During Alpha, the team working on the Get Into Teaching website decided that the adviser service sign-up form should sit on a separate site and have the look and feel of a GOV.UK transactional service rather than aligning with the campaign. Their hypothesis was that by users could then access this higher commitment service when they were ready without feeling pressured.
A separate service was created with a different subdomain (https://adviser-getintoteaching.education.gov.uk/):
During public beta, we carried out user research on the adviser sign-up. Users told us that the reassuring look, feel and tone they experienced on the Get Into Teaching website was missing from the GOV.UK site and it actually made them more concerned about the level of commitment needed, which was in contrast to the Alpha hypothesis. They found the transition from the Get Into Teaching website to a GOV.UK site jarring, didn’t always understand the benefits of adviser and often just weren’t aware of the adviser service.
However, once they were shown the Get Into Teaching page explaining the adviser service, they understood the value of getting an adviser. The data also showed that adviser sign-ups were decreasing year-on-year.
Benefits of integrating the services
We decided to investigate bringing the adviser sign-up into the Get Into Teaching website, so that all of the support services were in one place. We identified multiple benefits of doing this.
Firstly, the user experience would be improved. Having all the support services on one site would make it easier to explain the overall Get Into Teaching service proposition to users. We would also raise awareness of the adviser service and be better able to explain the benefits. There would be no jarring transition to a different site and users wouldn’t be confronted with two cookie opt-in processes.
Secondly, we would benefit from better data insights. If users landed on the separate adviser site first, we struggled to track them as cookie opt-in was very low. In addition, our Google Analytics account covered both sites which made tracking confusing. We also couldn’t split test on the separate adviser site.
Thirdly, consolidating the two applications in GitHub would reduce the maintenance effort for our developers, for example, only doing one set of dependency updates each month rather than two. It would also make it easier to keep our sign-up forms consistent.
Most importantly, our hypothesis was that the number of sign-ups to the adviser service would increase. We knew from the data that if a user had read the Get Into Teaching page on the benefits of getting an adviser first, they were much less likely to exit from the start page of the adviser site and much more likely to complete the form.
However, we were aware that there might be possible risks. For example, users may be less keen to give personal details if the site didn’t look like GOV.UK.
Testing the hypothesis
We decided to test our hypothesis. After a tech spike to make sure the move was feasible, we added the adviser sign-up to the main Get Into Teaching site, whilst keeping the separate Get an adviser site running.
Our intention was to run a split test using Google Optimize, but this proved extremely difficult across the two domains, as the two sets of cookies involved made the data very unreliable.
Instead, we switched to parallel running the funnels and benchmarked the form starts and completion rates.
The data showed that the completion rate on the Get Into Teaching site was better than that on the Get an adviser site. Having proved that the impact of the move was positive, we decommissioned the old Get an adviser site.