Skip to content

Optimizing the user experience with digital analytics

When we create or optimize digital solutions we base our decisions on qualitative and quantitative data. This is how Gjensidige's analytics experts work to improve the user experience.

1. Identify KPIs, case and goals

As with all forms of optimization, start with the problems. Is there for instance a user journey that doesn't quite work out? Are our customers having trouble finding a page, or not able to complete a task on our website?

When you've identified several problematic user journeys, pick one to investigate and solve. Narrow them down by looking at our business goals - find a problem that blocks our most important KPIs, for instance buying a car insurance. Also, select a case that isn't too large or complex. Scope the journey down to a manageable number of pages and paths.

When you've selected a case to investigate further, define a clear goal. This could be a list of action points that can be tested on a short or long term, for instance to increase sales.

2. Map out the traffic flow and user journeys

You've identified a problem for our users and our business goals, and you've set a clear goal on how to solve it. Now what?

Start by making an overview of the pages and paths that's part of the user journey. This could include a landing page, one or more pages you can enter from there and an application page or a form.

To find out which pages to include, ask yourself:

  • Which paths send the most traffic?
  • Which paths convert sales or forms better?
  • What differences in conversion do channels make? Do adverts convert better than organic searches?
There is lots of valuable information in Google Analytics that can help you get a good overview.

3. Fetch the baseline

Now that you know which pages and paths to include in your case, it's time to see how they're performing so far. 

Baseline is the term we use for numbers that show the current status for the use of pages, paths and solutions. The baseline offers insights that help us highlight the issues and possible solutions.

Gather these numbers from Google Analytics and other sources and include them in your overview. You could also get valuable baseline information by talking to customer advisors working with claims or sales, and look into documentation of solutions similar to your case. This could also help you form some hypotheses on why the problems arise.

4. Identify improvements

Hypotheses help you pinpoint specific issues that may contribute to the main problem. Having identified these, it's easier to suggest action points that may lead to a solution.

To find actions that might help improve the user experience, start by forming some hypotheses about what might cause the problems. For instance, there might be missing a link from the most visited page, or the CTA text might not work well with the other content and cause some confusion. Include several disciplines from the team to discuss probable causes.

5. Prioritize and test action points

Action points have to be prioritized within the team according to costs and resources. Therefore you should look out for "low-hanging fruits", actions that are quick and easy to implement, but likely to yield an effect.

Other changes might be tested with an A/B test, in which the new changes are shown to only half of the users, while the rest still use the old version. This way, you see whether the changes improve the solution's performance, and you make your decisions based on user data.

If you have many action points you'd like to test, try not to implement too many changes at once. If you test several changes simultaneously, you won't always be able to say which change made the most impact.

Don't implement too many changes at once.

6. Measure your changes

When you've found the action points you'd like to test, you naturally have to follow up your changes by comparing the new page performance to the baseline. Depending on your goals and what changes you've made, you could compare conversion rates such as numbers of purchases, or even the number of clicks on a specific button.

7. Deploy your successful changes to production

Having tested your hypotheses, for instance through A/B testing, you'll know whether or not your changes improved the user experience. If they did, you should of course display this new version to all customers.

8. You're never finished

During and after this case, you will of course have made many valuable experiences, observations, findings and new hypotheses. Make sure to save these for later - there might be an opportunity to continue the optimization work after implementing the first changes.

Good luck!