The problem we faced to tackle in this iteration of the cleaner application funnel, was our bounce rate. With a bounce rate of 65%, though the conversion rate we had was suprisingly high, we felt there were certain ways we could improve that metric.
A few problems we identified from the current landing page were:
To win the trust of cleaners upon entering the landing page, we wanted to eliminate the lack of 'focus' of the benefits of working as a cleaner at Hassle.com. To do this, we used data gathered from cleaner surveys. The most prominent being the following:
In addition to the excessive form weight, there was no information on how the application to Hassle.com itself proceeded.
We set a success metric of 40% bounce rate, once finishing the feature, we would review the data after a certain amount of tiem, to see whether the bounce rate had dropped.
Before wireframing, I ran a small excercise of stacking the 'sections' in level of importance; me and Ben, did this independantly, then showed the results. This helped me to have a general sense of heirarchy for the elements on the page.
My sketching led to some extremely low-fidelity wireframes, which we reviewed as a group of two designers. We quickly talked about what worked and what didn't in terms of intial prototypes; also allowing us to throw away quickly, any prototypes we weren't happy with.
This led me on to higher fidelity wireframes, using some dummy text, but also real text where applicable. We had decided to ship a 'condensed' version of the page and then individually ship each 'section' as a small feature, incrementally.
Having a fairly good base to start building, I also decided to make some changes further into the bid funnel, as I would have to chnage the layout due to the changes we'd made. I decided to make the following changes:
Upon shipping the changes above, we saw a reduction on bounce rate to around 43%, (a 22% decrease) thus practically hitting our success rate. The plan is to start shipping the individual tickets already wireframed, which we're hoping will decrease bounce rate even further.
In terms of conversion, we increased overall conversion by about 1%, not as high a statistic, but our hypothesis is that this shall also increase with the extra iterations.