HOW TO MAKE YOUR OPTIMISATION PROGRAM MORE EFFICIENT (PART 1)

This week we will be discussing the concept of a Conversion package.

Some Optimisation platforms only allow for a small number of conversions per test. There are many reasons for this:

  • Cost (GBP and Resources)
  • Dev Implementation time for adding tracking
  • QA time to make sure the tracking is working
  • Could lead to confusion of the main KPI

At Webtrends Optimize if you have a Managed Service Contract then your Consultant will write out your test plan documentation (which is an agreement of work for the client and the consultant) as well as a job spec for the Developer to build against and for the QA team to QA against.

Within this document we detail the test hypothesis, key metrics and additional conversion points. Detailing the hypothesis helps everyone to remain focused on what the test should be proving / disproving and how this can be reviewed by the success of a conversion(s).

The reason we at Webtrends Optimize like to also track additional conversion points other than just the key KPIs is to make sure no other detrimental impact is occurring on our client(s) website.

For example, a test may see sales increasing, but by monitoring AOV and Units per Order we can see  whether these have changed; if they have dropped for example the test may not be as successful as first thought or additionally a phase 2 of the test needs to be considered to strengthen any weaknesses you may be seeing on other conversion points.

The additional conversions provide greater insight and visibility. (Another post worth reading which shows the value in wider awareness is ‘What’s the optimal?’)

Is this alot more work? Agreed, these additional conversions can make the project build more complex and timely, but how we at Webtrends Optimize get round this is by building all of our clients a Conversion package.

The Conversion package is an additional piece of work we like to carry out for all of our clients at the start of our relationship, due to the enormous value we see in having these in place longer term.

Each clients Conversion package tracks the key overall business metrics for the website. For example, each step within the purchasing funnel, purchasing behaviours etc.

This then means the Conversion package will automatically run for every single test that has been built, whether built via our own Developers or as a client self-service test.

This helps to make sure:

  • No company valued KPI is missed
  • Reduces conversion build time and QA (Seeing as QA only need to QA this package when first built and not for every new test).
  • Additionally, any specific conversions per each test can then be built with more focus and clarity and can be more complex if required.

To create your own Conversion package, I would recommend you start by writing down the conversions that you see repeated in every test that you have built, and these should then be the base for your own Conversion package. From there add any additional elements you want to be sure you track for each test.

I hope this time efficient tip helps with your future testing!

If you have any questions or would like some help please do get in touch!