Why jumping straight to solutions can derail your experimentation efforts

It's tempting to jump straight into solutions rather than clearly defining the problem

In digital experimentation, it’s easy to dive right into solutions. However, doing so can derail your optimisation efforts before you've even started...

We often jump to solutions because it feels efficient, productive and maybe even intuitive. However, solutionising too soon can mean you end up solving for the wrong problem and have you running confidently in the wrong direction.

Let's take an example of low-converting SaaS homepage. Someone suggests making the CTA more prominent or tweaking the copy — a seemingly logical place to start. 

The test is launched, and...nothing. No uplift...or worse still, a drop in conversion.

Why? Not because the test build or QA was poor. Not because the pre-test analysis was inadequate. 

It was because a key step was skipped: understanding why users weren’t converting. 

This is where the Double Diamond of User-Centric Experimentation comes in.

---

Source: UK Design Council's Framework for Innovation Double Diamond (adapted for experimentation)

What is the Double Diamond?

The Double Diamond framework was originally developed by the UK Design Council to describe the design process. It breaks innovation into four clear phases: Discover, Define, Develop, and Deliver. Each phase switches between divergent and convergent thinking.

First, explore broadly. Then, narrow down and refine. 

There’s also a variation of this model used more broadly for innovation, known rather unimaginatively as the UK Design Council’s Framework for Innovation.

It builds on the Double Diamond by embedding the surrounding conditions that make innovation work: leadership support and engagement, alignment across teams and a process to bring it all together.

I’ve adapted this version into what I call the Double Diamond of User-Centric Experimentation. It keeps the same core structure, but shifts the focus toward CRO and experimentation - helping teams define the right problems, explore the best ideas, and align more effectively around what to test.

Diamond 1: Define the right problem

Going back to the example of the low converting SaaS homepage. You may feel tempted to make the CTA more prominent, change the CTA wording, shorten the text, or add social proof and other features to improve the user experience. 

However, before we jump to ideas, we need to confirm that we're actually solving for the right problem by widening our lens:

• Run a Thematic Analysis Workshop. Invite people from product, marketing, customer support, and commercial teams. Ask: what do we know, and what might we be missing?

• Use AI for problem exploration. You can ask it to cluster feedback, create hypotheses, or question assumptions. For example, try asking, "What else might explain this behaviour?"

This process can surface a whole range of possible problems, including:

• Mismatch between ad messaging and page content

• Unclear value proposition

• Cognitive overload

• Trust concerns

• Confusing micro-copy

Once you’ve mapped all the potential problems, it’s time to narrow down. What’s the single most important problem to solve here?

This prioritisation happens as part of the Thematic Analysis workshop itself. Start by reviewing the full list of problems identified during the research phase. Group similar themes, and use dot voting or ranking to identify which issues are:

• Most impactful on the user journey

• Closely tied to business outcomes

• Feasible to address through experimentation and other forms of validation

This could lead to a shortlist of one or two main problems to tackle. Ideally, these should be the ones that are both strategically important and have solid evidence supporting them.

Align the team around these priorities and document the reasoning. A clear problem statement, agreed upon by all stakeholders, creates momentum and clarity as we move into the solution phase.

Diamond 2: Design the right solution

Now that we’re clear on the problem, we can explore how to solve it.

Again, we don’t stop at one idea. We widen out again:

• Run an Insights to Action Workshop. Invite input from different teams. Present the main problem. Then, generate multiple testable ideas.

• Use AI for further solution exploration. Ask for different messaging styles, layout ideas, UX patterns, behavioural nudges. Ask it for some crazy suggestions too, it's surprising how many of them turn out to be good testable ideas (with some refinement). 

Then, we converge:

• Score ideas using a prioritisation framework (e.g. ICE, PIE, PXL, etc.) or impact/effort matrix

• Evaluate based on hypothesis clarity and alignment to the problem

• The goal isn’t just to test something. It’s to test something meaningful.

In the example of the SaaS homepage, you might discover through research that the biggest issue isn't users not being able to find the CTA, but an unclear value proposition or a poorly communicated on-boarding process that puts off potential customers.

Once you have your top issue(s) to solve for, you can run the experiments, confident that you're systematically addressing real customer problems.

Why this works

This process does a few things really well:

• Eliminates reactive testing and encourages critical thinking

• Lead with curiosity and shared ownership

• Creates shared understanding across teams

• Engage diverse voices across the business

• Blends qualitative depth with quantitative focus

• Identifies any gaps in data

• Speeds up execution by front-loading alignment

• Improves the quality and relevance of your tests

And crucially, it reminds us that in experimentation, the value is not just in what we test — but in how we decide what to test.

Closing thoughts

We often think the bottleneck in experimentation is development time, QA, traffic, or tooling. But in my experience, the biggest issue is clarity and alignment over where to focus our efforts. The ideation stage is therefore key in any experimentation process:

• Good ideas come from a strategic and systematic way of defining the problems to focus on.

• AI gives you a way to thoroughly explore all potential problems and solutions. 

• The Double Diamond of User-Centric Experimentation helps teams decide which problems to focus on.

If this approach resonates and you'd like to bring more structure into your experimentation, I run Thematic Analysis and Insights to Action workshops tailored to your business. Let's chat.

Share in the comments what keeps you and your team focused on finding the right solutions to the right problems.