How to prioritise when customers can't have everything

Most teams prioritise features based on internal assumptions, stakeholder opinions, or the loudest customer requests. They build roadmaps in conference rooms and wonder why adoption is disappointing. Meanwhile, customers would have chosen completely different priorities if anyone had asked them to make trade-offs.

Feature Voting forces customers to show you their real priorities by making them choose between alternatives with limited resources.

What's a Feature Voting test?

It's a prioritisation exercise where customers use a limited number of votes to choose the features they want most from a defined list. Instead of rating everything as "important," they have to make hard choices about what matters enough to spend their limited votes on.

It introduces a simple constraint that shows genuine preferences when customers can't have everything they want.

The results usually surprise teams who find out that their assumptions about customer priorities were completely wrong.

How a major retailer discovered what customers actually valued in a membership program

A leading retailer we work with was developing a paid membership program. They'd identified 18 potential benefits based on competitor analysis, focus groups, and stakeholder brainstorming. Internal teams were convinced that free delivery and exclusive discounts would be the main drivers of membership sign-ups.

Before launching the program, they decided to test these assumptions with a Feature Voting experiment.

The setup: They recruited 45 high-value customers who shopped regularly both online and in-store. Each customer received a list of 18 potential membership benefits with brief descriptions and 5 votes to allocate however they wanted.

What they expected: Free delivery and member-only discounts would dominate the voting.

What actually happened:

• Early access to sales won decisively: "48-hour early access to seasonal sales and new product launches" received 31 votes, more than any other benefit. This wasn't even in their core offering plans.

• Extended return periods ranked second: "90-day returns instead of 30 days" got 26 votes. The team had considered this a minor operational benefit.

• Free delivery ranked fourth: The benefit they planned to lead with received only 18 votes, behind several convenience features they hadn't prioritised.

• Personal shopping services emerged: "Free 30-minute styling consultations with store experts" received 15 votes despite being expensive to deliver and not in their original plans.

• The insight: Customers weren't primarily motivated by saving money on delivery or getting percentage discounts. They wanted access and convenience - getting first pick of sale items and hassle-free returns when purchases didn't work out.

Based on the results, the team completely restructured their membership program:

• Made early sale access the flagship benefit (addressing the highest vote-getter)

• Extended return windows for members as a key differentiator

• Positioned free delivery as a supporting benefit rather than the lead feature

• Added personal shopping services as a premium tier option

The repositioned membership program delivered:

• 78% higher sign-up rates in the first quarter compared to projections

• £8.4M in membership revenue within 12 months

• 34% increase in purchase frequency among members

• 91% member renewal rate after the first year

  
  

How to run Feature Voting properly

The power of Feature Voting lies in forcing real trade-offs. When customers have unlimited budget or no constraints, everything becomes "high priority." Limited votes show you what they actually value.

1. Create meaningful constraints

Give customers fewer votes than features (5 votes for 15-20 features works well). The constraint forces prioritisation and prevents everything from being "important."

2. Include features across different themes

Mix technical improvements, user experience enhancements, and operational features. Avoid lists that only include one type of improvement.

3. Use clear, benefit-focused descriptions

Don't just list feature names. Explain what each feature does and why customers would want it: "Automated maintenance alerts that prevent unexpected vehicle breakdowns" vs. "Maintenance module."

4. Allow vote stacking

Let customers put multiple votes on the same feature if they feel strongly about it. This reveals intensity of preference, not just broad appeal.

5. Segment results by customer type

Look at voting patterns across different customer segments, company sizes, or use cases. Different groups often have completely different priorities.

Try this next week

List 15-20 potential features or improvements for your current product or service. Give 5-10 customers exactly 5 votes each to allocate among these features. Pay attention to which features get multiple votes from the same customer and which get completely ignored.

You'll likely discover that customer priorities differ significantly from internal assumptions, and some features you considered essential receive little interest.