Does combining contact steps lift lead form conversion?
Our tax relief lead form asked for email on its own step after the other contact details. We moved the email field onto the earlier contact step to cut the funnel by 1 step. Conversion rose 65.5% over 8 days, and we implemented it.
Won Responsive 8 days Called at 90% Reviewed
The hypothesis
By combining the contact steps so the funnel is shorter, we will lose fewer visitors before the end of the form, resulting in more completed leads.
On a sister funnel, combining contact fields into fewer steps lifted conversion by more than half. Moving the email field had been a strong runner up there, and we thought it might suit this audience better.
We tested 2 ways to shorten the form: moving only the email field, and moving both email and phone.
What we changed
We moved the email field onto the earlier contact step.
In the control, email had its own step near the end of the form. Var A moved the email field onto the earlier contact step, which removed 1 step. Var B moved both the email and phone fields onto that earlier step. The questions and the design stayed the same.
- Email field moved to the earlier contact step
- Separate email step removed
- Email and phone moved to the earlier contact step
- Separate email and phone steps removed
The result
Moving the email field lifted conversion 65.5% over 8 days.
Both variations beat the control. Var A, which moved only the email field, gained 65.5%. Var B, which moved email and phone, gained about 32%. The test started on part of the traffic for a data check, then ran for 8 days on full traffic. The gap was large enough to call, and we implemented Var A on the live funnel.
What we took from it
Asking for email 1 step sooner cut drop off in this form.
Asking for email alongside the other contact details removed a step where visitors were dropping out. Moving the phone field too still helped, but less, which suggests that a heavier step can cost some of the gain from a shorter funnel. We then used Var A as the starting point for tests on other funnels.