Genetic evolution & A/B testing
Similarity between Genetic evolution and conversion optimisation
The first edition of Optimise Trends: Genetic evolution & A/B testing.
How does this newsletter work?
Every few weeks, we discuss various strategies and techniques that can be utilised to increase conversion rates, through A/B testing or personalisation. Here's how it works:
I scour through brands that successfully use personalisation to drive growth, then surface strategies for you to implement.
Break down specific changes you can A/B test.
We look at existing digital assets and explore how to increase conversions.
Sound good? Let's get started.
In This Newsletter:
How to apply genetic evolution theory to ongoing conversion optimisations.
Conversions and A/B testing
Breakdown: Optimising Nike.com
What your first A/B test should be
Genetic evolution theory & conversion optimisation.
When pondering on the question of “how to increase conversion rates?”, you often get led into a rabbit hole of attempting to figure out a solution to increase conversion. At this point, frustration kicks in while your thoughts cross through hundreds of ideas on how to improve conversions. Perhaps, “I should change that banner image?”, “the buttons are not large enough?” or perhaps, “what if I make a change and that doesn’t work?”. All of these “ideas” will lead back to the initial question and you’re stuck in a vicious cycle of “solution mode”.
Ok, so how do I approach the objective of increasing conversion rates? This is where we can start applying a systematic process, which in an oversimplified way mimics that of genetic evolution theory.
A simplified version of that process;
Define the objective (increasing conversions, duhh!)
Define the objective numerically (increase conversions on product A from X% to Y%)
Instantiate financial benefit ($) of achieving the objective (increase by X% will lead to an additional $Y revenue)
Gather all existing data, set a baseline and analyise/quantify which components of your site are likely culprits that contribute to the current low CR (Conversion Rate).
Document all proposed elements/components that are performing below the line, and recommend possible solutions.
At this point, we treat our conversion rate as a genetic profile(DNA) - that contains numerous “traits” that when combined result in your current CR.
so we have;
A Genetic Profile of our digital asset = Our current/desired outcome (Conversion Rate).
“Traits” = individual performance of components (example: individual pages within a funnel).
Adding everything together
Before we forcus on the outcome, we want to start tackling individual “traits/components” that are underperforming. Then, replacing them with a more efficient version.
Let’s take an example;
You have identified that the conversion rate of a funnel is low, and you’d like to change that (I mean, who doesn’t).
You also know that the first page in the funnel is causing 50% of users to leave. This is a negative trait that you want to be optimised.
Then, you define the hypothesis and document alternatives (better designs of that page) that are likely to perform optimally.
Once you have identified all of the “traits” that require improving, you can proceed in creating an A/B test for each trait/page element. This process of identifying negative traits and trying new variations of the same will eventually lead to highly optimised components (“traits”).
Conversions and A/B testing
So now you know exactly what you need to A/B test and you have your favourite tool ready to launch your A/B experiment. You run a test and get an uplift. At this point, you have made an existing “trait” perform more efficiently.
Few traps to avoid falling into;
“I ran an experiment and the red button performed 40% better. Let’s replace a few of the buttons on other pages, as we already know red performs better.”
Just as in genetic theory, different traits perform better in different environments than others. Simply transferring learned knowledge will not always work. Thus, always test before implementing a decision.Obsessing over a single trait.
It can be difficult when an A/B test result does not align with your hypothesis, so you try to logically frame the result positively. Remember that a “negative” result in an A/B test is more powerful than an inconclusive one.Focusing on results rather than the objective.
Who doesn’t want all A/B tests to result positively, falling into the trap of simply getting a good result in a test can hurt your overall goal. Focus on data, results, and the overall objective. Always.
Now onto an examples
Breakdown example: Optimising Nike.com
We will analyse few successful traits of Nike.com that contribute towards its conversion profile.
Localised geo-targeting
When visiting the site, based on your location, Nike will hyper personalise a user’s experience. Creating a sense of belonging and locality adds an additional layer of “relatability”. Despite Nike being a global brand, they become relatable to varying demographics and personas, all through personalisation.
In the screen capture below, viewing the site from a local state will render personalised content by promoting sponsored local events.
Element consistency
Let’s take a look at the primary navigation. Majority of sites will attempt to condense more pages than really required. Adding pages including “About us”, “Our story” would simply be redundant considering Nike has an established brand. That does not mean that you shouldn’t include those pages, it simply requires careful consideration “Is this really needed here and why?”.
Breakdown: Optimising au.Puma.com
Time relevance - Utilising emotional triggers
Puma has clear traits portrayed throughout its branding.
Speed (Puma)
Athleticism
Passion
…
Through association with sporting events, Puma takes advantage of large global sporting events to create partnerships.
While in theory, this association through sporting events appears trivial and unachievable for small brands, it can be extremely powerful. However, majority of smaller brands give up on the idea with the main blocker being “We don’t have that kind of budget to partner with global events” - The goal and value of promoting partnerships are not measured by the size of the event, rather the genuinity of it.
What can a small brand do?
Let’s take a parallel example to Puma, a small apparel brand servicing customers at a state or national level. While sponsoring a large event (even if you could financially) the “brand” value will not come across as genuine. Instead, sponsor a small local team, perhaps a youth soccer team.
Focus on local events first, then grow as your audience grows (naturally).
Try avoiding partnerships that are purely transactional, instead find local passionate teams to support.
Tribal - Creating a sense of belonging (picking aside)
Humans have tribal tendencies after all. There is nothing more powerful than passionately spupporting your local team, your country at a global event or your state.
While there is a fine line between healthy competition and creating division. Sporting events are a great way to generate fan loyalty as Puma has done. By providing content/products and associating them with various teams, allows customers to be part of the story, immerse themselves in the event, and therefore becoming part of the “tribe”.
Let us know if the type of content you’d like us to cover more in future editions.
Share this article with your team below