Does a simpler provider selection step lift lead form completions?
The energy provider step of our quote form showed a matched provider beside a manual entry field and an inactive button. We turned the match into a button and hid manual entry behind a link. Conversion moved +4.5% over 17 days, short of our confidence bar.
Not significant Responsive 17 days Not called
The hypothesis
By making the matched provider a clear button and hiding manual entry behind a link, we will remove friction at a key drop-off step, resulting in more visitors completing the form.
Our step tracking showed the energy provider question was a key drop-off point in the form. When the form matched a provider, the match did not look clickable, and it sat beside a manual entry field and an inactive button.
We wanted visitors with a match to confirm it in 1 tap, and visitors without a match to type their provider with help from autocomplete.
What we changed
We styled the matched provider as a button and hid manual entry.
When the form found a matching provider, Var A showed it as a button and hid the manual field and inactive button behind a link asking if visitors did not see their provider. Clicking the link swapped in a text field with autocomplete from a list of national providers. With no match, the text field and button showed right away.
- Matched provider styled as a button
- Manual field hidden behind a link
- Autocomplete from national providers
- Text field shown when nothing matches
The result
Conversion moved +4.5% over 17 days, short of our confidence bar.
After 17 days, completed forms were 4.5% above the control. Early results looked promising, but traffic quality dipped partway through, and by the end the variation looked flat against the control. The lift never reached our confidence bar. Because the new step was easier to use and did no harm, it was implemented after the test.
What we would do next
An easier step did not move completions enough to measure.
The provider step was a real drop-off point, but this traffic was too low and too mixed in quality to show a small gain. We would pair the provider change with other fixes to the same step, or test it again when traffic is steadier.