Building a product line-up from feature evidence

The question

A consumer technology company had a product in a young category and wanted a range that would reach more kinds of user.

It had to decide how many products to make and what each should be. Every feature, design and price in the line-up was a trade-off, and the company wanted evidence on how people would make those trade-offs before it committed to any of them.

Our approach

We built on what the company already knew, starting from its earlier research and personas, and its team helped shape the long list of features to test.

People sorted that list and picked the features they would pay a little more for, so features could be grouped into the products they belong in. A choice exercise then had people trade features, designs and prices against each other, with the groups that mattered most to the plan sampled large enough to read on their own.

What the team came away with

The team could test a line-up before committing to it.

A conjoint model and a simulator let it try ranges built from scratch, additions to its current products and responses to a competitor’s move, all within the features and prices the research measured.

Behind the model sat the features that mattered to each kind of user, sizes for the personas the company already worked with, and the clusters of features that point to distinct products.

Every output is a preference estimated under the conditions tested, never a sales forecast. Read that way, the model stays useful as plans change: the team can rerun scenarios whenever a new line-up is on the table.