Does a progress bar on a winning lead form lift conversion?

An earlier test had found a winning version of this lead form. We retested it with an updated progress bar, and with the progress bar plus 2 steps removed. Over 16 days, conversion fell 10.8% with the progress bar, short of our confidence bar.

Not significant Responsive 16 days Not called

SURFACE

Lead form

DEVICE

Responsive

VERTICAL

Lead Gen (CPL)

PRIMARY METRIC

Conversion rate

VARIATIONS

2

The hypothesis

By adding an updated progress bar to an earlier winning form, and removing steps, we will confirm the earlier lift and make the form feel shorter, resulting in more leads.

What we changed

An updated progress bar, alone or with 2 steps removed

Both versions started from the winner of an earlier test on this form. Var A added an updated progress bar. Var B kept that progress bar and also removed 2 steps: the state question and a short engagement step.

VAR A: PROGRESS BAR

Updated progress bar
Built on an earlier winning form

VAR B: REMOVE STEPS

Updated progress bar
State question removed
Engagement step removed

The result

Conversion fell 10.8% with the progress bar, short of our bar

CONVERSION RATE

-10.8%

LEAD VALUE PER VISITOR

-0.2%

VAR B CONVERSION

-9.9%

Over 16 days, the updated progress bar (Var A) converted 10.8% worse than the original, and the version with fewer steps (Var B) converted 9.9% worse. Neither gap reached our confidence bar. Lead value per visitor was flat for both, down 0.2% and 0.3%.

What we would do next

Move on from progress bar and step tests on this form

Neither change moved results in a clear direction, and earlier progress bar and step tests had given mixed results across our forms. Rather than keep tuning these elements, we would test a different part of the form next.