By the standards of six years ago, those things are free now. Of course, it’s never free as in beer, because in fact we’re now using all the bandwidth, storage and CPU power we thought we’d never need, doing things a lot of us though we’d never be able to do in real life, like watch streaming live TV on our laptops (and our phones).

Today, I listened to Jeff Hawkins’ talk at ETech a year ago, Why Can’t a Computer Be More Like A Brain?, where he announced the Numenta Platform for Intelligent Computing. I think the idea now is, “If you’re not planning for how you’ll use human-like pattern recognition in your system, you’re not planning.”

What are the possibilities? Jeff listed a few, but there will be lots we’re not even thinking about now (or think are impossible – like we thought watching TV on mobile phones would be). It’s “human-like,” but not human. It can do things we don’t have time or patience for (just like a computer can do arithmetic all day that would drive us bonkers, even if we could do it as fast). And it can do pattern matching on patterns we can’t see – whether those are in infrared, or in huge piles of data that we just don’t have the sensory apparatus for.

Here’s an example. I have a product that manages requirements and their relationships to customers and customer requests (among other things). I talk regularly to customers, and take notes on these conversations. They often mention desires or usage scenarios that are outside the scope of the topic of the conversation at the time, and that aren’t seem related to the product planning I’m doing. Two years later, when I’m actually considering adding features related to those comments, I don’t necessarily remember the conversation. Or if I remember it, I don’t remember which customer made the comment. Currently I use Google Desktop to help me find the appropriate notes, but it’s a pretty rough approach – I have to use my best guess at the appropriate key words, and I end up spending a heck of a lot of time combing through my notes again and again. I’d love to have a human-like agent do this trolling for me – on a weekly basis, review all my old notes against my current requirements and tell me where there are overlaps. And while we’re at it, maybe it can read my blogroll and find related articles I could be referring to.

I know there are a few products out there that (claim to) do some of this already, and perhaps they work (I haven’t tested them). If so, they are the vanguard of this new set of capabilities, and in a few years they will have been overshadowed by realities we’re only dreaming about now.

How are you planning for the technical capabilities – human-like pattern matching or whatever it might be – that we’ll have at our command in five or ten years?

Doug Hall, unlike most business consultants, makes quantitative claims in his book Jump Start Your Business Brain. According to his research, he can predict the effect of certain basic marketing techniques on the success of a business. For example, he says that businesses that articulate “an overt benefit, a dramatic difference, and a real reason to believe” have about a 47% chance of success, while having only two of the three gives about a 30% chance of success.

Today I just ran across another interesting set of quantitative data about success factors, from the Journal of Product Innovation Management, Success Factors in New Ventures: A Meta-analysis, by Song, Podoynitsyna, et al. (I believe the link I provided is free, although most issues of this journal are behind a paywall.)

Based on a meta-analysis of a number of other studies about successful and non-successful new ventures, the authors determined that eight factors – out of about 20 candidate factors – were the best predictors of success. The candidate factors included items like:

• A low-cost strategy
• Prior startup experience
• Patent protection

Interestingly, the eight success factors did not include prior startup experience, or what they call “product innovation”. Instead, the factors were:

1. supply chain integration
2. market scope
3. firm age
4. size of founding team
5. financial resources
6. founders’ marketing experience
7. founders’ industry experience
8. existence of patent protection

The article is a bit vague on some of these – for example, it doesn’t suggest the ideal size of founding team, just that it’s a success factor. And isn’t is obvious that a successful new venture will typically be in business longer than an unsuccessful one? As the authors acknowledge, to a large degree the article leaves more questions open than it answers.

So, getting back to the original question, if you are starting a new venture, or thinking about doing so, will you pay attention to the research and make sure you have all your success factors in a row? (Or that you have an overt benefit, dramatic difference, and real reason to believe?)