Monthly Archives: April 2017

I Went Fishing And All I Caught Was A Red Herring

The annual GRIT report should be required reading for marketing researchers; it provides a current snapshot of what senior players in our business are feeling about things. That said the reader should take a big grain of salt to go along with the red herring that winds its way through the report. The red herring is innovation, described as a holy grail with magical powers that should be self-evident. It is my contention that innovation for the sake of innovation is foolish. It only appears to be effective because so many buyers are enthralled with all things new and shiny, rarely digging beyond the surface validity of a lot of new techniques. Innovation needs to improve the predictability and reliability of our research – it does not have to be new, flashy, sexy, or disruptive.

The “innovate or die” movement persists, even though research has shown this to be untrue (e.g. Getz and Robinson, 2003). Yes, I know about the buggy whip company who didn’t see cars coming and Netflix changed the game much to Blockbuster’s consternation. Amazon brought us a new way to shop and one day their financial performance might justify what they do and how they do it – or not. Walmart’s latest challenge will certainly put a dent in Amazon’s growth estimates. Now take a look at all the “innovative” tools marketing researchers have developed. Are there any that blow you away?

In the GRIT report, John McGarr makes a critical point. He says, “MR providers need to keep in mind that no client owes the industry a trial of new methods for the sake of being innovative.”  Some will argue that we should expect buyers to try new things when they are (sorry about this) faster, better, or cheaper. I’d take the position that faster is fine, as long as it is at least as good and has a similar value proposition. I’d take the position that better is always better, even if it is not as fast and not as cheap – you should pay more for a better answer. I’d take the position that cheaper alone almost always has a hidden cost, usually in bad design, sampling, or survey construction.

The marketing research industry needs innovation, but the innovation needs to be directed at solving research problems. Here’s an example – we have a problem with predicting new product performance, as good as many models are. When our new product failure rates are over [70%, 80%, 90% – pick your number] we clearly do not understand the shopper dynamics beyond basic trial and repeat analyses. When was the last time we saw a new or better way to do our new product forecasting that was validated?

Instead of solving research problems, we pretend that automation and DIY are the innovations most needed because they make the research process less expensive. These types of innovations are technical innovations, but not problem-solving innovations. We miss the point – it’s not the price tag that matters, it’s the quality of answers we are able to give our clients. Of course budget limitations play a part in purchasing decisions. But, and it’s a big but, a methodology that is meaningfully more accurate will always be worth the cost. Greg Archibald, in summing up the report, says, “Over the next few years, we are going to see a continued focus on improving tools and methodologies…” I’d respectfully disagree – I don’t think we, as an industry, are very focused on improving our tool kit but rather we are trying to come up with the next new sexy thing to sell.

Innovations such as automation and AI are great for the business of marketing research, but that only provides a trivial benefit for our end-client. They need us to do our job better, they will pay us to do that, and we need to innovate with that in mind.

Originally published on on 8 February 2017

Big Data…It is ALIVE

So help me God, I thought we had killed it. The idea that Big Data, in and of itself, was something to embrace, that is. For two years now, the discussion shifted from why you MUST have Big Data (and invest millions to have it) and hire a fleet of data scientists to analyze it (because it’s too complicated for the average researcher). Instead, we started talking about small data, which might be a piece of big data, and how to use it. In many forums, including Greenbook and ESOMAR, I argued that it’s not data that should drive marketing; it’s the needs of your product.

And then today, I pick up the November issue of Quirk’s (sorry Steve, I’m a month behind), and there’s the lead article telling us we have to be data-driven or we’re doomed to failure. The key points of Lawrence Cowan’s missive ( are:

  • Data is one of the most important aspects of achieving a competitive advantage.
  • The ultimate goal is to create a business where data is leveraged to create real value (as opposed to fake value, I guess).
  • Data is a basic requirement for business, not a cost item.
  • You need a culture of “data-driven-ness” where you have to promote, train, and enforce the use of data (I’m picturing the corporate data police state when I read this).

I beg to differ. A lot. With all due respect to my friends who sell data for a living, you all mostly do a great job and provide a useful product. But data is data and data is not going to save a bad product or a bad company. Such a focus on data strategy and a culture of “data-driven-ness” across the company, as the author suggests, diverts attention from what is really important for business to thrive.

What’s really important is understanding where your product fits in the universe. Okay – maybe not in the universe, but in the store where people buy it, as part of a category of similar products. In these days of dwindling research budgets, data acquisition needs to be a focused activity. Otherwise, you are left with mounds of unused or unusable data that is not getting you the information you need.

We get to this focus by having a “theory” about your brand. I put “theory” in quotes because it does not have to have all the formal aspects of a scientific theory; its one formal requirement is that it has to be true. This theory will tell you:

  • Why shoppers buy your brand.
  • Why shoppers don’t buy your brand.
  • How sensitive is your brand to various marketing activities.

It is that simple. Once you know the answers to these questions, your marketing is dramatically simplified and your energies can be focused elsewhere. You might let this product go on autopilot. Or you might focus on improvements targeted to non-buyers. Or you might try and come up with some creative marketing that hasn’t been tried before. But most important – you don’t have to focus on data every day. Your knowledge gaps will tell you what types of data you need – and in a multi-brand company, that’s likely to be different for each brand. Your research needs will be focused on testing ideas generated by what you know – and what you don’t know.

We are not advocating that companies should ignore data. But they should not be data-driven either. The data comes from research needs which come from information needs which come from the brand theory – not the other way around. Let’s kill this idea before it once again stalks the countryside.

Originally published on 11 January 2017