NOTE: This article was first published by EFY Magazine in print format. Click here to download a pdf copy.
With this rather mouthful title, let me assert that there is a context to it.
Do you remember the famous quote of Henry Ford? “If I have asked people what they wanted, they would have said faster horses.” But think about it. Why should that be the case? That is when you start to see the limitations of incremental innovation and ideas around it.
In comparison, radical and transformative innovations are miles apart, especially when comparing their outcomes.
Moreover, the process of selling these (radical or transformative) innovations is significantly different. Incremental innovation is often the “pull-then-push” type, whereas radical and transformative innovations are exactly the opposite.
They need the “push-then-pull” methodology of selling the concept. It makes all the difference, along with my assertion and the title of this article. Let me explain.
Push-pull versus pull-push
With the car design Ford was working on, there was absolutely no comparison with whatever was available in the market. Therefore, there was no data to compare with and prove the effectiveness of the car. Comparing it with horses would have been meaningless and would not have yielded much. So, they “pushed” the concept in the market as a radical change. When people became accustomed to using cars, it was easier to “pull” their ideas, feedback, and suggestions and improve the car model. Accordingly, future models of the cars were easier to sell and compare with older cars.
“Push-then-pull,” therefore, has helped in this scenario.
Incremental innovation, on the other hand, is essentially “good getting better.” And that means the basis for comparison is available. Most changes are based on customer or market feedback. You can always put forth data-centric comparisons to convince customers to buy newer versions of your product.
Now that it is clear when you can or cannot use data let us see other issues with the data first approach in innovation.
Problems with the data-first approach
The fundamental requirement for any data is its existence. It means the data is only available from the past. A product or service must exist to give you data of any kind.
We often have seen organizations asking leaders to become data-driven or analytical. The problem with that approach is, you as a leader cannot foresee the future with crutches of data.
Those who wait for the data always lag and get blamed for being too late and slow. If you were to When you investigate the future, there is no data. There cannot be any. But in the absence of that data, you must have a good theory or hypothesis. So, when you use the hypothesis-based thinking lens and then use foresight, things become crystal clear.
Innovation is all about adapting to the ever-changing future. For organizations to survive, they need to adapt and hence innovate continuously.
Innovation is all about adapting to the ever-changing future.
But then, innovating for the future means you have no data to make any decisions. All you have is a good hypothesis. It is one of the strongest reasons why the data-first approach in innovation often feels good but makes you lag in the process.
Most innovations have no past, so there is no data available, and this is the basic problem with the data-first approach in innovation.
The sin of a business case
Another issue with data in innovation is business-case requirements. If your organization is one of those red-taped systems, then it is quite likely that you ask for a business case for innovation projects. The problem? There is no data to use in business case preparation unless you are making a case for incremental innovation.
If you are talking radical or transformative innovation, there exists no data. You are literally talking about the future. No matter how you look at it, you cannot plug any numbers into the business case spreadsheet. If forced, many would insert hypothetical numbers or best-guess scenarios. That brings us back to the hypothesis-driven approach.
There is no data to use in business case preparation unless you are making a case for incremental innovation.
If you are lucky, business case approvers will accept your hypothetical and best guess numbers. If not, they will ask you to justify and support those numbers. That is when you end up chasing a mirage. Trying to prove hypothetical numbers with a set of hypothetical scenarios and hypothetical market assumptions. The tower of assumptions builds rapidly until everyone kind of knows that all the business case is nothing but pages and pages of junk.
Suppose that gets rejected, congratulations on finding another method of wasting time and resources. However, if it is accepted, then I would argue, why not go with hypothesis-driven methodology in the first place?
Is there a solution?
So, not the question is – is there a way out? Is it possible to stay data-driven and innovative at the same time?
In black-and-white terms, no! You cannot have both. However, if you were to be realistic, there is a sort of middle ground.
If you were to stay a purely data-based decision-maker, you would always keep looking back in time. It will impair your forward-thinking. But it is also true that mere theory or hypothesis-based judgment could be difficult to convince others and implement with the necessary confidence.
You cannot be data-driven and innovative at the same time.
The solution is to design quick experiments based on your theory and gather small enough data to take or confirm that theory. Once you have enough data to give you and others reasonable confidence, then you can take the next step.
By the way, some of you may recognize it as a mixture of DoE (design of experiments) and hypothesis testing.
At first, you would lay out a hypothesis. Secondly, you would design a small enough experiment to conduct and validate or invalidate the hypothesis. With that experiment, you can obtain sufficient data to prove your point and make an informed decision. Now the key question is – how do you develop a good, reliable, and usable hypothesis?
A good, reliable, and usable hypothesis
In my previous article, I mentioned foresight skills, experimentation culture, and their importance. That is where it all can come in handy.
If you want to develop a good and reliable hypothesis that can be tested in the shortest possible time and with the least possible investment, you need reasonable foresight skills.
Additionally, you also need the ability to cross-reference multiple trends. Many polymathic and cross-functional experts can understand and gauge multiple trends efficiently and effectively. When I say multiple trends, I am referring to cross-industry trends and trends from sideways, i.e., those from industries that cater to the same or similar customer base as yours. Once these trends are identified and pieced together, you can see what that means for your industry or organization.
A hypothesis based on such information and analysis is more potent than something based on a hunch or leading questions.
The point is
While hypothesis testing is the key to generate new and relevant data, experimentation culture and discipline are essential to see that through its full lifecycle. Experimentation culture is difficult to build. But once it is there, it goes a long way in the innovation journey.
When it comes to innovation, the data is useless. Instead, support your innovation with a hypothesis-driven approach and experimentation culture. It will be the best transformation you will make in your organization. Not only to sustain but also to stay ahead in this VUCA market.
When it comes to innovation, the data is useless. Instead, support your innovation with a hypothesis-driven approach and experimentation culture.