As an optimization leader, you know the value A/B testing brings to your company. You’ve seen the ROI on your optimization investment, and you’re looking to grow your efforts.
So, where do you look next?
To increase opportunity, you have to expand your optimization team’s A/B testing focus beyond just the look and feel of your site. You have to A/B test deeper by branching into server-side optimization. For example, A/B test new areas of your site such as search algorithms, or new features to help minimize risk. This may require expanding your optimization across departments. So here are some best practices to make the move.
Document Your Program
Create a centralized document or presentation that highlights your optimization goals and recent successes. Include the principles that guide how the A/B test results affect your decisions. Once you have documented your team’s mission, everyone knows what to expect. Are you looking for guidance on where to start? Our team has helped many world-class businesses translate their optimization ideas into a documented plan.
Share Your Knowledge
During a customer visit, one meeting attendee stated:
“We held an information session for our development team about our A/B testing program, how we apply A/B test results, and what it means to what they are working on. We need to continually reinforce this as we work together.”
Educating your colleagues is very important. During your onboarding, SiteSpect delivers a custom training to your team. You can use what you learned to create your own training program. Be sure to provide clear examples that illustrate how the A/B test results affect other teams.
For example, let’s assume a guiding principle is to A/B test all new features prior to site-wide release. For one new feature, A/B test results indicate the feature has a negative impact on achieving your site goals. The team decides to not release the feature because of the impact. So what do you do next?
Involve Your Product and Development Teams Early On
If these teams are actively involved in your optimization program, the conversation will be much easier. When you talk to the developer, they will already have an understanding about why the feature may be pulled. It’s important to reiterate that the decision has nothing to do with the work and everything to do with the data.
Find Testing Champions
Chances are, as A/B testing continues, you will find a few optimization enthusiasts on other teams. These colleagues can be your greatest advocates within their own teams. Establish a dialogue to gauge the level of team interest and identify any areas of concern. If worries surface, work together to improve the situation. We have a lot of experience helping organizations develop, foster, and promote a culture of optimization. We can help you do the same at your company.
Invite the Team to the Table
Work with your champions to cultivate an ongoing dialogue between teams. Collaboration works much better than division. Share your A/B test plans on a regular basis and be open to any A/B test idea suggestions they provide. Their views may provide unique perspectives or help you develop A/B tests more efficiently. Once you have cultivated a list of A/B test ideas, our team can help you vet and prioritize your A/B testing options.
Share Your Successes
You learn something from every A/B test you run. Perhaps a recent A/B test made your company money or even stopped your company from losing money. When that happens, let everyone know. SiteSpect includes several reports to help you share your results. If you have a standard analytics tool, you can also use the SiteSpect API to pull necessary information from SiteSpect and create your own custom reports.
These are just a few ideas on how you can expand your optimization program and keep everyone singing along. We have an experienced and talented Professional Services team to help you along the way. Do you have any additional ideas to add? Or experiences to share? If so, we’d love to hear them!
To learn more about SiteSpect, visit our website.
Subscribe to our blog: