Reviewing the Reviewers

You are likely familiar with Yelp, the local business ratings and review platform. It's been a phenomenal success, but it also has a large number of detractors, in large part the businesses that are the subject of those ratings and reviews. Yelp is a business that, if examined dispassionately, really should never have succeeded. The primary reason was its business model: let anonymous users pick apart businesses in published reviews, then try to sell advertising to those same businesses. Yelp made this challenged business model even tougher by introducing a secretive filtering algorithm that would decide what reviews got published. The objective was to weed out spam, but all it did instead was to spur conspiracy theories among businesses that felt good reviews were being swallowed while bad reviews always seemed to get published.

Tough business, right? Well it gets tougher, particularly because Yelp showed little interest in mediating disputes (for example, there are documented cases of restaurants getting bad reviews on dishes they have never offered), essentially admitting it was too much work. Mix into this the inexperienced sales force Yelp fielded, giving rise to stories of reps offering to make bad reviews disappear  in exchange for advertising, with a raft of lawsuits claiming extortion quickly following.

Things seem to have calmed down for Yelp in the last year or so, but it's hard to imagine that the rift between the business community and Yelp has fully mended. Yelp is more powerful than ever, and can make or break a business. Yet it maintains as a core principle that it is a consumer empowerment tool, even though Yelp generates no revenue from consumers.

That's why I find it surprising that Yelp just announced the acquisition of Eat24, a service that lets people order food for home delivery. Yes, the company that controls the reputation and success of restaurants now wants to control their order flow as well. I see nothing to suggest that Yelp has become a friendly, trusted brand to the average local restaurateur. Yelp brings scale, but a lot of baggage as well.

What is the correct business model for a ratings and review business? There is no easy answer, especially as the consumers who typically provide the reviews show little appetite to pay to access them. One exception is Angie's List, which sells subscriptions, but even Angie's List now makes more money from advertising than subscriptions. Fortunately, Angie's List found a middle path that allowed this revenue pivot without compromising its credibility and integrity.

TripAdvisor is another reviews site with many of the same issues as Yelp. But TripAdvisor makes most of its money by selling eyeballs, a traditional media model. This means its doesn't have to rely on hotels for revenue, though it recently started to push in this direction.

The real estate website Zillow posts its estimate of a home's value right next to the (almost always higher) asking price. One can presume that's not helpful to making the sale. Awkward? Well, Zillow now asks consumers to rate and review the real estate agents to whom it sells advertising.

In many respects, the jury is still out on what does and doesn't work for review sites. What we've seen to date is that if you can build a big enough audience, the advertising dollars will follow, no matter how upside-down your business model. But just because they pay you doesn't mean they have to like you.  And this may come back to haunt these companies, if not in their core business, then by ultimately limiting their growth and expansion potential.

How Do Your Customers Rate?

It’s recently become known that Uber not only allows its customers to rate its drivers; it also allows its drivers to rate its customers. If you’re loud, abusive, demanding or otherwise unpleasant, you might find yourself silently boycotted by Uber drivers, none of whom are obliged to pick you up. Is this scary, unfair, an exercise in pure democracy or a wake-up call to consumers, who have to date exercised sometime life-and-death power over all types of small businesses with their comments, ratings and reviews, often furnished anonymously? It’s too early to know if this move by Uber foretells a trend, but it’s worth exploring.

Spend just a few minutes online, and you’ll quickly learn of the outrage of small businesses against rating and review websites such as TripAdvisor and Yelp, not to mention the periodic lawsuits. These businesses see an issue of basic fairness: why should unknown strangers (who may even be my competitors pretending to be customers) determine the success of my business in a manner in which I cannot even defend myself?

On the flip side, should businesses be empowered to rate their customers? In some cases, it’s simply not possible: customer transactions are largely anonymous. But for big dollar B2B transactions, rating and review platforms already exist.

An early example of this is a site called TheFunded.com. It had the audacity to let entrepreneurs rate venture capitalists, anonymously. It created a firestorm in the industry, with venture capitalists up in arms about the unfairness of anonymous reviews. In fact, the outrage really stemmed from the upending of the power dynamic in that business. Suddenly, entrepreneurs were no longer supplicants.

Business credit website Cortera has an interesting approach, creating online forums for credit managers in specific market areas to exchange information on companies. It’s a good concept, but one where it’s difficult to get a critical mass of interactions.

The idea of businesses rating customers is not completely new. Indeed, companies like ChexSystems operate “bad customer” databases used by banks to judge whether or not they want to do business with you. And in the apartment rental industry, numerous databases exist to report bad tenants, some with catchy names like badtenantslist.net and donotrento.com. These are not credit rating databases as much as they are places to report poor behavior.

So maybe widespread customer ratings will come along faster than we think. And if they do, a nifty data opportunity will arise: aggregate the ratings of customers that can then be used to weight the ratings these consumers assign to businesses. In algorithms, veritas.

How Many Ways Can You Monetize Data?

I watch the real estate sales vertical with great interest. There’s a lot of data, and money here, which in turn means a lot of innovation and competition. Companies like Trulia, Zillow (which are poised to merge shortly), Move (which operates the Realtor.com site) and a host of fascinating and scrappy regional players such as PropertyShark makes for endless creativity and impressive user experiences. The first thing you notice about all the online real estate information services is that none of them is trying to disintermediate real estate brokers. Indeed, these services typically have business models that depend on agents for revenue. Thus what has happened in this very unusual market is that customers have taken on the primary work of discovery (formerly a big part of the agent’s job), even though agents haven’t reduced their commissions to reflect this.

The second thing you notice is the wealth of structured data that is available for parametric searching. Search by zip code, price range, bedrooms, lot size, and much, much more. In fact, such powerful searching is table stakes now. Map integration? Done. Alerts? Done. Rich multimedia? Done. So what’s left to innovate?

Zillow burst onto the scene (beautifully timed to coincide with our late, great real estate boom a few years back) with its audacious system that put a price valuation on every home in the country. That brought it tremendous visibility, but also introduced consumers to the power of predictive analytics.

Trulia later upped the ante by overlaying neighborhood crime statistics on its database. Not to be outdone, its competitors overlaid school district boundaries to map the schools nearest to each home. Trulia then upped the ante again, licensing data from our Model of Excellence winner GreatSchools.org, that showed the relative quality of each school. And that’s where the market seems to be headed today – qualitative assessments of neighborhoods, along with more predictive analysis.

As you might expect, qualitative assessment starts with Census demographic overlays. Real estate site Movoto.com is already there, with zip-level income, education and ethnicity. Some other sites are hesitating because of the vagaries of real estate anti-discrimination laws. But that is not an impediment to third-party data providers such as Onboard Informatics, which provides a raft of local data, including an innovative “lifestyle search engine.” Other sites like neighborhoodscout.com provide sophisticated demographic views of local areas. And we’d be remiss not to acknowledge diedinhouse.com for those who need to know if former home occupants left on their own power or not.

But what’s most fascinating is that this lifestyle analysis of neighborhoods has even been elevated to a personalized, consultative model. The New York Times recently profiled a New York area firm called Suburban Jungle that helps homebuyers target areas based both on demographics and deep market knowledge. Suburban Jungle doesn’t sell real estate; it refers its clients to real estate agents in exchange for a fee-share, another great example of how many different ways data can be monetized.

CQ Roll Call: Going Big on Analysis

The superheated interest in Big Data and associated analytics continues unabated. As we have long noted however, while there are many ways that data publishers can tap into opportunities created by Big Data, they themselves tend not to be sources of Big Data. That’s not a bug: it’s a feature. So much of the value-add of data publishers come from distilling, organizing and synthesizing of data. Indeed, those data publishers that continue with the old-style model of delivering giant data dumps to their customers are the most challenged players in this more sophisticated and demanding environment.

But despite this intense focus on Big Data and analytics, let’s not forget that there is still real value in all types of analysis, no matter how it’s created. I was reminded of this today while reviewing a new offering from CQ Roll Call. If you’re not familiar with CQ Roll Call, you can think of it as a giant legislative tracking service. For the inside players in Washington, it’s a must-have: it’s the place to go to hear developments first, along with authoritative analysis. But CQ Roll Call also has an “inside sports” aspect to it: an endless series of details wash over you, and if you’re not deeply immersed in a specific piece of legislation, for example, it’s hard to get up to 40,000 feet and see what’s going on over time and in context.

CQ Roll Call is starting to address this need (and likely a latent market as well), with a series of detailed backgrounders, all on high profile topics. They’ve produced 45 of these policy backgrounders to start, with more to come.

What really caught my eye is that these backgrounders, unlike some many that are cranked out by publishers, are designed to be dynamic: they stay up to date. Also of interest is that they are not divorced from the company’s core data product: they link to CQ reporting, the actual text of each bill and other CQ research. It’s an elegant way to tie together a lot of the company’s products with a simple overview designed to encourage drill-downs by those who want to get deeper into the subject. CQ Roll Call touts these reports as a sort of “Wikipedia for policy wonks.”

The simple insight here is that even in the era of Big Data, ALL analysis products remain useful and valuable, and even more so if your content is fast-changing and intricate. Drawing meaning from your content is what your customer do with your data: do if for them and they’ll not only thank you for it, they’ll probably also pay you for it!

Content, Technology or Something Better?

There has been a recent flurry of head-scratching (here, here and here) on the topic of when a media company should better be thought of as a technology company. It’s a good question, but it’s a question muddied by the slippery, umbrella term "media." I am of the belief that if you create and publish articles, you are a media company, even if you happen to be running on a proprietary content management platform.

But when it comes to data publishers, things aren’t so clear. Data is a form of content that plays very well with software. In fact, most data products would be a lot less valuable if they couldn’t be used effectively by software. The real question is: whose software? Those data publishers whose roots were in print directories had the business mentality of most print publishers, which was to ship out big fat books filled with information and let the customer figure out to how extract value from them. When these publishers first began to offer electronic versions, they followed the same approach, shipping out Excel sheets and letting the customers once again figure out what to do with them. Those who did wrap software around their data were known mostly for creating really bad software, and found that their customers were asking, sometimes begging, just to get the raw data without the software. This led to the conventional wisdom that publishers couldn’t and shouldn’t create software, and technology companies couldn’t and shouldn’t create content. In theory, the two camps were supposed to partner, thus marrying great content and software. But it never seemed to work. There were too many issues around revenue splits and who owned the customer, not to mention a bevy of marketing, sales and operational issues. It’s only been fairly recently that most data publishers woke up to the fact that that selling raw data was not only leaving serious money on the table, it was eroding their perceived value as well. Thus the smart ones began to invest to bring in the talent and tools they needed to create top-notch software customized not only to their data, but to the needs of their customers. The results have been uniformly brilliant. Data wrapped in (good) software means higher price points, more customer engagement and better renewal rates. It’s also forced publishers to get a lot closer to their customers, because you can’t build good software unless you fully understand its use cases. As I see it, data publishers can fairly lay claim to being technology companies. Indeed, many now report spending more on software development than content. But when you think about it, why would a data publisher want to be considered a tech company? In a way, that’s slumming. After all, what’s more valuable: a salesforce productivity tool, or a salesforce productivity tool pre-populated with high quality and regularly updated sales leads?