“The future belongs to the companies and people that turn data into products.”
– Mike Loukides, O’Reilly Radar, June 2010.
It’s not as though product development alone is easy. Even when armed with the tears of hundreds of thousands of developers as a reference guide, you’re bound to contribute several of your own to the corpus.
Google Search is the most familiar example of turning data into product. And it has a few effects you know about.
We all know several dozens of people who can explain SEO in high detail. I know only four people who can describe the mechanic behind PageRank in high fidelity – and they no longer practice SEO at all (or exclusively) anymore. Yet, users of Google don’t care how relevant results are delivered. They just care that relevant results are delivered. Marketers don’t care how the algorithm works. They just want to be ranked first.
How do we make the data dissolve into the experience?
That’s the question. There’s no single answer.
Just lots of social and physical technologies and very open surface of possibilities.
There’s an intersection at aesthetic, execution, and market. (I’m not the first to note this.)
It’s is a lot like solving a linear system of equations while painting the Mona Lisa while shouting ‘just set it and forget it’. It’s relatively difficult for an organization to perform all three functions, and have them balanced both in absolute terms and temporally.
To expand on that point – everybody is subject to the tyranny of the Production Possibility Frontier. Resources are limited. Not everybody values aesthetic. Others do. Choices have to be made. And hidden in the nooks of those degrees of freedom is error.
Moreover, firms always confront the need for short-run profitability and long-run sustainability. It’s very easy to incur a massive amount of technical debt when you’re trying to keep the lights on and meet payroll. It’s very hard to say no in the short run for long-run sustainability. Such tradeoffs are accentuated by the stock market in particular, and within startups in general.
It’s because it’s risky that the rewards are huge
Turning data into product is hugely rewarding. And can be hazardous.
Ultimately, hopefully, it’s through far better analytics design and data science that we can make things better and solve major problems.