UI
it’s best to leave it for as long as you see fit. After
a while, the results will reveal whether or not
conversions have increased, which in this case can
be measured by comparing the CTA clicks before
versus after.
RINSE AND REPEAT
Decluttering offers the opportunity for other
elements to thrive. It’s vital that we didn’t remove
important elements in the last step because it may
be that the added cognitive load caused users to
ignore them. Our heatmaps might tell a different
story this time around, although if that’s not the
case then it may be that something is visually wrong
with those design elements.
After diving into the maps again, check elements
against the following design flaws and remember:
change/test only one thing at a time!
O Important elements not visible enough This
actually happened with us; both our CTA and book
cover blended into the background and, to make
matters worse, the cart button really stood out even
though there was nothing in it!
O Certain text snippets...
O Have suboptimal legibility
O Have suboptimal readability
O Don’t make any sense
O Don’t need to be there. This also applied to us! As
it turns out, users didn’t want our book description or
author bio, only the basic information.
O Too many forced or required steps
O Form or content too long/too boring
O Low-value visuals too distracting We stripped
the overuse of colour and imagery and instead used it
for our CTA, killing two birds with one stone. We also
swapped our flat ebook cover for a more-visible
iPad mockup.
O Element not as important as you thought
O Poor navigation (user can’t find the next step)Now looking at the website in question for the first
time, do you sense a change in usability?GIMME A CLUE!
Before we wrap up, let’s take note of a few more
observation tips that can help us solve some really
difficult mysteries.
Firstly, mis-clicks are better identified using
clickmaps. Are users clicking on something that’s
not clickable? This would inevitably lead to cognitive
overload and eventual frustration.
Are users scrolling and then stopping? They may
have found something interesting and this is an ideal
time to consider inserting a call-to-action below-
the-fold if you haven’t already. Scroll behaviours can
be observed using scrollmaps.
Just to throw a spanner in the works, could it be
that there’s actually something missing? Consider
user testing to find out. It could be something really
silly like a missing refund policy!Top Before vs after
reducing cognitive load
Above A heatmap showing
which areas are ‘hot’