Thinking in layers

The economist Milton Friedman once made the observation that no-one knows how to make a pencil. Human beings collectively obviously know how to make and manufacture pencils, but each component part is its own specialism. The end manufacturer does not know how to mine graphite, or how to grow the rubber for the eraser. A pencil, like almost all manufactured layers is a sum of specialized parts.

The same is true for many research projects. The consumer layer - what the customer buys and values is met from a chain of points of delivery. A sales specialist. A marketing specialist. The product designer. The technical team. The operations team that makes the manufacturing and logistics happen, and the finance team that assesses if the product is financially viable.

Each of the layers has different constraints and requirements. The consumer would like to pay less. The financial team would like to earn more. That means there are always tensions and compromises. Simply knowing what the customer wants is not enough as the business has to be able to deliver too. But also having the knowledge of how to make anything you want, does not tell you what you should make, and where the sweet-spot is for the business.

This is why good researchers reach out beyond the customer interface to ensure they think about the downstream implications of what the research could show. A segmentation study is not a survey with some cluster analysis.

Once a suitable set of clusters are found, they need to be targeted, managed and measured. If the segments are good enough, each individual segment needs its own resources. It could be a packaging redesign, or a different price or a new product add-on, or a dedicated segment manager.

The operational part of ‘doing a segmentation’ will require ways to ensure the segments can be researched and measured with enough depth that the product and marketing teams have clear reference points. What is good enough for a value segment, and what prices will they actual stand for a lower spec product?

The research needs to have hooks and answers to questions that only become relevant when the top-set of results are being implemented.

The layers also work in reverse.

Technology creates constraints that often mean a perfect answer is not possible, or where a customer’s demands have knock-on effects. The trifecta of faster, cheaper, better quality - which do you concentrate on?

For instance, a foldable screen for a mobile phone is an immensely challenging technical problem that then limits the ability to deliver a thin, lightweight device that may be dropped or that bangs around in a bag. Where are the compromises that are acceptable to get a go-to-market product?

The blue-sky idea has value for research, but unlike pure advertising which can promise the earth, if the business cannot deliver does it actually get any value from knowing customers want something they cannot have?

The layers stack up on top of each other. Each providing challenges and barriers that mean that the ‘easy’ customer need is much more of a balanced compromise. Insight comes not from knowing what customers want, but in the aha moments of spotting a way to shortcut through the compromises and limitations that each layer adds.

If you take a tablet device for instance, it was a format that had large numbers of predecessors - Apple Newton, Palm Pilot, UMPCs. It needed a compromise between screen-size, battery-life, weight and functionality. The compromise that eventually led to the iPad was to drop the productivity tools and to focus the layers on one simpler task - browsing. No typing or word-processor. Apple even started with some iPads that had phone connectivity, but even that was dropped.

The layers and compromises sharpen the research. And sometimes taking things away is as important as the constant pressure to add or shoe-horn in unnecessary things. Better research needs to understand the full stack and context to understand how things fit together, and where the core drivers of value are, and how they can be grouped and used. Just knowing what the customer wants is not enough.