link

Here's a mop-up-the-drool review of BlinkX, a search sidebar product. This is an example of contextual, implicit search, not unlike dashboard. It certainly seems useful to have relevant information automatically made available to you. But there are two (non-technical) challenges with bringing this kind of software to market. First, data off the public internet is usually not useful—and most people know to use Google when it can be. Secondly, there is a trust barrier that prevents users from opening their own data to such a product.

In the first case, an automatic sidebar might save you some typing, but if you want to do a Google search it's often just a text-selection and right-click away. However, when you're looking to get something done—rather than just be distracted by random stuff to read—you don't need constant google results. If you sell Widgets, you probably already know that the company you work for makes Widgets and that effective selling requires courtesy. You need to know much more specific stuff, especially from your past successes and mistakes. The stuff you need isn't on the public internet.

On the other hand, an individual or company would be profoundly naieve to open confidential data to an internet startup offering to index it for your convenience. That smells like spyware. After years of web-based email, even Google found that privacy norms weren't universally understood. And I wouldn't trust even Google with the contents of my hard drive. Dashboard, which is an open source application, overcomes these problems because it doesn't depend on any central entity, but I don't think it's anywhere near production-ready yet.

So, what about selling to companies? I think there is some potential here, but even a great contextual search application isn't going to sell on its own. Enterprise buyers will, however, see it as a major benefit if integrated into a broader knowledge management framework. Crucially, the search must acquire context from specific business applications and data, and provide results in useful, structured formats. For example, it should notice if you're entering a sales prospect form for General Motors and show you all previous sales attempts with GM, organized into sucesses and failures. It should show you current data on GM's profit in the location you are targeting. It should show you the latest intelligence on which executives to pitch to, along with their personal tastes. Then you know that your best bet is to offer John Doe in Dearborn a ride to golf in your CTS-V, but best wait until Q2 financials are done.

If I were developing a product like BlinkX, I'd start by partnering with enterprise KM (knowledge management) and CMS (content management) vendors and consultants. These markets have some special advantages. First, you can sell to end users and not just advertisers. Second, the business context can trump individual privacy concerns, at least at the corporate purchasing level. Finally, the ubiquitous search sidebar is a great motivator to get people into the habit of using and contributing to knowledge repositories. It's a natural symbiosis.