How does a company know where to start to innovate?

Grant Leboff: When a company has got a product offering, or a variety of product offerings, and they want to make sure that they stay relevant, they are looking at these incremental changes. There are so many variables; tweaking the product, tweaking the positioning, tweaking the pricing – whatever it is – where do they start? How do they know where to start as a business?

Costas Papaikonomou That’s a good question. It’s probably very difficult to answer precisely, and maybe that’s one of the misleading things that people want the precise answer in order to get started. My guess is that it’s probably driven, to a degree, by culture. I know that with the companies we work with, who are comfortable with incremental innovation, we have a system for it. They will default almost along a couple of dimensions. One may be always around claims on pack, for example. That’s just how they bring new news to their consumers. So, a new claim means, probably, also something new to talk about, in that a new claim may mean a new technology, for example, or something done in packaging. But it’s always led by a new claim. Whereas for others the starting point may not be the claim that is on packaging, but it may mean that they know that every year we just want a tweak to the actual product inside. We just want to change the formula, we know that whatever we do we have to update the formula in some way which makes it more relevant, so then we start there.

The key, then, is that once you have a default position to start from, what you do next? And I guess that differs a little bit by what we see. Often there is some kind of research project being ignited which can maybe, overdo it a little because you can’t, typically, research your way to innovation. What you can do through research is set out your case and say; my current consumers may have evolved new interests, new needs, or may be there are new trends which we should be playing into, which will then inform of us how to make that little change I just mentioned. So how can a particular new need that we’ve seen amongst our target [audience] trigger a new benefit that we might be able to provide, which then leads to a new claim or a new technology? Am I making sense?

Grant Leboff: Yes. That’s interesting because one of the things I wanted to ask you is about data. We live in a world that is becoming increasingly data driven and, of course, analysing data and looking at correlations in marketplaces and other things can play an important part. Think about Amazon, who now get a third of all their orders from [customers] looking at data led recommendations from correlations. So, to what extent do you think data should be playing a role in innovation and to what extent, in your experience, are companies referring to data to give themselves an eye into where they should be going next?

Costas Papaikonomou A very wise man once said to me “I have been measuring my p***s for 30 years and I still can’t control it”, so there is a lot being measured for the sake of measuring. I think there are two types of data, in my experience. One, is the data which comes from observing. When you could say that in an environment like Amazon, just tracking what happens on their sites, you are basically tracking stuff that actually happens.

The other, you could say, source of relatively big data, is just doing lots of surveys. That’s a very different type of data, because surveys are completely dependent on the people who fill then in. And you probably also know that the people who fill in surveys are in no way representative of what actually happens and, therefore, you could say the data is already flawed before you do any work with it.

Again, looking at innovation and, may be, in the context of incremental innovation, working with lots of data has probably its most value. Personally I’m not really a data person, definitely not a big data person, because the problem is it’s like – to bring Mr Ford back – people will just ask for faster horses.

When you are looking for new things to do, if it’s only incremental, it will kind of work. Basically going out to see what people are doing and what they’re asking for. If you want a little bit more, dramatic innovation is basically going to be, probably, one or two interesting clues, maybe three or four, but not 50 or 60,000 clues. And then it’s much more the responsibility of, let’s say, the smartness of the person tasked with innovating, of picking up the right bits of information to make up their mind on what’s really the insight here, and what is really the opportunity for us as a business to go after, rather than just, say, hogging lots of data and hoping that you might find a pattern.

The misleading thing about data, particularly if there is a lot of it, is that you will always find a pattern. You can always, in hindsight, prove that whatever happened was already waiting there for you in the dataset and you just happened to have not seen it. If anything is an interesting paradox, the more data you gather, the higher the likelihood that there will be conflicting data in there, rather than it all pointing down one particular direction. I mean, if it were easy, everyone would be doing that and, of course, everyone would be using big data to come up with the most fantastic innovations. It’s not happening for that simple reason. The more data you have, the more likely it is that it will conflict and also, if you have a hunch on, let’s say, a particular need, or motivational opportunity, or whatever… if you just collect enough data, you’ll find the answer right there. I mean it’s always in hindsight in that sense.

It’s maybe not the most optimum answer you are looking for, but I think it’s an area in which you need to tread carefully. Again, if it’s about small tweaks to your products, to your offering, to your positioning, to whatever, then you can use the nudges that data will reveal – sort of, the little movements in what’s happening in the market – but if it’s a slightly more dramatic innovation that you think is necessary, then you probably need to dive deeper into more qualitative type input. You should really go into depth in what is really happening, rather than just putting big numbers together in a vat.

There may be small changes to the spoken word in this transcript in order to facilitate the readability of the written English