"There's a lot of cynicism about search," Stephen Brady told me when we first met in August last year. He was working on a new personal search utility, then called BitLev and now called Found, that, he said, would help people find all their stuff, no matter where it was.
People have tried this before, and most have failed. X1, Copernic, and even Google's own Desktop Search app never got mainstream traction. And then Microsoft and Apple built desktop search into their operating systems. They made desktop search work better, but not soon enough. Nowadays, people store so much personal data in the "cloud," that searching your own local computer, no matter how fast you can do it, kind of misses the point.
Found is a personal search engine that spans local and cloud-based sources to get you the file or data you're looking for, no matter where you've put it. It's the right idea. Does Found deliver?
From the user's perspective, it has great potential. The Mac-only (so far) app is as easy to get into as Spotlight (press the control key twice) and is smarter than Spotlight in how it displays results.
Found shows local hits first, and puts the most recent results first, not just the most "relevant." This is subtle, but key. When you're searching for a file from your personal data store, time-order helps you recognize very quickly if what you're seeing in the result list is what you're looking for. With personal files, you know what you want, as a rule; you just don't know where it is. When searching for data on Google, in contrast, you usually don't know exactly what you want, and PageRank-like algorithms are much more relevant. (Plus, the whole concept of link-based relevancy doesn't apply to personal data.)
Found also shows hits from your Dropbox account, and from any number of Google Docs accounts. It also searches attachments in Gmail.
The app displays key information to help you narrow down your search. Like Google, it displays results as you type, along with time of last access and other helpful info (file path for local files, subject line for e-mail files). It has a lightning-fast preview window. It will also let you drag files or data directly from preview into another app, like a Gmail window.
The beta I tried is very early, but promising. It desperately needs more data sources (Gmail messages, Google calendars, Evernote, Box, Facebook, Twitter, etc.) and it needs a mobile app. Both are planned.
As a business
The economics of search -- especially personal search -- are challenging. To upload and index all a user's data in a cloud service is expensive at scale. Found doesn't do it. As Brady, Found co-founder and CEO, says, "The moment you take the data center out of the equation, the costs are dramatically different."
Found indexes users' data, both local and cloud, on local machines. This dramatically lowers Brady's cost. It also makes searching lightning-fast.
The search index a user has is mirrored onto other machines with the same log-in, so the work doesn't have to be duplicated. I presume that's how the mobile versions will work, since we can't expect an iPad to hold an index of a user's entire data set in its memory.
Found is a free app but it will have paid services, perhaps to include the indexing of professional data sets like Salesforce. This is the same model of competitor Greplin (which I use and like a great deal) but, Brady says, "We can go longer with the free service since we're not sucking everyone's data into a data center."
I was curious why Brady chose to make the first desktop version of Found on the Mac and not Windows. It's a more "concentrated" audience of enthusiasts, he said, and easier to market to. Especially if he can get some promotion in the Mac App Store. He also said that at this early stage, "It's OK to have a smaller number of users."
I'm not switching yet
Found's first release is a good start. I do hope that it gets several more data sources, as a personal search tool that is missing any one source is just not good enough.
My personal search and cloud toolkit, by the way, includes Alfred (instead of Spotlight) for local search and Greplin for cloud search, with the Greplin custom search in Alfred. It's a functional but imperfect setup, which leads me to believe that there's opportunity for challengers, like Found.