A Review of Reviews

Reviews are important. That’s no secret. Almost everyone uses them now as part of their pre-purchase research. We depend on them. We want them. And in the time and attention deficit world we all live in, we need them to help us quickly make smart decisions.

The basic premise of online reviews can be summed up as "in numbers, truth.” If you have enough people reviewing something, the real answer will emerge. And it will overwhelm all the cheaters, frauds and manipulators who are posting reviews as well.

But in order for a review site to build the volume of reviews, it needs to focus. If you want reviews on all the hotels in the world, you need to stay true to that mission. Same if you’re trying to be the authority on restaurants. There’s always time to expand your scope later, once you are established, known and successful. This is a simple, but key driver behind the success of sites like TripAdvisor and Yelp. If you want people to come to your site to read reviews, you better have reviews to read. And once they’re reading, getting them to post reviews is pretty easy as experience has shown.

That’s why I was puzzled to read recently that a Danish site called TrustPilot had just raised over $73 million in new funding. There must be innovation here, right?

Well, TrustPilot is indeed innovative, but not the way I had imagined. As far as I can see, TrustPilot wants to review every business in the world (and it’s already pushing into product reviews as well). Nothing wrong with being ambitious, but in this case is TrustPilot trying to be too ambitious?

Let’s look at the numbers: TrustPilot currently has about 10 million reviews of 90,000 businesses … worldwide. Further, it’s organized by overly broad categories such as “Services” and “Transportation.” In a nice feature, it ranks the top companies in each category based on their review scores, but in all the categories I examined, I had trouble finding any companies whose names I actually knew. TrustPilot is a great vehicle to post reviews, but as a purchase research tool, it’s a mile wide and an inch deep.

Sure, $73 million buys a lot of growth. But it seems like long odds against TrustPilot getting enough review volume across all its categories to reach critical mass. First they need enough companies to become a real go-to destination. Then they need enough reviews of each company for the truth to emerge.

My review: in a business that depends on volume, don’t start out by trying to be everything to everybody.

Everyone into the (data) Pool

There’s a quiet revolution going on in agriculture, much of it riding under the label of “precision agriculture.” What this means is that farms are finding they can use data both to increase their productivity and their crop yields.

To provide just one vivid example, unmanned tractors now routinely plow fields, guided by GPS and information on how deep to dig in which sections of the field for optimal results. Seeds are being planted variably as well. Instead of just dumping seeds in the earth and hoping for the best, precision machinery, guided by soil data, now determines what seeds are planted and where, almost on an inch-by-inch basis.

It’s a big opportunity, with big dollars attached to it, and everyone is jockeying to collect and own this data. The seed companies want to own it. The farm equipment companies want to own it. Even farm supply stores – the folks who sell farmers their fertilizer and other supplies want to own it. In fact, everyone is clamoring to own the data, except perhaps the farmer.

Why not? Because a farmer’s own soil data is effectively a sample size of one. Not too valuable. Value is added when it  is aggregated to data from other farmers to find patterns and establish benchmarks. It’s a natural opportunity for someone to enable farmers to share their data to mutual benefit. This is a content model we call the “closed data pool,” where a carefully selected group agrees to contribute its data, and pay to receive back the insights gleaned from the aggregated dataset.

One great example of this model is Farmers Business Network. Farmers pool their data and pay $500 per year to access the benchmarks and insights it generates. Farmers Business Network is staffed with data scientists to make sense of the data. Very importantly, Farmers Business Network is a neutral player: it doesn’t sell seeds or tractors. Its business model is transparent, and farmers can get data insights without being tied to a particular vendor. Farmers Business Network makes its case brilliantly in its promotional video, which is well worth watching: https://www.youtube.com/watch?v=IS4KIrcRMMU

Market neutrality and a high level of trust are essential to building content using the closed data pool model. But it’s a powerful, sticky model that benefits every player involved. Many data publishers and other media companies are well positioned to create products using this model because they already have the neutral market position and market trust. Closed data pools are worth a closer look. Google certainly agrees: it just invested $15 million into Farmers Business Network.

There's No Substitute for Structured Data

Cloud-based contact management software provider Nimble recently introduced a new feature called its “Smart Contacts App.” Load the app to a supported browser, and if you see the name of a person or company that interests you, whether reading a news story or in Facebook or Twitter, just highlight the name and Nimble constructs a full profile on the fly. In addition to basic background information, Nimble also searches a number of social networks to find matching accounts. The goal is to build the richest possible profile of the person or organization, and it’s all real-time. With one more click, you can load the profile into your Nimble contact manager.

This isn’t an entirely new concept, but it’s slickly executed. After putting a magnifying glass up to the various screen captures provided by Nimble, what I think I see is that a lot of the magic depends on LinkedIn. And guess what? LinkedIn is a data product. Nimble’s ability to associate social media accounts is impressive, but still imperfect. Indeed, it asks the user to explicitly confirm every social media account match. Nimble also does a nice job integrating with email so that it can pop up a profile of anyone who sends you an email. Microsoft has offered this for a while now, but this is part of a bigger push by Nimble to have its customers do all their work in Nimble so all prospect and customer data resides in one place, all tightly linked and readily accessible.

I draw two insights from all this:

•  The push to tightly integrate sales prospecting data is serious and intense. The idea of any contact manager (and this includes Salesforce) having a button that says “click to view profile” is quickly getting dated. That means data has to be more tightly integrated into these systems to a degree we haven’t yet seen, and that means software companies will need to license more data from data publishers to get this level of deep integration.
   
• For all its sizzle, this new offering from Nimble isn’t creating data; it’s assembling data from other data sources. To be valuable, Nimble needs data that is accurate, rich and most importantly, structured. You can’t assemble that out of thin air. And that unique characteristic – structure – is what makes data so powerful and so valuable.

When You Centralize Data, You Too Become Central

One of the ways that bricks and clicks are starting to merge is through a technology called beacons. It’s all the rage in retail right now. Acme places specialized transmitters in each of its stores. When a customer with an Acme app on his or her phone enters the store, the transmitter can push real-time, targeted promotional messages to that customer. Even better, the customer doesn’t have to access the app – it’s designed to wake up and alert the customer.
Cool stuff, and what better time to target customers then when they are inches from your cash register. Yet, not every promotional message generates a sale. Despite your best efforts, the customer leaves your store. Now what?

This is the interesting area where a start-up called Unacast is playing. It wants to marry the data you have on the customer who just left your store to online ad re-targeting platforms, so you can continue to advertise to these customers, in the hope of making the sale. Again, cool stuff.

But Unacast is taking this a step further. It is going around to all the manufacturers of these beacon transmitters and positioning itself as a central back-end data repository for this valuable shopping data. As a central repository, Unacast can watch where else the customer is going to gain both marketing and segmentation insights. Did the customer go to a competitor? Better re-target with your best deal then. Does the customer go to discount stores or high-end retailers? A retailer can not only learn a lot more about its customers, but is better able to serve them highly customized advertising messages as well.

It’s a data bonanza that will yield endless benefits, and Unacast is moving fast to lock up this opportunity. That’s important because there’s typically only room for one central clearinghouse in a market.

This is a model you might apply to your own vertical. If you are seeing numerous companies collecting similar pots of proprietary data, chances are there is both a need and an opportunity to be the central repository. Why you? Why not? You’re established, know the data business and you’re a neutral player. Central clearinghouse opportunities typically go to the fleet of foot, especially now because the value of data is much more broadly appreciated. Do you have your running shoes on?

Is Your Data "Datanyzed"?

A new product by a cool young company called Datanyze is capitalizing on some well-established infocommerce best practices. Here’s how they did it.

The core business of Datanyze is identifying what SaaS software companies are using (sometimes called a company’s “technology stack”). To do this, Datanyze interrogates millions of company websites on a daily basis, looking for telltale clues as to the specific software they are employing online, and apparently a lot of categories of software can be divined this way. Datanyze aggregates and normalizes these data, then overlays company firmographic data (Alexa website rank, contact information, revenue estimates) to create a complete company profile.

Datanyze links directly to the Salesforce accounts of its customers, so it can add and update prospects on a real-time basis. At a basic level, the use case for this product is straightforward: a marketing automation platform like Eloqua could use it to find companies using a competitor or no marketing automation at all. But wait, there’s more!

Datanyze’s new product essentially flips this service. Now, Datanyze clients can have Datanyze analyze their existing best customers, and Datanyze will build a profile of these customers that can be used to predictively rank all their prospects, current and future. Here are the best practices to note:

• The transition of Datanyze from a data provider to an analytics provider, something that’s happening industry-wide
• The shift from passive (we supply the data, you figure out what to do with it), to active (here are top-rated prospects we’ve identified for you), and the associated increase in value being delivered by the data provider
• The tight integration with Salesforce means that Datanyze customers just need to say “yes” and Datanyze can get to work – no IT involvement, no data manipulation, no delays
• Datanyze is pouring leads into critical, core systems of its customers, a strong example of workflow integration
• The use of inferential data. Boil down a lot of the analytical nuance, and Datanyze has discovered that companies that buy expensive SaaS software are better prospects for other kinds of expensive SaaS software. Datanyze doesn’t know these companies have big budgets; but it does know that these companies use software that implies they have big budgets

Datanyze offers a concrete example of how data companies are evolving from generating mountains of moderate value data to much more precise, filtered and valuable answers. Are you still selling data dumps or analytics and answers?