Growth hacking is not only applicable to online marketing as we know it. You can "growth-hack" nearly every aspect of your business. And if it comes to recruitment, it's the best way to go. By leveraging growth hacking techniques we helped our client, Dura Vermeer, increase the number of applicants per month by 25% compared to the previous year. How we achieved that? Read this case study to learn about our successful framework.
Scroll down to see the results of (some) of the changes we made
Dura Vermeer is one of the largest construction companies in the Netherlands. Their expertise is infrastructure and real estate. They’re executing many different projects from constructing buildings to improving the public space. Before Dura Vermeer started cooperation with RockBoost, they had their brand new website developed by DotControl. After that, they turned to us to step up their game in recruitment.
The OMTM was to increase the number of applicants per month by at least 25% compared to the previous year. However, we weren’t responsible for acquisition but rather the whole funnel:
Let me break it down into consecutive steps we followed in this project.
We proud to say there has been a 48.6% lift in the number of applicants (comparing 2018 to 2017). We see a reduced drop-off at the end of the funnel, as well as additional applicants via email (as a result of the smart bar). We also see that the new channel mix provides more candidates at the same costs (lower cost of acquisition). The next step was to reduce the drop-off between viewing a vacancy and applying for a vacancy.
Less bounce rate
Lower cost per client
Click through rate
More page views
In order to get the data right, we had to start by making everything measurable. It isn’t the sexiest part of the process, but our 53-step Google Analytics Audit is the key to ensuring there is a proper Google Analytics setup:
Part of making this happen was also setting up a measurement plan that included event tracking. This gave us an extra layer of valuable behavioural data.
Whilst the improvements in Google Analytics and event tracking already gave us a far better understanding of the visitors, there was still one major detail to be fixed: the actual application form was located on a completely different domain.
The first step was to set up event tracking on that domain so we could measure how many people clicked on the CTA. Even though it was only a micro-conversion and not the end-goal, the event tracking gave us a basic understanding of the applicants per vacancy.
In the meantime, we discussed cross-domain tracking possibilities with the third party that managed the application form.
Once we had that setup, we could understand far better:
By ensuring that all website changes were data-driven we identified how we should use the website budget to improve the customer journey and get more applicants. In Google Analytics we found the biggest drop-offs and combined this quantitative data with Hotjar data. This tool provided us with heatmaps and user recordings to better understand the reason for the drop-offs.
For example, we noticed that many visitors abandoned the funnel right at the beginning. After having a look at the data we realized that even though applicants had to choose between “starter” and “professional” vacancies they eventually ended up on the general page with all vacancies anyway. When we dove deeper into the user recordings, these screamed out hesitation and confusion about to which group they belonged and which vacancies they should go to.
When we asked recruiters what their experience with this was they confirmed that applicants indicated they weren’t sure which bucket they fell into. They were either looking for an internship or a full-time job. We simplified this by leading straight to the overview of vacancies and offered them an option to choose between internships and vacancies once they landed on that page.
In the ideal world, we would A/B test all these changes first. However, we did not reach the minimum number of applicants per month.
We decided to take the “ before / after” approach:
We required at least a 15-% lift to ensure that the effects aren’t due to other changes.
Whenever possible, we got creative. We used Sumo smart bars and pop-ups to A/B test changes before implementing them on the website.
For example, we noticed that a huge amount of traffic came from mobile. However, only a few people had their CVs available on their mobile devices. So we created a smart bar targeted at mobile users only encouraging them to leave their email addresses. Then we identified which vacancies they were applying for via a hidden field that captured the URL. We used this URL as a merged field in Sumo that sent the visitor back to the vacancy. If they still didn’t apply, they received automated follow-ups. We captured hundreds of extra applicants using this method.
The fact that we now had the data to optimise the website was not enough. We wanted our client to understand and use the data too. This way they could make better decisions and they were not reliant on us. That’s how we ensured we could use more time to help build the website further.
The recruitment team was enthusiastic about this! However, they had no previous experience in marketing. So after creating actionable dashboards in Google Data Studio, we provided the team with training every 2 - 3 months to bring new members up to speed and answer questions.
Based on their input we kept improving and developing the dashboards further to enable them to understand their target audience better.
A large company like Dura Vermeer has multiple different traction channels to find the best possible talents for their teams. However, with so many channels, how do you know what is working?
We had already customised the channel grouping in Google Analytics to provide them with a clear overview of their actual channels, but that wasn’t enough.
So we gathered data from the various stakeholders and calculated the CPA per channel. We looked at how channels worked together in the customer journey to attribute scores to earlier channels in the customer journey based on how much traffic they brought in.
We optimised the channel distribution further to improve the use of budget in getting new applicants.
The next step was to see if the ratio of applicants on candidates differs per channel to weigh high-quality channels more.
One of the best parts of the project was close cooperation with Dura Vermeer product owner. She fully trusted us in all the website and channel changes. She also helped us get other people at Dura Vermeer onboard.
This resulted in a strong collaboration with the recruitment team. Next to regular training, they often reach out to us with additional questions. They suggested making the dashboard a continuous project so that every month it’s improved and optimised as the team gets more advanced and understands the data better. Starting as an MVP of 1 page with the most basic metrics the dashboard grew proudly to 10 pages of actionable data sets adjusted to different needs of the recruitment team.
Another factor that made this project successful was the “baby steps” approach that we took. We didn’t try to get the perfect data setup in one go. We got the best out of what we had available and worked on improving the data quality in the meantime.
This project was a collaboration between RockBoost & DotControl.
Drop us a line and say
Van Nelle Fabriek
Van Nelleweg 1
3044 BC Rotterdam
3004 HC Rotterdam
Our main office is located in the most prominent industrial monument of Dutch architecture, the Van Nelle Fabriek.
Open via Google Maps