Could StackExchange become a better answer engine than True Knowledge?

Could StackExchange do better than TrueKnowledge and Wolfram Alpha at answering real-world questions?

TrueKnowledge is working to gather and structure all known information so that it can answer any question put to it. This is extremely ambitious and quite hard.

Stackexchange is a spin-off from StackOverflow - a Q&A site for developers that has become enormously popular very quickly (like a year or so). Developers ask a question, other developers answer it, and the community votes up the best answer. Submitting a good answer gets you reputational points. Having more points gives you more voting power. The best answers float to the top, which is why the site is so popular - developers know they are likely to get or find a good answer to their question.

The folks who built StackOverflow (Joel Spolsky and Jeff Attwood) now offer, through StackExchange, their core voting / reputation technology to anyone who wants to build a Q&A site. Like mothers, or business travellers.

TrueKnowledge and Stackexchange are both are aiming to make it easier for you to find an accurate answer to a real-world question. But they are using completely different approaches.

It seems to me that StackExchange sites could well build a useful knowledge base much quicker than TrueKnowledge or Wolfram Alpha could. It is hard enough for TrueKnowledge to answer questions like:

In which city would you find the Rialto?

But do people really need answers to "factoid" questions like that? Could TrueKnowledge ever manage to answer real-world questions like:

How can I detect which line I’m processing in Perl?

or

What to pack in your labour and hospital bags?

Even if TrueKnowledge eventually can answer questions like this, I think they will be too late. My money would be on StackExchange sites gathering and organising that knowledge faster than TrueKnowledge and Wolfram Alpha will.

37signals on "pre-revenue" valuations

From 37signals:

In order to increase the value of the company, 37signals has decided to stop generating revenues. “When it comes to valuation, making money is a real obstacle. Our profitability has been a real drag on our valuation,” said Mr. Fried. “Once you have profits, it’s impossible to just make stuff up."

Hilarious. Read the full post here.

Is silliness of name a predictor of financial success?

Why do so many webby startups have such silly names?

Look at the Seedcamp finalists, all these totally made-up names:

Brainient

Codility

Comufy

Erply

Joobili

Kukunu

TechCrunch50 also peppered with start-ups called things like Clixtr, lissn, yext, etc.

What's going on? What's wrong with real words? Its like the first dot-com boom all over again: boo, flooz, kozmo, kibu, roolz, etc.

Is silliness of name in some way (thought to be) a predictor of financial success? I think we should be told.

Selling to corporates (2)

Loving this post from Mark Suster comparing micro-businesses, SMEs and corporates to hunting bunnies, deer and elephants respectively.

John Crane toys - amazing service

A chap (think he said he was the sales director) from John Crane toys has just dropped by our house to deliver a new one of these rocking horses, because ours broke.

No charge, no fuss, all done because we phoned up to say the old one had broken (expecting maybe to get a replacement bit we could fit ourselves).

Looking at the new one, it is obvious that our old one broke because we removed the safety rails but not the backrest (which is held in place mainly by the rails - without the rails, the backrest only has a couple of screws holding it in place).

Had we read the instructions properly, the backrest wouldn't have broken. Was that our fault? Probably, but not according our man from John Crane who simply said the instructions obviously weren't clear enough and that he'd email "Mrs Crane" right away, she was en route to the factory in Thailand now, and she'd pick up the email when she landed, and they'd figure out a way to make it clearer in the instructions that the rails shouldn't be removed without also removing the backrest. That's pretty amazing service.

(Incidentally, though, making the instructions clearer probably isn't going to help. The user doesn't have the instructions. In this case, they probably really don't: You might use the instructions to assemble the rocking horse in the first place, but by the time your precious bundle is big enough not to need the safety rails, you will surely have discarded the instructions.

So John Crane could perhaps do this:

1. Make the rails and the removable backrest the same colour, different to other nearby bits of the horse. That'll make them feel like one part, not to be separated.

2. Make it impossible to separate the rails from the backrest until you have removed the entire assembly (rails + backrest) from the horse. You are unlikely then to attempt to re-fix the backrest on its own, it will be obvious there is nothing really supporting it.)

No one believes your service is perfect

From 37signals:

The lesson for anyone trying to pick up fans/customers: Don’t be afraid to reveal those little mistakes everyone faces.

Couldn't agree more. Glad someone shares my view about publishing negative comments from customers. We used to do that at 0800handyman, and pleased to see the new owners continue to do so. What other home maintenance service would have the comment

One blind was not fitted correctly

near the top of their customer feedback page?

Selling to corporates

From Signal vs Noise:

The problem with selling your product to big companies: Middle men. The people who buy your products aren’t the same ones who use your products

Couldn't agree more. I can't get that excited about selling to big businesses. I get much more buzz out of knowing we've made someone's life easier than I do out of landing a big contract. That's one reason I left ViaPost: ViaPost needed to focus on bigger, corporate accounts in order to make any money, and that just didn't excite me very much. I'd much rather have focussed on the SME and micro businesses, but that probably wouldn't have been the best strategy for the business.

The idea that someone will start using your product because you take them to lunch, keep calling them so you are "top of mind", send them a Christmas card, etc. just seems really odd to me, even though that's how all corporate sales work. I'd much rather build a good product / service, market it, and then let people decide for themselves whether they want use it or not.

Very naive, probably.

Game over for on-premise software?

Wow:

• 13% of small businesses planning to switch from Microsoft to Google Apps within 12 months.
• 62% would prefer their software to work "through a browser"

(from Accredited Supplier, via TechCrunch)

Internet OS

This is interesting - Google announced back in July that they are working on an internet operating system, but someone else is already well ahead of the game:

Jolicloud