## The 3rd Industrial Revolution: Considerations For Developing An Innovation Factory

Christian Blumhoff

In this series, we are discussing the 3rd industrial revolution and its impact on information practices.  This blog is the first of three parts for best practices for developing an innovation factory.

Personally, I like a good puzzle. I am not particularly into Mensa, but their puzzle from the 9th of July 2015 was interesting for two reasons. For one, I actually didn’t find the solution. But for another, once I looked at the solution, I had to admit to myself that I had been narrow-minded.

You can have a go at it yourself here:

Question: What number should replace the question mark in the grid?

I won’t make you read to the end of the article to find out what the solution is, but maybe I can illustrate why I was so disappointed in myself. I considered myself to be a reasonable, intelligent, and outward-looking person. So when I started to look at the puzzle, I was fairly it is math-related. Most are. So going with that probability, I pretty quickly saw that 8–2 is 6, 8–5 is 3, and 8–1 is 7. But this did not hold true for other areas (2–1 and 2–6 are not 7). I also noticed that 8 is larger (>) than 5 >3 >2 >0. However, I kept running in circles after that.

The solution is more holistic. It forces us to look at the data in larger chunks. Subtract each row from the one above, to get the one below. By deduction, the missing number is a three. I could of course make excuses. The puzzle is well-designed The 5-by-5 grid is symmetrical and drives the mind to subdivide. Resolving the pattern requires us to start at the right of each line, which is counterintuitive to Western culture. But the truth is, I was looking too closely. I was focusing on breaking things down, instead of seeing the whole.

What really upset me was the fact that I had recently come across a very useful approach to all puzzles. A system that increases the odds to find approaches and answers to any question. It’s from MIT’s Allen Rabinovich.

Reforming his puzzle approach to a business context brought me to the following list of practices:

1. Any business problem comes with context. Context data is just as important as the actual data itself. Take, for example, a claims process in the insurance sector. The claim data itself is surrounded by metadata such as phone-call locations, duration, and language. It also comes with background information like weather conditions, who processed the claim, adjusters, and so on. It always makes sense for any process to evaluate the context and the data it brings. You should answer at this stage: “Is my data and context information complete and accurate?”
1. Spend time in discovery. Initial insights are often very valuable. It is no coincidence that trusted data discovery is a rapidly growing business intelligence practice. More importantly, make notes of your initial findings, share them, and discuss – but, don’t get too heavily invested. What you’re really looking for is patterns.
1. If you have a set of early insights that can be turned into quick wins, deliver them. For example, finding that your storage replenishment policy contradicts the personal bonus plan. One such instance in one-for-one replenishment in retail versus MBO on distribution costs for the warehouse manager. This will definitely lead to issues and should be aligned quickly. This is something that can be passed on immediately.
1. In business, we should not only look for business potential, but spend more time on finding white space. While it’s often easy to see that we sell less ice cream in winter, it’s much harder to find the opportunity such as Ben and Jerry’s holiday destination. Keeping an open mind helps us see the market opportunity. We often get stuck in our company’s internal wisdom that it can be hard to take the customer’s point of view. There are methods that help from many sources including design thinking.
1. We don’t like anomalies, so we have a tendency to explain them away. When we look for opportunity, it’s critical to pay attention to exceptions. Easier said than done, you say? Why not bring in automation at this point? Today’s predictive technologies are strong enough to automate many analytics steps.
1. Organizing is of critical importance. Collecting data is all well and good. However, a good taxonomy and a detailed information architecture brings not only rigor, but also flexibility. In addition, don’t be afraid to reorganise as your innovation grows. Arguably many an inspiration have come from trawling through data and knowledge repositories. (For some reason, I liked this book on the subject of trawling for inspiration).
1. It cannot be overstated (in line with bullets 2 and 6 of this list) that sharing and widening the net are of paramount importance. Just because you can’t see it, doesn’t mean it doesn’t exist. (For lateral thinkers, here is an easy puzzle: What common chemical substance can be represented here: H, I, J, K, L, M, N, O? Answer in a bit).
1. Avoid power thinking sessions. Interlace analysis with more mundane work-like modelling or publishing. In addition, drink plenty of water and exercise (walking is best – on why click here).
1. This last point is the most important. Have fun and commit. We’ve already discussed that failing is okay as long as it’s part of a processes to improve. Commitment to the process of innovation is critical. Own the failure. Don’t abdicate responsibility. Any member of the team should share the fruits of success and learn from failure. Commitment leads to driving for simplicity and perfection. In the words of Winston Churchill “I am easily satisfied with the very best.”

By combining these best-practice ideas with our innovation process (see graphic above), we start to see what it takes to create an innovation factory. In part two on this topic, we’ll look at an example of an actual innovation hothouse, compare it to Google, and come up with practical approaches to keeping results high and costs low.

Which leaves me with one final task. The solution to the brain teaser. The answer is of course H2O (H to O).

Sorry.

Past blogs in this series:

Still to come:

• Considerations for developing an Innovation Factory (Part 2 and 3)
• Simplify our concepts and measures of success
• Why and how an organization improves its memory

Connect with me on Twitter at @CBlumhoff.

For more on tools that boost business, see Networks that drive net worth.