UX + CRO = PROFIT: How to Use Your UX Skills to Improve Conversion Rates (Part Two)

Albert Wavering, Former Digital Analyst, and

Laura Sweltz, Director of UX Research and Strategic Initiatives

Article Categories: #Strategy, #Data & Analytics

Posted on

Use your UX skills to improve the bottom line with this CRO guide, from implementation to analysis.

Previously on UX + CRO = PROFIT… We provided an overview of Conversion Rate Optimization (CRO) and introduced a framework to help you run your own CRO experiments. If you haven’t read part one yet, go read that first before starting part two. Don’t worry. We won’t start without you.

Now that you’re all caught up, let’s jump into part two of our CRO adventure. As a reminder, our framework has 7 steps:

  1. Generate ideas.

  2. Prioritize experiments.

  3. Create an experiment plan.

  4. Run your experiment.

  5. Conduct analysis.

  6. Share results.

  7. Implement the winner.

In part one, we covered how to generate ideas, prioritize experiments, and create an experiment plan. Now that we’ve completed the planning steps of the framework, we’re ready to move on to conducting and analyzing experiments!

Step 4: Run Your Most Important Experiment

Now, set up your most important experiment in your experiment tool of choice. If you’re making simple changes, you may be able to use one of the visual editors offered by Optimizely or VWO. For more complicated changes, you’ll want to talk with your front- or back-end development team to implement your planned changes.

Run your experiment for the amount of time estimated by your conversion volume, rate, and estimated improvement.

Several factors affect the time needed to reach conclusive tests:

  • total visitors

  • number of test variants

  • the current conversion rate

  • the expected conversion rate improvement (this tends to be the greatest source of uncertainty)

You should also document the following information for each experiment:

  • Experiment Details

    • Experiment Name

    • Suggested By

    • Started Date

    • End Date

    • Pages Where Experiment Runs

    • Experiment Hypothesis

    • Primary Goal

    • Control

    • Variants

    • Winner

    • Uplift (% lift / loss)

    • Statistical Significance

    • Data Last Updated

    • Interpretation

  • Results

    • Control Conversions

    • Control Visitors

    • Variation Conversions

    • Variation Visitors

    • Implemented? (Y/N)

    • Does the outcome lead to additional experimentation?

    • Notes

  • Analytics Info (Tool Dependent)

    • Optimizely ID

    • GA Custom Dimension/Variable #

Tracking this information will help you analyze the results of your experiment and determine whether or not your experiment improved your KPIs. It will also help you remember the course of events of your CRO program as team members change and memories fade as team members change and memories fade, as well as  report more accurately on why visitor behavior may have changed during different periods.

Step 5: Conduct Analysis

Analysis Tips

  • To get started with site optimization, it’s essential to have proper tracking (including custom events and goals) on your website. Tag every page and track major events (including major conversion points, but also micro-conversions such as social media button clicks). Put a value on macro and micro-conversions, filter your internal traffic, remove bot traffic, and double check with your web analyst or friendly neighborhood agency if you need help setting this up.

  • Track specific cohorts (groups of people who share a similar experience with your site or application). Some popular cohorts include users who joined at the same time, or came through the same acquisition channel. By comparing how cohorts behave and perform over time, you can identify factors that are associated with successful conversions.

  • Control for outliers when analyzing conversion rates. Controlling for outliers can involve scrubbing the data, isolating manually-defined ranges, or using statistical techniques such as median values instead of average values.

  • Set up heatmaps to measure more granular interaction data. Heatmaps are useful for identifying tactical challenges, like people who click on an image and expect it to be a link.

Step 6: Share Results

Once you’ve analyzed your data, it’s time to share your results with other people in your company so everyone can learn from your experiment. An assortment of tools are used to share results — from spreadsheets to Trello to dedicated tools like GrowthHackers Projects.

Graphing a series of experiments together can show your CRO progress over time.

Step 7:  Implement the Winner

Implement the winning variation by moving it out of your experiment platform and into your CMS, and move on to the next experiment in your calendar.

In conclusion

We hope that this overview has provided you with an understanding of how you can make conversion rate optimization part of your process. UX can and should play a role in optimizing conversions. Incorporating a UX role helps to expand the focus of these efforts. UXers can bring additional context to design work and iterations. Ideally, a conversion rate optimization effort will involve a cross-disciplinary team

. By working together, a team can incorporate experiments directly into the design process and those experiments can be deliberate.

Now it’s time to put the 7-step framework into practice. Good luck!

Laura Sweltz

Laura is Viget's Director of UX Research and Strategic Initiatives. She works from our Durham, NC office, where she helps clients like Rotary International, AARP, and Time Life understand the needs and behaviors of their users.

More articles by Laura

Related Articles