Welcome to the WIRe Series. Recorded live in Austin, this series is bringing interviews straight to you from the WIRe MRx Meet & Mingle event. In this interview, host Jamin Brazil interviews John Martin, Founder and CTO of Measure Protocol.
Contact John Online:
John Martin, Chief Technology Officer, Measure Protocol. Thanks for being on the Happy Market Research Podcast.
My pleasure, my pleasure.
So, we’re at the WIRe event, part of… connected to the IIeX Austin events. What are your thoughts so far regarding IIeX?
I mean it’s fantastic. To be honest, I haven’t been there yet today. I flew in.
You just got in.
But I was just checking my Bizzabo app trying to pick out all of the…
Yeah, and all of the sessions tomorrow that I’m going to go to. Looks like there’s a bunch of System 1 stuff, which is super interesting; a bunch of Big Data stuff, which is interesting. So, yeah, it’s a really good lineup this year.
CTO of Measure Protocol. Tell me a little bit about Measure Protocol. What do you guys do?
So, we’re doing sampling on the blockchain. So, we started life about 18 months ago right, I guess, at the height, of blockchain madness, of ICO madness, and started working on this idea that we just couldn’t run away from. You know two out of three of us founders had a background in market research, had worked at Comscore and Kantar, and had another startup, which was a market research startup that we’d exited awhile ago. And the deeper we got into blockchain, the more we just could not run away from how perfect a fit we thought that it was for this one particular use case that we knew so much about, given our history. And so, we’ve been sort of digging away at it for the past 18 months.
Your executive team, Ben? Is that right?
Owen and Paul.
Owen and Paul. Sorry about that. And you guys have had a successful exit, which is really unique, especially in the context of blockchain since the very early stages of that whole technology. What are you seeing as the central role or, even more specifically, the practical role of blockchain in market research? What is the problem that it’s solving?
Right. I think it’s going to end up coming on in two different phases. When I think about what we’re doing, which is just quite literally putting sampling on the blockchain. So, instead of a sample provider being in the middle of a researcher and a population of respondents, there’s software on a blockchain, which is not to say that we’re disintermediating all the sample companies, right? It’s just to say that, from a technical perspective, this stuff is possible to be done in software alone. But, when I think about what we’re doing, there’s sort of three buckets of benefits: privacy, transparency, and economics. And I think with the current state of the cryptocurrency market, the economics argument is somewhat off the table, right?
There’s this sort of really interesting, idealistic, futuristic business models that are possible when you have a cryptocurrency in the middle of these things where users become, to a certain extent, shareholders of what you are building, which is super interesting and can sort of inspire all of these accelerating positive feedback loops. But, because of the crypto-markets now, that’s not as easily possible to get off the ground, but it certainly should be possible in the future. What we focus on today is privacy and transparency. So it really forces us to build essentially a sample company that is private by design and is… essentially has a sort of 24/7, 365-day year audit on what we’re doing. So buyers of data and respondents can all see the providence of offers, can see payments go through, can see the validity or not of data that’s being claimed on the network and so on and so forth.
I would have thought that security would have been part of the stool.
I guess I bucket that under the privacy. In fact, my talk tomorrow is about the privacy aspect. And one of the somewhat surprising consequences of trying to build this stuff out on a blockchain protocol is that you have nowhere to put respondent data – not the data that they’re entering in surveys necessarily but when we’re doing sampling, when we’re running research, all day long we’re sort of collecting and curating and digging through demographic profile data, which is incredibly sensitive and it’s not at all the type of data you want to put on a public, immutable database, right? So, you end up doing away with the respondent database; there’s no respondent database. And what we’ve done in our particular example is push profile data out to the edges of the network. So respondents have all of their profile data on their devices and then the challenge in building this business becomes… If you don’t have a respondent database, how do you run feasibility on a study? How do you know what the composition of the network is? How do you filter somebody into a particular study? So, we’ve sort of come up with these baskets of cryptographic techniques to try to fish that information out, but in a way where we, as the service provider in the middle, have no access to the actual data itself. So it’s sort of privacy by design. You almost don’t have to worry about security because there’s nothing to secure. You keep all of the sensitive stuff off the blockchain.
The pilot project: Tell us a little bit about that. How did that work?
So, I guess starting at the beginning of year we gathered together about a dozen clients and sample providers and research agencies and pulled a whole bunch of projects from them, cobbled together several hundred users on the system, the system which was really fresh out of the oven. We’d kind of just baked this thing and released it to the world, and we set the thing running. And there was really three things that we were trying to do: The first one was just a technological smoke test. (Does this thing work? Does it do the things that we claim?) The second one was education for our clients, for these agencies and for the brands to try to… because it’s a difficult thing to explain: this sort of notion of privacy, of not having a respondent database; this notion of transparency. It’s a difficult thing for somebody to just sit down and intuit, based on reading it from a paper or hearing it from somebody like me. So what we were doing was we said, “Look, let’s just run a bunch of studies for you guys and we’ll come in and we’ll show you what you get out of this, right, and we’ll show you all of the new types of data and we’ll sort of illustrate to you what lives on the blockchain, what lives off the blockchain, and why that might be beneficial. And then, the third part of it was just to do research on research, was to see how users respond to this stuff, which was… I mean it warms my heart; it was bloody good, like the users were very, very engaged. We always think about privacy as being a hygiene factor, rather than a motivating factor. You know it doesn’t get people to show up but it might get them to be more active and to hang around longer than they would have otherwise. And so, it’s hard to say anything definitely just yet. And we’re going to publish our results in the coming weeks, but there was certainly nothing dispositive in what we did. Really, really encouraging, and we’ve turned that into real customers and real partners right now, which we’re just lining up on the network.
That’s beautiful. So, two things: one – like how do you guys get paid? Where do you fit from a dollar perspective? Is it an existing dollar that’s been budgeted by the customer for that particular sample? Or is it a new dollar?
So, today it’s probably coming out of two existing budgets. It’s coming out of the sample budget. I mean we’re taking a technology fee today for each…
Sort of analogous with the marketplace fees.
Yeah, yeah. And then the other bucket that it seems to be coming out of so far is the fraud bucket, right? We have what I think is a quite rigorous, interesting story about fraud and about validating people’s claimed profile. And so, I think more and more we’re going to go up against the companies that are sort of taking 50 cents or a dollar off of every complete to run fingerprinting and so on. I mean that’s not a threat; that’s just my prediction. I think we have a very good story about fraud for several different reasons. And I think there’s just a lot of survey fraud on web surveys, and the companies are spending a lot of money on it.
There was a presentation that was given a couple weeks ago by Proctor & Gamble, and she said that they had done some root-cause analysis of some research failures, and it boiled down to fraud at the respondent level. And, when they looked under the hood, it was approximately 30%. Another great example is a friend of mine did a study with… I probably even shouldn’t say too much about it, but anyway it worked out where they asked a red herring question, “What is 1 + 1 + 4?” and as a pick list of six answers choice, of which 48% of the people got it right, which is not great. But, anyways, you’re always going to get some level of error in data; it just seems like it shouldn’t be over half of the sample. I mean it’s exciting work that you guys are doing. So congratulations on that, and I look forward to your talk tomorrow. If someone wants to get in contact with you, how would they do that?
Ahh, MeasureProtocol.com and I’m on Twitter at JohnM.
Perfect. You got to join our Twitter Chat. So, MRX…
I see it, I see it, every other week.
Yeah, yeah, yeah, totally. So, jump on there. Be a great way for other people to be able to get in contact with you. And, of course, we’ll include this in the show notes. Thanks for being on the Happy Market Research Podcast, and I hope you enjoy the WIRe event.
Thank you, mate. You too.