Action plan

An action plan is a systematic plan for identifying, prioritising and carrying out improvements that raise the conversion rate.

Without a structured action plan, data and analysis rarely lead to changes on the website.

A well-designed action plan turns data-driven insights into concrete, prioritised actions with clear ownership and a timeline. It makes conversion rate optimisation (CRO) a structured, continuous process that delivers results, rather than a series of one-off, ad hoc projects.

What is an action plan?

How an action plan is defined

An action plan is a prioritised list of actions intended to improve the conversion rate, based on an analysis of user behaviour and the friction points identified.

Why is it needed?

Two main starting points

The priorities and focus depend on the company’s current situation, as shown by the conversion analysis.

When many visitors start to buy but do not finish

Problem: A large share of visitors start the conversion process (click “Buy”, begin filling in a form), but many of them drop off before the purchase or the form is complete.

Focus of the action plan: Fix the weak links in the conversion chain. Identify where users drop off and why.

When few visitors start to buy at all

Problem: Only a small share of visitors start a conversion at all.

Focus of the action plan: Increase the number who start to convert by making the calls to action clearer, improving the value proposition and reducing the initial friction.

How to build an action plan in five phases

Phase 1: Identify the problems

The first step is to identify problems and opportunities.

Methods

Heuristic analysis (expert evaluation): A CRO specialist goes through the website with a critical eye and identifies obvious friction based on proven methods.

Result

A list of 20-50 identified problems and opportunities.

Quantitative analysis (data)

From the conversion analysis: Where in the funnel do most people drop off? Which pages have the highest exit rate?

Qualitative analysis (users)

Session recordings, user tests and surveys: Why do users behave the way they do?

Summary

All identified problems are gathered in one document with:

• A description of the problem
• Where it is (URL, location on the page)
• How serious it is (impact on conversion)
• Evidence (data showing that it is a problem)

Phase 2: Prioritise the actions

Not every problem can be fixed at once. Prioritisation decides what is done first.

How to prioritise with the impact versus effort matrix

High impact, low effort (quick wins): Do these first.

High impact, high effort: Do these next.

Low impact, low effort: Do when time allows.

Low impact, high effort: Skip.

How to score actions with the PIE framework instead

Potential (potential for improvement)
Importance (importance to the business)
Ease (how easy it is to implement)

Each action is scored 1-10 on each dimension. The total score determines the priority.

Example

Phase 3: Plan the implementation

Once prioritisation is done, an implementation plan is drawn up.

For each action, the following is defined

What

Exactly what is to be done.

Example

“Add a Trustpilot widget to all product pages, with the average rating and number of reviews visible above the ‘Buy’ button.”

Why

The expected effect.

Example

“Increase trust and reduce hesitation. A possible target: 10-15 per cent more people adding to basket.”

Who

The person or team responsible.

Example

“Front-end developer and designer for the implementation.”

When

The deadline.

Example

“Sprint 3, weeks 10-12.”

How the result is measured

Which KPI is tracked.

Example

“Add-to-cart rate per product page, before/after comparison.”

Document format

Phase 4: Carry out the actions

Quick wins (low effort)

Implemented straight away without an A/B test if the risk is low.

Larger changes (high effort)

A/B tested before a full rollout to show whether the change improves conversion.

Proven changes

Some actions are so well proven that they can be implemented straight away.

Experimental changes: Changes whose effect is uncertain should be tested.

Phase 5: Measure and follow up

After the rollout

Set the baseline

What was the conversion rate before the change?

Measure the results

What is the conversion rate after the change?

A before/after comparison is a weaker method than an A/B test, because seasonality, campaigns and a changed traffic mix can explain the difference. Compare equivalent periods and note what happened during the measurement.

Document

What was done

A description of the action.

Results

Figures (conversion rate before and after, and statistical significance when the change was tested in an A/B test).

Insights

What did you learn? Why did it work (or not)?

Next steps

What should be tested or improved next?

Actions that are often needed

Improve the value proposition

Problem: Visitors do not understand what they get or why they should choose this company.

Expected result: A lower bounce rate and more people starting to convert.

Optimise the calls to action

Problem: CTA buttons are not visible, are unclear or are badly placed.

Expected result: A higher click-through rate on the CTA and more people starting to convert.

Simplify forms

Problem: The form is too long and too complex, and asks for unnecessary information.

Expected result: More people complete the form, and fewer drop off halfway through.

Build trust

Problem: Visitors do not trust the website or feel unsure.

Expected result: Higher conversion, especially among new visitors.

Optimise speed

Problem: Slow loading times, especially on mobile.

Expected result: A lower bounce rate and higher conversion (especially on mobile).

Improve the mobile experience

Problem: Desktop converts well, mobile converts poorly.

Expected result: Higher mobile conversion (closer to desktop levels).

Reduce distraction

Problem: Too many choices and too much noise overwhelm visitors.

Expected result: Sharper focus and fewer people leaving.

Training and knowledge transfer

Why training matters

For an action plan to work in the long term, the company needs to build in-house expertise.

Solution: Train early. Build an understanding of CRO principles and processes.

Who should be trained?

Developers

Understand the technical side (A/B tests in your own systems, tracking, performance).

Designers

Understand UX principles for conversion (CTA design, layout, friction).

Content and marketing

Understand the value proposition and copywriting for conversion.

Product owners and management

Understand the business value of CRO and how resources are prioritised.

How the training can be structured

Collaboration between teams

CRO requires cross-functional collaboration.

Who does what in a CRO team

Communication

A shared tool

Everyone should have access to the action plan (Jira, Trello, Google Sheets).

Open reporting

Results from tests and implemented changes are shared with all the teams involved.

What a complete action plan can look like

The situation in a hypothetical online shop where few complete their purchase

Problem: 85 per cent of those who added to basket drop off.

The actions in order of priority

Priority 1 (quick wins, weeks 1-2)

Action 1: Add payment logos at checkout

• For example Visa, Mastercard and Klarna
• Effort: Low (a few hours of design and development)
• Impact: Medium to high (builds trust)

Action 2: Shorten the delivery form

• Remove “Company name” (optional) and “Phone 2” (unnecessary)
• Effort: Low (backend adjustment)
• Impact: Medium (fewer fields are quicker to fill in)

Action 3: A progress indicator in the checkout

• Show “Step 1 of 3” clearly
• Effort: Low (front end)
• Impact: Medium (reduces uncertainty about how much is left)

Priority 2 (rolled out in weeks 3-6, with before and after measurement)

Action 4: Try a one-page checkout instead of several steps

• Hypothesis: Fewer pages give fewer exits
• With around 60 purchases a month after Priority 1, the shop is too small for an A/B test of the checkout. Instead, track the share of people who move on between the steps, before and after the change.
• Effort: Medium to high (redesign of the checkout flow)
• Impact: Potentially high

Action 5: Try guest checkout instead of requiring an account

• Hypothesis: Guest checkout reduces friction
• Effort: Medium (backend logic)
• Impact: Potentially high

Priority 3 (long term, weeks 7-12)

Action 6: Implement address autofill (Google Places API)

• Effort: High (API integration, testing)
• Impact: Medium (faster form filling)

Action 7: Optimise the mobile checkout

• Effort: High (a dedicated mobile design)
• Impact: High (if mobile conversion is much lower than on desktop)

How much conversion can increase in the example, if the assumptions hold

Keep the action plan running

CRO is never “done”

Mistake

Thinking of CRO as a project with an end.

Reality

CRO is a continuous process, and there is always something more to improve.

Update the plan when circumstances change

These may call for an update

A new target group

If the company starts targeting B2B instead of B2C, completely different friction points may appear.

A new product

Launching a new product line may require changes to the checkout or product pages.

Technical changes

A new platform or design can create new problems or solve old ones.

Market trends

Customer expectations change (for example faster delivery, easier returns).

Document and learn

Build a knowledge base

An internal wiki or document where all tests and results are recorded.

Share insights

Give regular presentations to other teams about what you have learnt from CRO.

How to get from plan to results

An action plan turns insights into improvements.

Process

Contact us about action plans

Carnaby helps companies create and carry out structured action plans, from identifying problems to implementing solutions that improve conversion.

Contact us →