Follow the premise, not the experiment

Most experimentation teams treat their tests as isolated events. They start with a hypothesis, run a test, log the result, and move on.

This is a tidy workflow, but it misses something crucial: the connections between wins and losses, and the chance to compound what’s working into something bigger.

Rhys Mohun, founder of Formentor Labs and creator of the Room to Think newsletter, has spent over 16 years building experimentation programs, including developing Intuit’s Canadian experimentation practice. His approach is different: follow a premise, an underlying buyer motivation, and pull that thread across the business until it stops producing results.

Then find a new one.

What is a premise in experimentation?

It’s easy to focus on the surface-level mechanics of a test: design, copy, placement.

The real power, however, lies one layer deeper: the buyer motivation that makes a test work in the first place.

{{quote}}

The premise is the “why” behind the “what;” the reason the customer converts. The button, headline, or element that triggers the action is important, but understanding the “why” is crucial.

Mohun describes a hierarchy of motivation:

• Trust: Does the customer feel safe? 
• Usability: Can the customer complete the task?
• Suitability: Does the product solve the customer’s problem and does the messaging say so?

At the top of the hierarchy are more marketing-oriented motivations: FOMO, urgency, luxury, and similar ideas. Any of these may be the premise that connects your experiments. The trick is to identify which one matters the most and commit to it.

{{cta-block}}

How a premise compounds results

In the earlier example, Mohun discusses comparability (whether mobile users could easily compare products) as a strong potential motivator for a given segment.

Discovering this could be a huge win for an experimentation team. Rather than celebrating the win and moving on, however, Mohun’s advice is more practical: keep pulling that comparability thread across moments in the customer journey. Each new test should focus on the same underlying premise. 

This accelerates experimentation velocity. When teams share a clear premise, they spend less time debating what to test next. 

{{blue-block-1}}

When a winning premise breaks down

Of course, not every winning premise transfers. Mohun’s team at Intuit led initiatives that drove double-digit growth. Following this success, he tried to replicate the strategy on a sales-led business line, to poor results.

{{quote1}}

From that failure, Mohun developed the “three-legged stool:” he imagines his premise as a stool supported by three legs: context, customer, and content. If all three “legs” remain stable, the premise is probably portable.

If even one changes, however, you need to retest. 

{{quote3}}

Before you carry a winning experiment to a new audience, check the stool; if a leg has changed, plan to retest, and never assume.

How to start experimenting with premises

• Name the premise behind your last win. When a test succeeds, ask yourself what buyer motivation drove the result. Was it trust? Urgency? Clarity? Name it explicitly.
• Pull the thread deliberately by looking for other surfaces in the customer experience where that same motivation applies. Build your next experiment(s) around that premise, rather than unrelated ideas.
• Check the stool before you scale. Before carrying a winning idea to a new challenge, audience, or format, check whether the context, customer, and content resemble the original win.
• Reset your expectations every time. Even when following a strong premise, remember that every experiment is its own event. You must plan for the possibility of failure, and know what to do next if it happens.

{{cta-block1}}