The power of statistics and why the “why” doesn't matter
What do books like ‘Freakonomics‘ and ‘The Long Tail‘ (which I haven’t read yet) teach us? That statistics is an immensely powerful tool for getting at the underlying behaviour of people. It’s rationalised observation really, for books like ‘Blink‘ teach us the same thing… minus some of the numbers perhaps.
In “real life,” so far I have found out several things through statistical collection and analysis: what percentage of businesses use what type of hardware set-up (in the 90s, so no longer relevant); what percentage of people is interested in luxury leisure services; and what percentage of businesses use what type of financing. I also found out many of the “why’s,” though in statistics, numbers speak louder than words. For example: An entrepreneur can tell me he doesn’t trust banks, business angels, and venture capitalists. But on the aggregate, I also see that his ambition-level is far lower than those that do “trust” and use these institutions. The more data you can collect, the less “talking” needs to be done.
A friend of mine told me about some of the statistical methods he uses on his successful e-commerce site, and I’m sure many others do too. The method is called “A/B split testing” and works as follows. You create several copies of pages, all with different looks (fonts, colours, levels of “noise,” etc.) and you see which one is more likely to turn the visitor into a buyer. Very simple to do on a web-site, you can profile by region, browser, or just pick every 2nd/3rd/…/50th visitor, and very quickly, after a 100 hits or so, you see what works and what doesn’t. And you don’t need to concern yourself with the “why.” On a more negative note, A/B is also the reason why you also need to clear your cookies when shopping for your flight. More on A/B here.
Another interesting method is that employed by Amazon, Pandora, and Last.fm (I think). That of “collaborative filtering” Essentially, it allows you to quickly check what other products/services a user/listener/buyer of a certain product/service will like, based on the past behaviour of other users, and suggest it in real-time. More on that method here.
The advantage of (internet-)software is that everything is measurable and easily changed to match tastes and behaviour. With what I call “hard” businesses, those that operate with tangible things, it is far less easy to change and adapt, though companies like Dell or Toyota are certainly a forerunner there. The problem can also be observed with software-vendors. If they provide a static piece of software, which does not send back user-behaviour, they will have much more difficulty in adapting future products to users’ needs. That’s where software as a service (SaaS), which Jeremy has been raving so much about, has a real advantage, and will, technology-permitting, eventually replace the (crappy) desktop software we are so used to.
Imagine the kind of statistics Last.fm is collecting on music-usage; Facebook on social trends; Linkedin on business-trends; and Twitter on god-knows-what. And imagine both what revenue-streams can come from it and how these services can be improved based on that data. That is the power of statistics. But also don’t forget that, as these methods become and are already part of hard businesses, improvements happen there too. With it, you will get a better cup of coffee from Starbucks, a better burger from McDonalds (I’m sceptical), more logical product layouts in your local supermarket, and a better car from Mercedes. With the internet we are only scraping the surface—process innovation. The truly radical stuff will happen in the world outside your computer-room.
Vincent is a co-author on Tech IT Easy, and prefers getting shot to working in Excel. While he’s loading his gun, you can find out more about him on this blog’s initial announcement or on his site.
LikeRelated posts:
Books, Business Process Management, Business intelligence, Business strategy, CIO, CRM, ERP, Economics, Entrepreneurship, Finance, Globalization, Internet, Networks, Organization, Security, Software, Supply Chain, USA, Venture Capital, innovation, marketing, media, open-source, operations, retail, user-generated content
[...] I thought. A smart entrepreneur at the last incubator I worked at, suggested that it is a form of A/B marketing. Coke is marketing the silver Diet-brand at women and Zero at men. Made sense, I thought: black is [...]
[...] and psychographic data and I don’t want to cheapen that. I’ve written before about how statistics only matter as much as where your data comes from. But I’ve also written that triangulation is a large part of my research philosophy, which [...]