More information about their users is being made public by social networks, and the tools to work with massive data sets are getting cheaper. A lot of companies are trying to figure out ways to make money from these two trends, so I wanted to give an overview of some practical revenue streams either potential customers have asked me about, or that I’ve seen competitors in this space using.
I’ll focus on cataloging what I’ve seen, rather than digging into the ethical debates that some of the applications raise. It’s important to understand what’s possible and happening right now so we can have a meaningful argument about what the rules should be.
Improved search results
I got started working this area when one of my products needed to match up email contacts with their social network accounts. I wanted to automate the process of Googling a person’s name when you first exchange emails with them, and so my first thought was using an API to one of the existing search engines. Unfortunately Google actually blocks most Facebook results from their API, and Bing and Yahoo have very spotty coverage, missing a lot of users. That led me to write my own simple crawler to catalog Facebook profiles myself just to do those name/location lookups.
I later realized how much other fascinating information was available in those public profiles, but I ran across several smaller search startups willing to pay for just the information matching a name and location to a profile. It’s definitely not a massive market, but there’s money to be made, and since it’s identical to Google’s functionality it doesn’t raise many ethical questions.
Better targeting for direct email marketing
This is one of the least known but most lucrative uses for public profile data. A company with a large email list will run all the addresses through a lookup service that gives them a list of their customer’s social network accounts. That knowledge can then be used in all sorts of ways to target customers, from sending special Twitter offers only to people you know are on the service, to pulling detailed location information for localized campaigns. I’ve even heard rumors of a Vegas casino that upgrades guests to suites if it spots they have a lot of Twitter followers! If you want to see something like this in action, Flowtown offers a lot of these features thanks to the Rapleaf API.
Any business-to-business use of our personal information is inherently a bit creepy, but direct marketing firms have been doing similar analysis for decades using traditional data sources like magazine subscriber surveys, so this seems fairly uncontroversial.
The most direct link between information and money is in the financial world. For example, if you can detect that a brand is becoming popular before anyone else, you can buy shares in that firm and benefit from the price rise when that success shows up in their profits.
Hedge funds have been using non-traditional metrics for years, doing things like running their own focus groups and opinion polls, but recently there’s been a lot of interest in the flood of information flowing through social networks. Twitter is the most obvious example of a data source, but the audience is both small and heavily skewed towards geeks, making it hard to pull out meaningful information. My feeling is that this will only become really useful once mass-market data is more available. Imagine being able to spot companies where a lot of employees have recently updated their LinkedIn profiles, for an early warning of firms in trouble.
One challenge to this approach is that you need some kind of historical baseline to compare current figures against, to tell if they represent something real or are just noise. That’s a barrier because it means you need to have been collecting the data for some time before it starts to become valuable to hedge funds.
Again this seems to be an extension of existing processes, just slotting in public profiles as a new data source, so it’s hard to see what new ethical ground is being broken.
General marketing intelligence
Marketing managers for big brands constantly have to make decisions about how to allocate their resources and craft their messages, and they need the right information to make good choices. My FanPageAnalytics project was aimed at those people, giving them unique information about who their and their competitor’s fans were, what else they were fans of and where they lived.
There’s definitely money to be made in this area, but brand managers are busy and non-technical, so they require something very targeted to their needs and don’t seek out new solutions. My feeling is that makes the leaders like Radian6 hard to beat even as the technology changes, because they have built relationships with most brand managers that gives them a defensible distribution channel.
Reaching influencers for PR purposes
Public relations people want to persuade influential people to write about their clients. One problem is that they may not know who the influential people are in a given area, or they may know but be unable to reach them effectively. Ever since I did my Twitter visualization, I’ve been asked about this use case repeatedly. The holy grail is being able to enter a topic, see who the most influential people are *and* who they are influenced by. Very often there are lesser-known specialists who are read by more popular writers for story ideas, and those sources may be an easier route to getting your stories to those mainstream influencers than approaching them directly.
This is one of the few areas where Twitter’s comparatively small user base is not a issue since most people who broadcast to an audience are using the service as another channel. Using information from other networks to reach them can feel like stalking though, so I expect that the increasing availability of public data will be countered by celebrities locking down their privacy settings.
Weak relationships, people you met once at a trade show, are surprisingly effective when it comes to getting a job. Recruiters contribute a massive chunk of LinkedIn’s revenue, and people are largely happy to see their resumes and connections shared for job-hunting purposes. It’s a pretty sweet position for LinkedIn, since it makes them the only customer-facing business that’s able to sell their users’ private data to other companies without fear of a backlash. It’s an area that could be helped by the new flood of public profile data too, especially if you can get some information about people’s connections. I’ve run across two different firms who’ve tapped into their employees’ friends networks on Facebook and Twitter to help fill positions, and I imagine there has to be a lot more innovation coming in this area.