Free VC-Backable Idea of the Day

I’m a fan of tchotkes, so here is one for readers: A free VC-backable idea.

It goes like this: Web sites are, in aggregate, as Google has ably demonstrated, outsourced unstructured databases. It is like having an agreement with some deranged person who contracts to enter immense amounts of data for you, so long as you can live with the corresponding illogic and chaos of the resulting database. But impose a little structure on that unstructured or semi-structured data … and, well, you’re going like Google.

The trouble with Google, however, is that it is trying to be all things to all people. It spends some time coming up with special-purpose searches (i.e., you can enter a ticker and get a stock, you can enter person’s name and city and get a phone number and address, etc.), but Google mostly sticks to the tried and true of simply using immense spidering horsepower to find stuff without too much regard for source or underlying structure.

But the fun thing is that there is lots of data out there that gets missed by that approach. For example, there is a venture-backed firm in Boston that has built a fine and valuable business by spidering only the “About Us” section of websites. In effect, it has built a dynamic database of the comings and goings of people by geography, industry, and position. That sort of data is immensely valuable to recruiters and others (e.g. “Get me a software sales exec in the Boston area with HRMS experience”).

There are many variants on that sort of thing, all semi-structured data that is out there, being managed by someone else, but that would be valuable if some structure were imposed.

So, I’ll give you one: The Careers section of company websites. But here is the twist: I’m not suggesting you spider the careers section of companies’ websites and then pour that data back into your own meta-recruiting site. That’s boring, consumer-centric and 1999. What I’d like to see instead is a service that spiders Career sections and builds a competitive intelligence service.

Want to be alerted if a competitor is upping (or reducing) their hiring of engineers? You’ll be fed an alert, likely via RSS, when it happens. Want to be alerted when a company starts looking for staff in new areas, whether it’s scientists with expertise in monoclonal antibodies or Linux programmers? You’ll be alerted.

I imagine companies, equity analysts, and industry analysts all as subscribers. They’re a valuable bunch, a group to whom many folks would like to cross-sell other data-centric products. So, you build it up to $25-million in sales, then sell it to Thomson for 4x. QED.


  1. The only problem w/ the approach is that it will drive more and more companies to advertise through those annoying recruiters you see e.g. in the Economist: “Our client, a large pharmaceutical company, is looking for Linux experts with background in monoclonal antibodies”. That trend will make it harder for companies to recruit based on their image and harder for employees to gauge the worth of a listing prima facie, just putting money into recruiting firms.

  2. David —
    No question that there is a potential free-rider problem here, as well as companies becoming more discrete about their hiring practises. But that said, people who believe that their careers sites aren’t already being scrutinized by competitors — or soon won’t be — are practising a dysfunctional variant of security through obscurity.
    At the same time, it’s kind of a trade-off. It is expensive using recruiters for those monoclonal Linux people, so I really wonder if companies won’t continue to post positions on their own sites, at least for quite a while longer — even if my hypothetical competitive intelligence services springs up.
    Either way, there is a market anomaly here that is worth exploiting: Some data is not being efficiently distributed to all interested parties — so I’d expect the anomaly to be exploited, at least until the profits go away ….

  3. Here’s another freebie:
    a machine that turns junk mail into toilet paper.

  4. A good idea, but sadly not VC-backable … 😉