So, what can we learn from the typical contestants coupling and uncoupling, and how can we apply it to a successful testing approach?
100% My Type on Paper
How many times have we heard that infamous line in Love Island, “they’re 100% my type on paper”? And more importantly why am I trying to link that to setting up a successful test?
Well, if you inspect it a little closer, what is being revealed is a contestant’s hypothesis for finding their match on the show. Hypothetically our subject believes that a counterpart with the right build, hair color, or “banter” will be their best match.
For a variety of reasons these initial impressions are often proven to be incorrect. This doesn’t devalue the outcome of the test, of course, and instead makes room for another, more accurate hypothesis to be made.
The purpose of a hypothesis is to ensure that your test stays on track, as having an expectation for what will change will make it easier to review results at the end and disregard anomalies.
For example, if you were testing targeting a customer profile with tailored ad copy, your hypothesis would be: by calling out USP X to persona Y, we expect higher CTR than with generic creative/copy as the message is more aligned to their priorities. This makes the parameters of the test clear (who you are targeting and what you will be changing), as well as what you are testing (CTR).
Before you approach setting up a test you need to make sure that your hypothesis is 100% established on paper.
Shooting your shot
Invariably the next step for a contestant past finding someone 100% their type (establishing a hypothesis), would be “shooting their shot” – or as I’d like to call it, coordinating the testing approach and plan.
For the average Love Islander, this would involve getting a pep talk from their pals, giving the heads up to other islanders that might be “stepping on their toes”, and of course getting their best chat lines ready to go.
Here in digital marketing land, it’s very much the same. We need to make sure that we communicate our testing plans with other team members to make sure that they are aware of any changes, and we need to plan the test accordingly.
The easiest way to keep track of your test plans is via a testing matrix. Here you can lay out the hypothesis of your test, start and end dates, testing details such as ad copy or bid strategy changes, and also monitor your results in an easy- to- access manner.
Can I pull you for a chat?
So, you’ve nabbed your ideal partner, coupled up and the so-called test has been running for long enough to start reviewing the results.
Islanders handle this through pulling each other to cozy corners of the villa to check out whether they’ve got the ick, want to take their relationship to the next level, or need to give it a bit more time to sort their head out.
I’m sure you’ve guessed it – testing is exactly the same.
Before calling it quits, deciding the test is over and starting to make assumptions from the test data you need to make sure that the results are significant. This means double checking your numbers to make sure that enough data has been collected to reliably take a learning from the results.
A slight uplift on a small sample size could over egg the potential value of the element you are testing. You may want to run your test for longer to ensure you pass significance levels, if even with more data the results are still not significant, the logical conclusion is that there isn’t any measurable impact for the topic of your hypothesis.
In love island terms this is best compared to not throwing in the towel with your initial partner for a head-turning-bombshell, and instead making sure you’re taking the time to pull them in for a chat before making a final decision. Otherwise it’s very possible you run the risk of looking like a “total mug”.
The moment of truth. The gut-wrenching, edge- of- your- seat moment. The Re-coupling.
All testing comes to an end now as couples now share their results and put them into practice.
For digital marketers it’s equally as important to review the results at the end of a test and take action upon them. Whether the result was positive, negative or inconclusive there are always ways we can apply the data to future testing.
This exponentially increases the value of the test being run, as continually reviewing past learnings to inform future tests encourages a growth mindset and greater understanding of what works or doesn’t.
Sadly Love Island is just a flash in the pan in our TV viewing, with the last episode expected to land on our tellies in mid-March. Luckily testing is a year-round programme, and should definitely be given prime time in your strategic approach.