3 Steps for designing your innovation experiments

  • Business Model Innovation

At BMI, we work with our clients to make innovation happen in order to move their businesses forward. But before finding out if an idea is worth pursuing, we first run a series of experiments to test the market, end user, and business assumptions so that we can set ourselves up for success.

This process starts by identifying and prioritizing assumptions, uncovering the areas that require the most validation (or invalidation) before moving forward. Once those are set, we carefully design our experiment within the context of the business and the project at hand.

Here’s what that looks like when designing your own experiments.

Considerations for your experiment

Before you design your experiment, there are a number of considerations to run through that will help shape how you run it.

  • Suitability: Determine what experiment format makes the most sense for the assumption you want to test. Other contributing factors will be any deadlines you’re trying to meet as well as the availability of skills and resources to bring an experiment to life. For instance, if you need to build a landing page, you’d need a web designer or an outside partner to help you do that. All thing considered, experiments are just tools – you don’t want to spend months setting them up.
  • Confidence level: The perfect experiment doesn’t exist. So as you design your experiment, you need to go in understanding how reliable the data you’re collecting will be. Experimentation can be often messy, which can make it a challenge to gather reliable data. We recommend doing multiple quick experiments that provide different perspectives or patterns, collectively providing a broader data set — and higher confidence level.

Designing the right experiment for your innovation project

Once you’ve landed on the considerations that will guide your design, there are several steps you can take.

1. Set a hypothesis

Creating a hypothesis up front is crucial. This will help you establish the type of information you’re looking for, as well as a benchmark to measure your results against. Without this foundational element, it can be difficult to validate or invalidate your assumption, and see what you’ve learned from the experiment. Setting a hypothesis is also a good way to avoid confirmation bias.

2. Format your experiment

When you’re constructing your experiment, there are a number of small design decisions that have an impact on how it’s implemented. For example, if you’re trying to get your test audience to choose the best feature on a list, you have to decide if you want them to rank these features or simply select their favourite. While a ranking will give you a deeper data set, having participants choose a single item will make the experiment go faster. Here, the choices you make should align with the data you’re trying to collect and the broader business needs tied with the project. You should ensure the data you collect can help you to test your hypothesis (and defined metric).

3. Remain rigorous

As you prepare to run your experiment, it’s important to not make any premature conclusions. Throughout the process, you need to stay conscious of the confidence level of the data, and be ready to conduct additional experiments that validate these initial results. At the end of the day, the more insights you have, the better informed you’ll be as you choose where to take your innovation project. This way, you can make an informed pivot – based on collected data that holds up in real life.

If you’d like to learn more about our approach to experimentation and what it could look like at your organization, let’s chat.

Download the Experimentation Playbook and get started today.

Download playbook

Comments

Stay posted.

Keep up to date on what our clients have to tell, new tools, training and cool content.