Life is to short for only testing small changes! – Conversion Sweetspots Series
Recognize this? Please leave a comment
UX-ers, marketeers and text & visual editors take 100’s of small decisions while designing & releasing pages; Where does that Unique Selling Point go, is there a Heading and a Subheading, do we show the product below each other or next to each other and many more decisions like that. Usually these are informed by best practises or what Airbnb, Deliveroo or whatever is ‘cool’ with the peer group of the UX-er or marketeer. Sometimes the decisions are informed by sheer ‘divine’ inspiration of the ‘creators’ of the page.
Life is to short for testing small changes
The ‘traditional’ way A/B testers or CRO’s are ‘raised’ by their peers is to make a single change on a page and wait 2 -5 weeks until the 80% power 95% significant ‘verdict’ of the A/B tests arrives. Unless you are blessed with Booking.com, Ebay or ABN Amro like traffic & conversions you will never succeed in staying ahead of proofing their unproven -best practice & divine inspiration based- output.
Radical A/B testing
Be testing super small iterations while being flooded by 100’s of untested assumptions being made by well meaning coworkers in scrum teams? So what to do……..?!!!! Jump ahead of the curve and do some radical A/B testing; Big Changes First! Why?
- You have limited places/time to test; Not every part of your site is a sweet spot for testing. So you need to make it count!
- Your colleagues are taking 100’s of unproven small decisions which you can only proof or falsify one test at the time. You need to get yourself in the driving seat: test a few BIG ideas first.
- Find a Big winner; then you start testing for small improvements in the BIG winner.
Don’t get stuck in a local optimum!
So realizing life is to short for testing small changes will get you ‘back in control’. Of course A/B testing is about finding improvements, high velocity; climbing the Conversion and Customer Satisfaction Mountain small steps. But never forget you sometimes need to switch mountains because the other one might have a higher top.
The sweet spot: get involved before pages get released. So if a new design is abut to happen. First test important elements on the current website: then launch. You might call this: factbased redesign.
Please share your experience
Do you recognize this situation? Please share your experiences!