InfoArmy is an initiative that aims to create “competitive intelligence reports for every strategically significant company on the planet.” The company’s founder and CEO, Jim Fowler, is recruiting the crowd to help his company achieve this ambitious task. We caught up with Fowler earlier today to ask a few questions about his newly launched company and its goals.

Anton Root, Crowdsourcing.org: How did the idea for InfoArmy come about?

Jim Fowler, InfoArmy founder and CEO: My background is in crowdsourcing. I’m not sure if you are familiar with Jigsaw?

Yes, I’ve read about it.

We sold Jigsaw to Salesforce back in 2010 for $175 million. We had understood the crowdsourcing model, and after Jigsaw, I knew I wanted to start another business. I started to get some thoughts on where the holes in the market are, and of course I wanted to do something completely different from Jigsaw. So the idea came up about creating a comprehensive, competitive intelligence view on companies, and keeping it up to date. I knew the value was in understanding the change in data. So InfoArmy just kind of morphed out of what I knew from Jigsaw. I wanted to do something more ambitious and valuable, if we could pull it off.

You touched upon crowdsourcing there. How exactly do you use crowdsourcing at InfoArmy? I also read that you guys have MBAs and professors signed up to do some of the research. Can you elaborate on these topics?

So the basic idea behind InfoArmy is that each report is built by a team of two researchers. They don’t have to be MBAs or professors, in fact we’re not doing analysis, though there are some estimations. It’s really about doing diligent research and making sure that we collect all the raw data, and then use the raw data in a very standardized, structured way to create estimates on these companies – estimated revenue, estimated number of employees, estimated customers, etc.

We use a huge amount of gamification in this. Our rank structure is a globally known hierarchy. Once you publish your first competitive intelligence report, you get your first stripe, and you become a private. As soon as you publish your second one, you become a corporal; your third one, you become a sergeant, and so on. We also have a bunch of badges. The way it works is that a team of two researchers builds a report and once its published, then the whole army can come in, look at the report, and add value to it. But the team of two really is responsible for it.

And how long does the process take, from someone claiming a company or expressing interest in doing research on a company, to publishing the final version?

It depends on the size of the company. Larger companies take longer because there is more information, but it takes about two to four hours. It’s usually closer to four hours the first time you do it, and then it gets much quicker after. The most important thing, though, is that the research team has to update this report per quarter. So they build the report the first time and every quarter they update it.

There are two reasons why they do this, but the main one is that we are splitting the revenue 50/50 from our report sales with the army [of researchers]. So there is no guaranteed payment, but if this model works, the upside is huge. That is why we have a lot of researchers already, doing this globally – every single time your report gets bought, you get half of the money, basically. There is huge upside, and there is a bit of a land rush going on because people know that the hot companies are going to get bought most of the time.

I was going to ask about that, actually. It’s unpredictable which company makes the news on what week, but do you give preference to higher-ranking researchers for potentially bigger companies, or is it whoever gets there first gets to do the research?

There is a little bit of whoever gets there first, but there are a couple of structures in the way that works. Every report is created by what we call a primary researcher, and they do all the handwork – they go in and collect all the data. Once they have that done, a senior researcher gets assigned to the report, and they work together and collaborate in order to publish it. But nothing gets published until the senior researcher gives the okay.

There are a couple of ways to get on these companies – you have to work your way up the ranks. You have to actually be an officer in the army and become a certified researcher. In order to be a senior researcher, you have to pass a test. So in that way, seniority matters. But in terms of doing primary research, it’s a first come, first served model. We put companies that we want into a queue, and researchers jump on them and build reports.

You mentioned that updating the data is just as important, if not more important than the initial research. Do you encourage the same people to keep updating the data, or will there be circumstances where it will be someone new coming in?

It’s the same team. Keep in mind that there is a big, big incentive to update. The hardest part is to build a report the first time, because you have to go collect all the data. For each part of this report, you have to give sources. We’re not dealing with any sort of insider information, this is all publicly available information, and there is a lot of value in tracking that over time. So it’s usually the same team that keeps the report up to date, but if you stop updating the report, someone else will just come in and get it, and then they’ll start getting the revenue for whenever the report gets bought. So there is a lot of incentive to keep the report up to date.

And you said that’s once per quarter, right?

You have to update it every single quarter. But it’s pretty easy to update. It takes a couple of hours to build a report, it takes a half an hour, maximum, every quarter to update it.

How do you ensure that the information is accurate? I know you mentioned that you give sources, but do you use the crowd to verify information, or anything like that?

We sure do. The crowd plays a role in several different ways. Of course you have two people that are highly motivated for this data to be right.

By the way, let’s really be super clear – these are estimates. So, some of the data you can confirm: how much money have they raised, what funding rounds have they done, who is on their leadership team, what their products are, what press have they gotten. Those are all facts, and they are pretty easy to verify. The harder stuff is figuring out who they compete with; harder still is estimating the revenue. And let’s be super clear, estimates of revenue are not facts. They’re estimates, so by definition, they’re not entirely correct, and we don’t publish them as facts. What we’re saying is – if you were going to ask a question about any company that you can think of right now, where would you get that information? The answer is, you’re probably going to do it yourself. That’s how most people do it. There is no reliable information out there.

Our goal is that our researchers are going to do this far better than anyone else and give people the best possible data on this. Once they publish the report, any member of the army can come in and create what we call tasks. They can point to incorrect or questionable data. Also, comments come in from customers. Finally, every time we publish a report, we send that report to every person on the leadership team and the board of directors, for comments. There is a lot of feedback coming in. But at the end of the day, the research team decides what to do with the information.

What’s the incentive for members of the crowd to check everybody else’s work? Is it just that if everyone’s information is more accurate, the reports are more likely to be purchased across the board because you guys have a better reputation, or is there any monetary incentive?

Do you have an iPad nearby, by chance?

No, I don’t.

It would be easiest to understand if you could see it. One of the most important jobs in these reports is saying who a company is competing with. The report brings in all the data for the other companies that are in the competitive ecosystem, so data from a bunch of other reports goes into your report. And if their data makes your report look bad, then you’re not going to be excited about that even a little bit. In this way, you’re highly motivated for everyone to work on the reports. Plus, in order to be a senior researcher, you have to prove that you know how to collaborate. The way that you do that is you go in into reports of companies that you are familiar with and help those reports get better. That’s the way you get accepted as a senior researcher. The whole point is that our researchers are highly motivated. Any crowdsourcing that’s based on money alone is usually a failed model.

This takes a lot of time, but we’re under no illusion that we can get people to do this without a big financial incentive. Basically, there is a concept in the real army that one bad soldier gets the whole platoon killed. With InfoArmy, one bad report kills everyone’s ability to make money. Everyone’s in this together. What I learned at Jigsaw was that people have a gamer and builder gene – they love to build datasets. We’re seeing that a lot of our early members have the same mentality. They say, ‘This is great, I get to make money! As I move up the rank, I get more and more access to reports on the database.’ At the same time, a lot of people just really like building datasets.

Related to that, could you elaborate a bit more on the gamification? There is a lot of incentive to be a senior researcher, and there is incentive to work your way up through the ranks. But are there any midlevel perks?

Yeah, there are. Basically, we have this concept called ranks, and anyone can get rank as long as you just keep publishing the reports. Every time you publish another report, you move up in rank. But we also have badges. And they’re all described on the website – there is an Inspector badge, there is a Certified InfoArmy Professional badge, a recruiter badge, and more. But the two most important badges are Senior Researcher and Judge.

Any time there is a dispute in the army – maybe the primary researcher and the senior researcher don’t agree or we have to take a report away from someone who is not doing a good job – a judge gets called. So everything is policed by the army itself, and we have badges that empower them to do it. In a way, these badges act a little bit like roles, if you will. It gives them power within the army. As you know from gamification, people love this kind of stuff.

My last question is about translating the reports. I know you mentioned that you want to put out these reports in multiple languages. Are there reports out already in foreign languages?

No, no, in fact, we don’t have reports on foreign companies yet. But it is part of our vision as we move forward. What we need to do is get critical mass, which is about 100,000 reports. We’ll probably end up doing all of those in English. Once we can hit critical mass and this model gets proven, then we’ll move forward with translation. We’ll have translation teams that work exactly like report-creation teams – a primary translator and a senior translator that ensures quality. Again, we will do a revenue share with them. But that’s in the future. Our mission is to have one of these reports for every company on the planet in multiple languages. But it will take us a while before we get there. One thing I can guarantee you, it’s going to take a crowd to do it. (laughs) There is no way to do this without the crowd, period.