March 31, 2005
Infomediaries--Where Are They?
I have been thinking a lot about “infomediaries.” If you’re not familiar with the term, John Hagel first described it in a 1997 Harvard Business Review article The Coming Battle for Customer Information (with Rayport) and then fleshed out his vision in the 1999 book Net Worth (with Singer).
Infomediaries interpose themselves between marketers and consumers to facilitate better marketing matches. Consumers disclose their personal preferences to an infomediary, who can then offer marketers the ability to engage in highly targeted marketing without knowing consumers' personal identities. Further, infomediaries will use their aggregated consumer demand to cut consumer-favorable deals with marketers. To make this work, consumers must completely trust that infomediaries will respect their privacy and will not become a biased shill for marketers based on which marketer pays the infomediary the most.
From an academic’s perspective, I think infomediaries would substantially improve social welfare. Consumers get what they want—relevant and trustworthy marketing without sacrificing privacy; marketers get what they want—a cost-effective source of interested consumers; and infomediaries profit by taking cuts of the deal. Society wins due to lowered transaction/search costs and fewer marketing mismatches between consumers who don’t want the marketing and marketers who cannot target granularly enough.
Compare this with our current marketing environment, where consumers lack an easy one-stop way to disclose their preferences (and many consumers refuse to do so due to privacy fears). More regulated solutions of marketing communications have high transaction costs (for marketers, and sometimes for consumers too) and a high risk of Type I and Type II errors (i.e., relevant marketing is squashed; unwanted marketing is unregulated).
Despite all of these benefits, as far as I can tell, the infomediary industry has failed to materialize. In Feb. 1999, James Glave wrote a Wired News story called The Dawn of the Infomediary listing five companies trying to enter the infomediary business: Lumeria, PrivaSeek, InterOmni, @YourCommand, and PrivacyBank. On January 24, 2005, I visited the purported websites of all five infomediaries discussed in Glave’s article. Lumeria’s site still exists but appears not to have been updated since 2000. InterOmni was acquired by Lumeria in 1999. The PrivaSeek and @yourcommand domains appear to have lapsed and been reregistered by others. InfoSpace.com bought PrivacyBank in 2000; it is unclear what happened thereafter.
In other words, it appears that all of these infomediaries are out of the business. Also gone are the group buying sites (like Mercata and Accompany) that aggregated consumer interests to negotiate better deals with merchants.
We have some more success if we broaden our definition of infomediaries further. In some industry verticals, infomediary-like businesses have emerged, like LendingTree for loans and Autobytel for cars. However, to some extent, Autobytel act like messaging services—I submit my information, a message goes to interested dealers, then the dealers spam me directly (sometimes relentlessly). Rather than protecting my privacy (whatever that means), Autobytel just ratchet up the email volume. There is still value to consumers to messaging systems, but I don’t think they rise to the infomediary level. LendingTree actually makes offers, not just referrals. However, I'm not entirely clear how these offers are ordered.
We could also try to analogize the shopbots to infomediaries. Shopbots like Shopping.com, Shopzilla and PriceGrabber have survived the dot com crash and offer some infomediary-like services, such as organizing marketing information by product and pitting merchants against each other. However, shopbots do not personalize the offers based on a consumer’s preferences or try to act as a consumer agent; instead, like some industry vertical sites, shopbots view their role as referral services (i.e., send the consumers to the merchant and get out of the way). Further, merchant listings are generally presented based on merchant willingness-to-pay, so consumers may feel like shopbots put merchant interests ahead of their own.
Why haven’t infomediaries emerged? I am struggling to answer this question. Some of the possible theories I’ve come up with:
· Infomediaries do exist but I’m not defining the term expansively enough.
· Infomediaries cannot convince consumers that they are trustworthy. In my experience, my clients would routinely start out saying that they wanted to protect their customers’ privacy, but inevitably they would, over time, look for ways to monetize their customers’ information. Further, companies usually cater to those who pay the bills; so any infomediary will inevitably be tempted to put merchants’ interests over consumers.
· Consumers’ privacy concerns are not strong enough that they need infomediaries. The empirical evidence here is sharply split. Consumers routinely say that privacy concerns inhibit their online actions, but consumer behavior routinely belies this. There are plenty of good reasons to use an infomediary beyond privacy protection, but perhaps this motivation is not as strong as Hagel predicted.
· There is no viable profitable business here (i.e., the economics simply don’t work).
· There is a market failure that prevents companies from entering the market. If we could find a market failure, would this support government intervention to sponsor the creation/operation of one or more infomediaries?
As you can see, I’m stuck. I ask for your help, and I’m opening comments on this post. (Unfortunately, to prevent comment spam, registration is required—sorry). Why do you think infomediaries have not arisen?
Eric, your post reminds me of the problems faced by the B2B hubs in the late '90s. Although they were not completely analogous to infomediaries, one of the roles that they wanted to play was similar: interposing themselves between buyers and sellers, thus increasing costs, but creating more opportunities for exchange and reducing transaction costs. Some of the problems were: (1) They never worked. The technology remained vaporware, enriching only consultants. It was harder to make these things work than everybody thought. (2) Collective action problems / lack of network effects. They only were worth the trouble if a lot of players (both buyers and sellers) joined up. Everyone waited for everyone else to commit. (3) Distrust on the part of the sellers. Sure, it would be nice for buyers to set up a scenario where suppliers come in and bid against each other. But why should the sellers do it? Access to a lot of customers would motivate a seller, but the motivation for sellers to wait for customers to materialize exacerbates problem #2.
Fourth, and most important, the costs failed to outweigh the benefits. Like so many others in the late '90s, B2B hub and infomediary enthusiasts underestimated transaction and transition costs and overestimated benefits. The costs of B2B hubs and infomediaries include: (a) Extra step in transaction; (b) Learning how to use a new service; (c) Paying for the service (even free registration for an infomediary required filling out a lengthy form and, most likely, being bothered with e-mail); (d) Loss of control (not just a trust issue involving privacy, but a trust issue issue regarding competence/effectiveness--are you getting a better deal than you would find on your own?). There likely are other costs that I am missing. Every new cost added weighs heavily on a business model that relies on reducing transaction costs and exploiting network effects. These costs make it harder to convince people to come on board ("will I *really* come out ahead?"), which in turn keeps the business from getting off the ground (no network effects). Also, costs to your customers reduce savings, thus reducing what they are willing to pay for the service. B2B hubs faced this problem: They had a hard time convincing people of the value of paying even infinitesimal transaction fees.
The business can solve some of the cost and network effects problems by giving away its services at first. Such a strategy takes time and money and increases risk. Venture capitalists have learned the hard way (and perhaps "overlearned") to be skeptical of such a model. Moreover, as I noted in point (1), these things generally proved more costly and harder to create than anyone thought. The technology was not "there" yet.
Still, I was also "drinking the kool aid" (as we liked to say back then). My clients had some compelling pitches. The economics seemed to work. We were too optimistic. If the infomediary model made senses all, however, somebody may pick it up at a later time. Perhaps the technology will be more favorable. Perhaps the players will be ready now if they were not earlier. Perhaps the model will work better if an established business incorporates it into a wider business model (one thinks of how eBay and Amazon have gradually incorporated complementary business models).