#06: 3 steps to drive weekly experimentation

Here are 3 strategic steps that propel teams from nervous waiting periods into continuous product development.

Step 1: Anchor to a Learning Goal

Step 2: Design an Experiment

Step 3: Ask for the Evidence

You don't need permission from others to get started.

We have all been there. We have a new project at work. There are lots of unknowns and you are unsure how to get started. So, you wait.

• You wait for direction from the senior leader.
• You wait for budget approval.
• You wait until someone tells you that you have a good idea
• You wait until the “alignment meeting”
• You wait until the key stakeholder is back from vacation next week
• You wait for the perfect conditions to get started.

Focus on learning and experimentation instead

There is a better way to activate a team. And it involves new language and behaviors around learning goals and experimentation. Some of the benefits are:

• Creates action with the team
• Turns unknowns into knowns
• Creates a structure for the team during ambiguity
• Allows for safe risk-taking to build confidence

3 Steps to accelerate evidence-based experiments

Step 1: Anchor to a weekly learning goal

Here are 3 examples from projects I am leading/coaching and the blockers I experienced:

Example 1: Governance dashboards: waiting on senior executive buy-in on strategic direction

Example 2: Product Scoping Templates: debate on whether we have the right solution

Example 3: GenAI for Support data: waiting for budget approval from a business stakeholder

At first glance, each scenario would require the team to wait for approval, direction or buy-in. Yes, we need those from leadership but they don’t have to derail us from taking action in the week.

So, instead, ask your team:

What do you want to learn this week?

You can think about your learning goals through the lens of customer desirability, business viability or technical feasibility?

desirability: does your customer/user want it?

viability: is it valuable to the business/organization

feasibility: can you actually make it?

Based on the 3 examples above, here are my learning goals this week:

1. Governance dashboards: What kind of data do senior executives want to see for their product governance structure? Is this even the right audience?
2. Product Scoping Templates: Would an internal discovery team find a 1-page template helpful to clarify their product scope?
3. GenAI for Support data: Can a GenAI solution make the analyst team more efficient?

Framing these questions as learning goals creates a sense of curiosity and wonder. It helps surface the main unknowns in the project or product direction.

Time-boxing it to 7 days sets expectation of the amount of time we, as a team, are going to be thinking about this question.

Once you have the question framed, it is time to take a new approach on how to learn it:

Step 2: Design an experiment to gather evidence

I am very purposeful on the language of experimentation vs ‘the answer.’ Language matters here.

Experiments evoke imagery of a scientist in her lab trying different chemical combinations out to see what works. Trial and Error. Exploration. Making to Learn.

If you recall, all my project examples are waiting for approval or budget, so we have to be scrappy to test our learning goal to keep moving.

What we are trying this week:

Governance Dashboard

Learning Goal: What kind of information do senior executives want to see for their product governance structure? Is this even the right audience?

Experiment: Async message to 2 execs with Excel charts with fake data about their teams and ask them to take an action.

Product Scoping Templates

Learning Goal: Would an internal discovery team find a new 1-page Template helpful to clarify their product scope?

Experiment: A 55 min workshop the following week to get 5 teams to use the new Template and ask for live feedback.

GenAI for Support data

Learning Goal: Can a GenAI solution make the analyst team more efficient with their support database?

Experiment: Take the most popular support question and make up 3 different answers and feed them to analyst to see their reaction.

--

Note, all the above examples involve getting some first-hand evidence toward your learning goal. Push your teams to get real data - this is what is going to drive decision making and next iterations of your project or product.

Crafting the right experiment is part art and science. There are lots of frameworks and examples out there.

I am currently using the Say/Do framework in the new book: The Experimentation Field Book. 

Step 3: Ask for the evidence at next week’s meeting

One of two things will happen.

1. The will run the experiment
2. They won’t run the experiment

Both are helpful.

If they run the experiment, you have data to talk about! What did you learn? How did it go? Are we on to something? What are we missing? It will be a very constructive meeting and will help inform your next iteration to start back at Step 1: What do we need to learn this next week?

If they don’t run the experiment, you say: "Great! What prevented you from running it?"

Typically, reasons fall into a few categories:

• No access to data
• No access to customers/users
• I didn’t know how (capability)
• I didn’t have time (capacity)

As the team lead and meeting facilitator, you are now getting weekly evidence to evolve your product or surface team blockers.

Either one is helpful and actionable to de-risk your strategy your idea.