This blogpost is part of a series inspired by a great article from the Gogroup Digital. In this article they have asked some of the industry’s most respected thinkers why CXO programs fail. In the coming weeks we will dive into the most important reasons and discuss how to make sure your program won’t fail.
1 – Why experimentation programs fail: The experimentation program is not tied to an overarching business purpose and lacks a central guiding metric.
Running a CXO program costs money, experiments are expensive. You need a bunch of tools and a team of specialists to run the program. On top of that, not every experiment is a winner. As a rule of thumb: only 30-40% of experiments lead to a successful outcome. It’s the nature of the game. On the flipside, done it the right way, there is no other activity in digital marketing that can make a bigger impact than CXO.
So if running an experimentation program has already so many challenges from the start, you have to make sure that your experimentation program is directly tied to business objectives and makes it crystal clear how experiments contribute to overall business objectives. It is your only option to show the added value of your program. Remember, if you are not able to explain and show the added value in numbers that anyone can understand, your program is about to be ended.
A good way to do this is to break down business objectives into metrics that you can use in your day to day experiments. These metrics should be coherent and together they form your KPI framework. This framework will give you a clear understanding how metrics relate and even more important, will give you guidance towards new experimentation opportunities.
Here is an example for a retailer that has both a webshop and physical stores. Let’s assume the overall business objective for this company is profit. The break down could look like the image below. This example is for your inspiration. The KPI tree can be as long as you want. You could add funnel steps or break down “Costs” in smaller steps to include return rates for example. You can create endless variations for this set up. Just create a version that works well for you.
Here is another example. In this case we are looking at a KPI framework for a SaaS company. Their overall business objective is MRR: Monthly Recurring Revenue. The breakdown in relevant metrics could be as follows:
Now your business objectives are clear and you know what the relevant metrics are that drive your business outcome, you can organise your CXO program accordingly. Below are the typical steps of a CXO program. This is a generic set up, your program might be different but in general the steps mentioned below, cover the CXO process.
It is a best practice to do research before you start with your experiments. As you have defined which customer journeys have the biggest impact on overall business objectives, then this is where to explore CXO opportunities.
From your research you draft your hypothesis. As your research is centered around the customer journey with the highest impact, you are more likely to draft a hypothesis that will make a meaningful impact.
Resources are always limited and only so many sprints can be done per year. For that reason you have to prioritise your hypotheses. There are many prioritisation models around. The PXL and PIE models are widely adopted within the CXO industry. But if you want to make sure you spend your scarce resources on the experiment that will lead you to CXO nirvana, then create your own prioritisation model. It is okay to take a standard model as a starting point but make sure you include your business objectives in your prioritisation model. It is the best way to assure that the most important tests will get the highest priority.
After the three previous steps, running the experiment itself is rather easy. It is important though to collect and store all test information in a single place: your CXO knowledge base. Organise your experiments using tags and filters. Create filters for every customer journey and for every business objective.
As stated before in the post, not all experiments are winners. Make sure you always learn. From the set up with filters and tags that relate to customer journeys and business objectives, if done properly, will allow you to do powerful analysis over time. You will learn which journeys work best, what struggles customers and prospects face and how your program generates value for your company.
The set up explained in this article will take some time and effort. The good news is: it doesn’t have to be perfect the first time around. But not having a north star for your experimentation program means your CXO program will never thrive like it should (and you won’t get paid what you deserve).