Models for Being the Best

There is endless innovation and variety in what I call the “best guides” segment of the market. These are guides, print and online, that help consumers find and select the best of something – from hotels to restaurants to contractors to consumer electronics.

It’s a huge market segment. Consider such vast scale businesses as Yelp and TripAdvisor that largely focus on restaurants and hotels respectively, though you can find on their platforms crowd-sourced reviews for just about anything. If you have huge ambitions, it’s pretty well established that the fastest way to develop massive amounts of content is via crowd-sourcing. If you build it, the crowd will comment on it and rate it for you. The downside to crowdsourcing has always been lack of control. Too many people posting ratings and reviews have malign motives. More fundamentally, two people can honestly have diametrically opposed opinions about a restaurant, for example, and it’s close to impossible to reflect this in an overall rating. Crowdsourcing depends on sheer volume for accurate reviews to drown out inaccurate and biased reviews. It works, more or less.

Knowing the limitations of crowdsourcing, there have been many who have tried to refine the concept. Angie’s List was an early pioneer, aggregating reviews but only fromits members and only forits members, what we call a “closed pool” model in our Business Information Framework. The concept worked, but was difficult to scale, in large part because members didn’t want to pay for ongoing memberships. Angie’s List has since shifted to a lead generation model. I also wrote in 2016 about a company called BestPickReports.com, that’s building its business both online and with expensive print guides mailed to consumers. It appears to be a long-term play, and is backed, somewhat surprisingly, by EBSCO.

 We all know about Consumer Reports. For decades, Consumer Reports was the first place to check before making a major consumer purchase such as a car or a dishwasher. Consumer Reports did all the testing and rating itself, and understood that its reputation was everything, so much so that it prohibited manufacturers from citing its reviews, and the owner was a non-profit organization. In the days of print, Consumer Reports did very well selling subscriptions to its print magazine. It wasn’t an ideal way to distribute information (how do you buy that new car in March when the new car reviews didn’t come out until the May issue?) but it worked for a long time because there weren’t a lot of options. That’s why Consumer Reports had some struggles when it moved online because consumers didn’t want an online subscription as much as they wanted to be able to buy just dishwasher reviews and only when they were in the market for dishwashers. Consumer Reports continues to flourish, the result of momentum, its pristine reputation and quality reviews, but it’s quite possible its business model will come under increasing pressure for the same reason as Angie’s List: selling a continuous information service to consumers who don’t continuously need information is just plain hard.

Finally, let’s look at the original arbiters of what’s good, better and best: newspapers and magazines. 

Hearst Magazines has a review site called BestProducts.com. While the name might imply product testing, the site recommendations appear to be closer to the traditional “editor’s picks.” There is heavy use of the phrase “what welike,” and the site overall seems to be much more about informed personal preferences of the writer – more taste-making than research. Indeed, aside from a great (and arguably misleading) domain name, these are product recommendations that would not look out of place as print magazine articles from ten or twenty years ago. Online forced a change in business model, however. Hearst links to vendors of all the items it recommends, hoping to profit from online referral fees.

 The New York Times blends a few models together through its Wirecutter.com site, a business it acquired in 2016. Wirecutter offers much more than the personal opinion model of Hearst, but less than the rigorous product testing of Consumer Reports. It walks a middle ground, doing real product research, but not actual product testing. In terms of business model, Wirecutter follows Hearst, generating revenue from product referral fees.

 Depending on product referral fees is a risky business because of “leakage.” Simply put, it’s too easy to take your recommendation but not click your link. When that happens, the business generates no revenue. The only real solution to the leakage problem is sheer traffic volume, something both Hearst and the New York Timesalready have and can easily leverage. The New York Times, for example, is increasingly citing Wirecutter in its own news stories, albeit with full disclosure of its ownership.

 There is no single best model, and here’s a rundown of the tradeoffs. Crowdsourcing works, and it is cheap, but the quality of the content is uneven. Closed pool crowdsourcing yields a huge step-up in quality, but it’s a tough model to execute. 

 You can generate your own reviews to guarantee the quality, but you have to fight the trend towards unbundling. Consumers will pay for reviews and recommendations, but only the ones they want when they want them. It’s tough to generate adequate revenue on that basis.

 Online referral fees are an inherently dicey business because it’s too hard to mask the name of the manufacturer, and there are far too many sellers, all a click away. You can make it work if you have gobs of traffic, and this is even a better business if you can leverage your existing traffic and not start from scratch.

 If I was trying to build a “best guide” site, I’d select Wirecutter as my starting point. It has the benefit of offering true product research without the huge testing costs incurred by Consumer Reports. It totally controls both the research process and resulting recommendations. It can leverage the brand and traffic of its parent, the New York Times. What would I change? First, I’d see if I could sell recommendations on an a la carte basis. Buying a dishwasher? Then buy our dishwasher reviews. I might also be able to generate some additional revenue from national retailers or manufacturers who could offer special deals along with the recommendations, though I would need to be careful to make it clear nobody had paid for a preferential rating. I’d ditch the referral fee model because it’s catnip for free-riders. Finally, I’d wrap the New York Timesbrand more aggressively around Wirecutter to reinforce the quality of the recommendations. I understand why the New York Times is moving cautiously here, but at some point, if you want to be in this business, you need to be in this business. You can’t hold it at arms-length. 

If you’re considering getting into the business, leverage your strengths, choose the right content and business model, and plan for the worst and hope for the best, because as of yet, there is no clear pathway to success in this huge and tantalizing area.

AI in Action

Two well-known and highly successful data producers, Morningstar and Spiceworks, have both just announced new capabilities built on artificial intelligence (AI) technology. 

Artificial Intelligence is a much-abused umbrella term for a number of distinctive technologies. Speaking very generally, the power of AI initially came from sheer computer processing power. Consider how early AI was applied to the game of chess. The “AI advantage” came from the ability to quickly assess every possible combination of moves and likely responses, as well as having access to a library of all the best moves of the world’s best chess players. It was a brute force approach, and it worked.

Machine learning is a more nuanced approach to AI where the system is fed both large amounts of raw data and examples of desirable outcomes. The software actually learns from these examples and is able to generate successful outcomes of its own using the raw data it is supplied. 

There’s more, much more, to AI, but the power and potential is clear.

So how are data producers using AI? In the case of Morningstar, it has partnered with a company called Mercer to create a huge pool of quantitative and qualitative data, to help investment advisors make smarter decisions for their clients. The application of AI here is to create what is essentially a next generation search engine that moves far beyond keyword searching to make powerful connections between disparate collections of data to identify not only the most relevant results, but to pull meaning out of those search results as well.

 At Spiceworks (a 2010 Model of Excellence), AI is powering two uses. The first is also a supercharged search function, designed to make it easier for IT buyers to more quickly access relevant buying information, something that is particularly important in an industry with so much volatility and change.

Spiceworks is also using AI to power a sell-side application that ingests the billions of data signals created on the Spiceworks platform each day to help marketers better target in-market buyers of specific products and services.

As the data business has evolved from offering fast access to the most data to fast access to the most relevant data, AI looks to play an increasingly important and central role. These two industry innovators, both past Models of Excellence m are blazing the trail for the rest of us, and they are well worth watching to see how their integration of AI into their businesses evolves over time.

For reference:

Spiceworks Model of Excellence profile
Morningstar Model of Excellence Profile

LinkedIn: A D&B For People?

I joined LinkedIn in 2004. I didn’t discover LinkedIn on my own; like many of you, I received an invitation to connect with someone already on LinkedIn, and this required me to create a profile. I did, and became part of what I still believe is one of the most remarkable contributory databases ever created.

Those of you who remember LinkedIn in its early days (it was one of our Models of Excellence in 2004), remember its original premise: making connections – the concept of “six degrees of separation” brought to life. With LinkedIn, you would be able to contact anyone by leveraging “friend of a friend” connections.

It was an original idea, and a nifty piece of programming, but it proved hard to monetize. The key problem is that the people most interested in the idea of contacting someone three hops removed from them were salespeople. People proved remarkably resistant to helping strangers access their friends to make sales pitches. LinkedIn tried all sorts of clever tweaks, but there clearly wasn’t a business opportunity in this approach.

What saved LinkedIn in this early phase was a pivot to selling database access to recruiters. A database this big, deep and current was an obvious winner and it generated significant revenue. But there are ultimately only so many recruiters and large employers to sell to, and that was a problem for LinkedIn, whose ambitions had always been huge.

Where things got off the tracks for LinkedIn was the rise of Facebook, Twitter and the other social networks. Superficially, LinkedIn looked like a B2B social network, and LinkedIn was under tremendous pressure to accept this characterization, because it did wonders for both its profile and its valuation. LinkedIn created a Twitter-like newsfeed (albeit one without character limits), and invested massive resources to promote it. Did it work? My sense is that it didn’t. I never go into LinkedIn with the goal of reading my news feed, and I have the same complaint about it as I have about Twitter: it’s a massive, relentless steam of unorganized content, very little of which is original, and very little of which is useful. 

Today, LinkedIn to me is an endless stream of connection requests from strangers who want to sell me something. LinkedIn today is regular emails reminding me of birthdays of people I barely know because I, like everyone else, have been remarkably undisciplined about accepting new connection requests over the years. LinkedIn is also just one more content dump that I barely glance at, and it’s less and less useful as a database as both its data and search tools are increasingly restricted in order to incent me to become a paid subscriber.

Am I predicting the demise of LinkedIn? Absolutely not! What LinkedIn needs now is another pivot, back to its database roots. It needs to back away from its social media framing, and think of itself more like a Dun & Bradstreet for people. LinkedIn has to use its proven creativity and the resources of its parent to embed itself so deeply into the fabric of business that one’s career is dependent on a current LinkedIn profile. LinkedIn should create tools for HR departments to access and leverage all the structured content in the LinkedIn database so that they will in turn insist on a LinkedIn profile from all candidates and employees. Resurrect the idea of serving as the internal company directory for companies (and deeply integrate it into Microsoft network management tools). Most exciting of all to me is the opportunity to leverage LinkedIn data within Outlook for filtering and prioritizing email – big opportunities that go far beyond the baby steps we’ve seen so far.

I think LinkedIn’s future is bright indeed, but it depends on management focusing on its remarkable data trove, rather than being a Facebook for business. 

Crossbeam’s Mission Impossible

I write often about the opportunity for data companies to operate as central information exchanges because they have a central position in their markets, and this neutral market position makes them trustworthy.

Lots of sensitive market information gets exchanged through central data hubs. Companies routinely exchange credit data, pricing data, business metrics and much more. They do this because they know the data they submit will only be released in aggregate or anonymized form. As importantly, they do this because they need the answers that only data exchanges can provide.

This is why I got excited when I heard about a stealthy start-up called Crossbeam. Crossbeam wants to build a database that consists of company customer lists. Yes, they are asking companies to upload their entire customer files to the Crossbeam database!

Mission impossible? Not at all. Consider when companies discuss merging. One big, burning question is always how much customer overlap there is between the two companies. Even in merger situations, companies are reluctant to hand over their crown jewels to what often is a direct competitor. Crossbeam is offering to compare those two customer files on a confidential basis and report out the results, something that demands a neutral market position, and the trust that goes along with it.

You might think that this idea, while interesting, isn’t all that big. Think again. Crossbeam aims to be a business development tool for those in charge of partnering and strategic alliances. Using Crossbeam, a partnership manager can easily search out companies with a large overlap in customers – almost always the key to a successful partnership or business alliance. It’s an efficient, quantitative way to take the guesswork out of developing alliances, affiliates and business partnerships, because you know in advance you are selling to the same customers they are.

Crossbeam never releases customer data of course. It simply flags companies where there is a large overlap between your customer file and theirs. This is a wonderful example of the distilled magic of the central information exchange: companies contribute data that they would ordinarily not share because it provides back information they cannot otherwise get.

In the course of helping to accelerate business partnering, the other data and business insights that Crossbeam will be able to access are potentially staggering. Of course, Crossbeam also has the challenge of protecting all this sensitive data, making sure it can’t be used in unintended ways, and making sure it doesn’t kill the golden goose by mining all the data in its possession too aggressively. Still, those are manageable issues, and all part of the mission Crossbeam has chosen to accept!

Being in the Middle of a New Data Product

I’ve written before about the application model called the “Closed Data Pool.” In this model, companies (and many times they are competitors) contribute proprietary data to a central, neutral data company. The data company aggregates the data and sells aggregate views of the data back to the very companies that contributed it. Madness you say? Not really, because these companies get great benefit from those aggregated views (think market share, average pricing and other vital business metrics). It’s the neutral, trusted data provider in the middle who makes it possible. 

But there is another twist on the closed data pool that represents an even more profitable business for the data provider in the middle. Consider a company called The Work Number.

The Work Number came into being because a lot of credit grantors need to be able to quickly verify employment status and income. At the same time, companies hated getting an endless stream of calls from creditors seeking to verify employment data. The Work Number came up with an ingenious solution. It went to big companies and said that they could outsource all these nuisance calls to The Work Number. All the company had to do was supply a feed of its payroll data. 

The Work Number then went to major credit grantors such as banks and said that instead of those painful verification calls they were making, credit grantors could just do a lookup on The Work Number website and instantaneously get the exact data they needed.

The best part? The Work Number was able to charge credit grantors for access to the database because of the big productivity gains it offered. But The Work Number was also able to charge the companies supplyingthe data because it increased their productivity as well by eliminating all these annoying verification calls. Yes, The Work Number charges both to collect the data and provide access to it!

If this sounds like an interesting but one-off opportunity to you, it’s not. Opportunities exist in vertical markets as well. Consider National Student Clearinghouse, which does the same thing as The Work Number, only with college transcripts.

Is there an opportunity in your market? Look for areas where relatively important or high-value information is being exchanged by phone or one-off emails or even by fax. If the information exchange constitutes a serious pain point or productivity drag for either or both parties, you’ve probably got a new data product.