we just— we’re going to really make this huge. And then all of a sudden it was like it just we didn’t sell again for like two quarters. We had lots of demos and lots of prospects we just did not get… people weren’t signing contracts.
Part of it is exactly the enterprise sales cycles are long one. Part of it was just the adoption cycle of a new technology – we were in a category that really didn’t exist before, people didn’t have budget set aside for tag management or this sort of digital marketing platform optimization in this way, and so they would see the demo and say “Wow, that’s awesome. I want it. But I don’t have any money, I don’t know how I would get this maybe I’ll ask for it next year.” That kind of discussion would go on.
So it was really about six months we were just struggling, just closing 1-2 customers through that period of time and then all of a sudden the market started to mature to the point where those early customers that we had were starting to have great success and have phenomenal ROI stories that made it easier to sell the other ones. It just started to become a little bit more mainstream people say “Okay, I know somebody who’s done that who’s had experience with it.” We started to get some of our early customers onto the speaking circuit within sort of the industry trade shows and so on, so they were publicizing and helping us publicize. There was just sort of that intangible timing element which is so tricky when you’re doing a startup because you’re going to have the best idea and if you’re timing is just off you can blow through your cash just be done. But yes, that was definitely a challenge for us because we weren’t sure when the market would truly sort of mature and be ready for it.
Martin: Josh, I totally understand the advantage of using tag management like Ensighten, only single line of codes and you join lots of data sources, totally get it. But what are the downsides of using something like this? I don’t know, maybe page load speed, performance?
Josh: Well, the page load speed actually accelerates. It accelerates for a few reasons:
- The second is that we can control how they load so we can as opposed to loading the mall synchronously and say you wait until they’re all, which is how some poorly implemented sites would affect were able to dynamically insert them just the right place in the sites so that the site becomes interactive and that it becomes functional much sooner that would have otherwise.
- And then the third piece is if you think about a high complexity site, a site somebody who’s got fifty tags or more most of those tags literally don’t apply to every visitor, they apply to some of the visitors. So if you’re coming from this campaign but run this code but I might have done twenty times in there. So any one visitor will only have one of those twenty tags actually be applicable to them. In the old world, all 20 would run, all twenty would collect that visitor’s data which is a privacy issue even though it’s irrelevant to them. In our world, Ensighten would actually dynamically decide, “Hey, this visitor has only this piece is relevant to them” and so in that case you’d actually save 95% to load because we would only bring down the tag that needs to run for that one individual and so the other nineteen tags literally don’t even need to download or execute for that visitor. For the next one it might be a different one but it’s able to essentially personalize which tags go until you can take a huge amount of the payload that needs to both be downloaded and the burden you put on the browser in terms of processing that code and actually executing it.
So the downside case is honestly I would say it’s there really isn’t a strong one in the sense of like… The value proposition is sort of so self-evident, it’s very similar to content management in the sense that you would generally not want to start running a website without a platform like this, it just doesn’t make sense to do this stuff by hand. If you were going to dig deep for downside case you’d say well IT’s giving up control just like they would do the content management system someone could post a piece of inappropriate content or in this case they could literally do something without the appropriate sort of guard rails in place and we invest heavily to create those guard rails for people but no matter what you do somebody could potentially hurt themselves with it or break a site or launch something that interferes with some of other piece of functionality on site if there isn’t the appropriate workflow and sort of testing in place. And that’s the biggest objection that we typically hear is that IT and security, in particular financial services, they really want to make sure that there’s a really enterprise-class solution in place that has security workflows, permissioning, that the data is secure because what we’re really talking about is a sensitive asset for global enterprise or any of our European customers in particular that can be very sensitive topic.
Martin: How do you go about this safe harbor topic because, for example, for European customers?
Josh: You mean in terms of storage of their data? So for European customers who want to take advantage of that universal profile, then that data will literally be stored in European data centers and not allowed to transport back to North America which is little bit more challenging from an infrastructure standpoint but again I think one of the reasons that you want to go with a real enterprise-class vendor in the space as opposed to some of the smaller guys that are out there.
Martin: Josh, how’s the revenue model working and how did you decide about the pricing?
Josh: The revenue model is pretty traditional SaaS based software, so it is generally an annual subscription to our service. And that’s like any of the SaaS based technologies out there so you just sign up for a year, two or three years depending on how long you want commit and how big discount you want to get. Generally that is also based on the volume of transactions that you have; going on so how frequently are you interacting with our platform. If you’re used to buy web analytics tool it’ll generally be the same kind of model, I used to buy a billion pages from my site then I would buy a billion server calls from Ensighten to help support that because generally each interaction would going to be responding with here’s a customized for this visitor in these tags that apply to them.
Let’s see, the pricing and how we started with it. I think anytime you’re starting out it’s very hard to figure out what is the value to your customer and you generally have very little idea what your own costs are you going to deliver because you are very early stages, you are not at scale. You are losing money no matter what you charge. It’s tricky. We actually started with the domain based pricing models so trying to say that people could put it all over their domain and it wasn’t server call based, that turned out to be a little bit cumbersome for some of our big clients. So as we started to work with really big organizations like Microsoft they would say well that’s great but it kind of— here’s an example with our store where we have a hundred domains across hundred different countries and that kind of works for you but then if we go and put you on microsoft.com and that’s one domain and its billions and billions of requests that kind of works for us. Now, when I want to get kind of balances out in a customer like that so it’s not that big of a deal but they were actually the one that helped drive us to the decision to move to server call base pricing because they were much more able to allocate costs across different business units and so they would buy large quantity and then allocate across many different business units. So they found that model made it a lot easier than domain based that might have a small group consuming a very little amount of traffic but for a lot of domains that was not necessarily good.
Martin: And for the pricing, did you rather use the cost based pricing, value based pricing or just by looking at competitive pricing?
Josh: So when we started there really wasn’t a competitive example. There was actually a company in Europe called TagMan who we bought last year in 2014, but we weren’t seeing them in a lot of the North American deals when we were just starting. So our primary goal was to try to find a value based price that worked which was associated with the value of the customers paid and where I think we ultimately ended up was that you tended to be a combination of value-based pricing but also so correlating based on what other guys in the marketing tag stack were charging and what people were used to. So if they were used to pay $100,000 a year for solution then they might be equally happy to pay $100,000 for ours but they would have paid $200,000. So it was that sort of tweaking the model that took place in the first couple years.
Martin: And Josh, you said that especially in the beginning you don’t have a good idea of how the cost functions working, how the revenue functions working. How often did you need to adjust the pricing model?
Josh: We tried not to adjust it that often because it was tricky in that first say five customers. It got adjusted a lot because we were you’re not really even sure if you’re in the right ballpark and you’re trying to calibrate your customers are telling you what’s appropriate what they feel good about and you want to be in that zone where they feel really good they’re getting a really strong ROI but it’s also you know when you start dropping that in your models you say, “Okay, I can see how that will work for me It’s not like I’m going to charge you $10 a month.”
ADVICE TO ENTREPRENEUR FROM JOSH MANION
Martin: Josh, let’s talk about your advice for first time entrepreneurs. What type of learnings have you learned over the last 5 years?
Josh: I think there are a few things. One thing is that’s always been true in my own mind here and starting companies is the faster you can get a product in front of the customer and you can actually get live feedback and the closer you can get, even if it’s not a paid sale but the idea of getting you know live product working in a customer environment and seeing that, that to me is the fastest path to creating value. I see a lot of entrepreneurs that I talked with and had coffee with that feel like they can’t do that and I’m always finding myself encouraging them do anything… it doesn’t need to be pretty, you don’t need to have a UI, you just need to demonstrate that what you’re talking about, number one works and number two creates value for the customer. If you do those two things and if you even in the third element of they’re actually paying you anything for it, now we’ve got something to work with in terms of creating a business or if you’re interested in raising outside funds. Those are things that the vast majority of companies that investors see don’t have, at least early stage investors, and so you’re taking these massive elements of risk and just eliminating it. We talked a little bit about one of the things that I would also add in that category which is timing. You have to be very conscious of the timing and not only have that patience when you have a market that you may be a little bit ahead of but making sure that you create the strategy for the company, and in particular the financing strategy for the company, that’s associated with that. I think, for example, if we don’t raise, we ended up bootstrapping for almost three years before we ended up raising our first sort of professional investment and I honestly believe if we had raised money earlier especially during that period I was describing we weren’t selling to so many customers, we would have probably burned through a ton of money and I don’t think any amount of money was going to change the fact that the market wasn’t ready to buy and so we could have gotten ourselves into a very bad situation where we’d set expectations with investors, we’d be disappointed, we’ve already raised money, taking delusion and now we were much more nimble and being able to navigate the situation at the bootstrap company that was just like, “Ok well we won’t hire anybody, we’re not selling to customers” and that’s very easy to make there but if you raised your series A, there goes your expectations, “Are you hiring these people, are you scaling?” And it allows it to progress a bit more natural on its own.
Martin: Josh, thank you so much for your time.
Josh: I appreciate the time.
Martin: And next time you are thinking about what other customers actually are doing on my website and really in a combined dataset maybe you can think of Ensighten. Thanks.
In San Francisco (CA), we meet Co-Founder & CEO of Quantcast, Konrad Feldman. Konrad talks about his …
In Madrid we interviewed the founders of the secondary ticket market platform Ticketbis, Jon …
In New York (NY), we meet DigitalOcean CEO Ben Uretsky. He shares how he came up with the idea …