Before
After
Context
Kolabtree helps businesses hire freelance experts and scientists online. Users are asked to post a project, during which they select the service they are looking for. An internal algorithm works to pull up a list of recommended experts.
The problem
Kolabtree clients were taking a lot of time to choose a service as the options were confusing. Many users were also choosing the “Other” option even though their service was listed. This meant that they found decision-making difficult at this point and wanted to quickly move on to the next step. However, choosing a wrong service impacts matchmaking, and so users did not see the right set of experts, leading to a drop in conversions.
The solution
With the aim of reducing the amount of time users spend making a decision on this page, we wanted to classify Kolabtree’s services into not more than five categories.
My role and approach
My role was to research how users were interacting with the product, specifically understanding the language they were using while posting a project. I found that many users were choosing the “Other” option even though their requirement was listed in the options, which meant that they wanted to quickly move on to the next step. Choosing a wrong service impacted matchmaking, and users were not able to see the best-suited experts.
I used a five step approach:
- Research existing user behaviour
- Understand the language clients were using and mapping them to our services
- Editing and merging categories
- Testing out new categories with existing and new categories to make sure nothing was missed
- Looking at heatmaps and GA data to analyze drop-offs and flow to next step
The benefits
- The number of users filling out this information and moving to next step grew by 30%.
- This step guided the rest of the steps on the form, making it easy for users to see a list of matched experts. This improved conversions.
- The information captured at this point helped us refine subcategories and show pricing recommendations.
