Creating Momentum with Marketing Micro-Experiments

Marketers are drowning in data.

Open rates. Impressions. Time-on-page. Bounce rates. Conversion rates (if they're lucky). Reach. Shares. Dashboards that keep on growing until you run out of Hubspot reports 😉

And yet, week after week, most marketing teams keep making the same things, in the same ways, for the same audiences. The data sits in a tab. And the content calendar marches on.

That's not a data-volume problem. That's a data-doing problem.

Because data collection, on its own, does nothing; except pile up to the point where you can't even process it any more.

Data only matters when it changes what you do next.

And how do we change what we do next?

Well, first, it takes a decision. And second, it takes an experiment.

Part 1: Making the Decision

Here's the uncomfortable truth about most marketing dashboards: they're not waiting on more data. They're waiting on a decision.

We've gotten really good at collecting. Really good at reporting. Really good at sitting in a weekly meeting where someone says, "interesting, let's keep an eye on that," and then nobody does anything different on Tuesday.

A decision is the moment you stop letting the data be interesting and start letting it be useful.

And deciding to do something doesn't mean deciding to launch a campaign. It means deciding to stop treating a piece of data like wallpaper and start treating it like a question worth answering.

So before we get to creating an experiment, let's start with the decision. Because you don't need to act on everything. In fact, it's unwise from a functional hypothesis point of view. But you do need to act on SOMETHING, and the below questions are a great place to start:

• What's one piece of data I've looked at more than three times in the last month and still haven't done anything with?

• What's the thing in my dashboard that surprises me, even a little, when I see it?

• Where is the gap between what I expected to see and what I'm actually seeing?

• What would I want to know more about if I had nothing else on my plate?

• What am I afraid the data might be telling me?

That last one is usually the most honest; which can also mean that it's the most difficult to answer. After all, humans tend to avoid the data that would require us to change something we've already committed to.

Try picking one. Just one. It doesn't even have to be the most important thing (we'll get there). The practice is to pick ONE thing you're actually going to do something about.

Part 2: Committing to the Decision

Two things separate a decision from a passing thought:

It has a verb attached to it. "I'm curious about our LinkedIn engagement" isn't a decision. "I'm going to find out what's driving the difference between our top and bottom posts" is.

It has a deadline attached to it. Not "someday." This week. Because momentum, by definition, requires motion. And motion requires a start date.

If your decision doesn't have a verb and a deadline, it's still a thought. Which is fine. But thoughts don't change what gets created on Wednesday.

Part 3: Asking the Right Question

Once you've decided to do something, the next move isn't to plan a campaign. It's to design a micro-experiment.

A micro-experiment is a small, time-boxed test of one thing you believe might be true, designed to give you information, not validation. Let's clarify the difference between those two things👍

Information means you're trying to learn. Validation means you're trying to be right. Those are two very different goals, and they produce very different experiments.

When marketers run experiments to be right, they design them to succeed. Big sample sizes. Long timelines. Everything optimized to confirm what they already planned to do.

When marketers run experiments to learn, they design them small enough to actually finish, sharp enough to actually answer something, and honest enough to tell them when they were wrong. Because after all, from a scientific measurement point of view, you don't have a "good hypothesis" (and subsequent test), if you don't have a 50% chance of being wrong 💡

Personally, I get excited about the possibility of what we could learn in 7 days. Why 7 days? Because it's long enough to get a real signal, yet short enough that you remember why you started. And it's cool to be able to see how different something can be - with intention - within only ONE WEEK.

Part 4: Designing a Meaningful Experiment

A meaningful experiment isn't a complicated one. It's one where you can clearly answer four things before you start.

1. What do I think is true?

Say it as a sentence. Out loud, ideally (because you get to taste your words). If you can't finish the sentence "I think ____ because ____," you don't have a hypothesis yet, you have a hunch. Hunches are great starting points, but they're not testable.

• What story am I telling myself about what's happening in the data?

• What would have to be true for that story to make sense?

• What's the part of this I'm most sure about? The part I'm least sure about?

2. What's the smallest version of a test that could confirm (or deny) that assumption?

This is where most marketers over-engineer. They turn a question into a campaign, a campaign into a quarter, and a quarter into a reason not to start.

• What's the simplest thing I could change, add, or try this week?

• If I only had one post, one email, one conversation to test this, what would I do?

• What would I need to stop doing to make room for this test?

• What's the version of this experiment that I could launch by Wednesday?

If you can't launch by Wednesday, your experiment is too big. Shrink it.

3. What signal will tell me something useful?

A signal isn't a KPI. A KPI is a number you're held accountable to. It's not even a metric necessarily (something you're used to tracking). A signal is a piece of information that helps you decide what to do next.

• What's the one thing I'll look at on day 8?

• What would I expect to see if my hypothesis is right?

• What would surprise me?

• What would change my mind?

Decide your signal before you run the test. Otherwise, you'll find a way to read the results in favour of whatever you wanted to believe.

4. What will I do with what I learn?

This is the question that most experiments are missing. And it's the one that turns a test into momentum.

• If the answer is yes, what's my next move?

• If the answer is no, what's my next move?

• If the answer is "I'm not sure," what would I need to test next?

You're not just designing an experiment. You're designing the decision that comes after it. Because the decision after the experiment is where momentum actually lives.

Part 5: Compounding Experiments

One micro-experiment doesn't change a marketing team. A practice of micro-experimenting does.

Not a check box. A shift in DOING.

When you run one every week, or every other week, or whenever you spot something worth testing, you stop being a team that produces content and start being a team that produces learning. The output on the calendar might look the same. The thinking behind it is completely different.

You build a library of things you actually know about your audience, your channels, and your message; not things the industry told you to believe.

You stop hiding behind dashboards and start using them.

And you trade the exhausting MORE of constant creation for the energising MORE of compounding momentum. Same effort. Completely different return.

More on the MORE Trade

If you've been reading anything I write, you know I'm not anti-MORE. I'm anti-MORE-of-the-wrong-thing. MORE blog posts. MORE channels. MORE campaigns. MORE dashboards. MORE meetings to talk about the dashboards. That's the MORE that's burning marketers out and producing thinner and thinner results.

The MORE we actually want is MORE momentum. MORE learning. MORE of what's working. MORE of what we couldn't have predicted from a planning doc six months ago.

And that bridge between "data we collected" and "momentum we built"? It fits inside a week.

The Next 7-Days

Ready to try a micro-experiment with your marketing data? Add the below To-Do's to your calendar:

• Today: Pick the one piece of data you're testing. Make a decision about it. Give your decision a verb and a deadline (7 days from now).

• Tomorrow: Write your hypothesis in one sentence. Design the smallest test you can launch by end of day. Make sure you've defined your signal!

• Day 7: Look at your signal. Decide what to do next.

• Day 8: Do the thing you decided to do next ;)

That's it. That's the practice.

It's simple really. And the marketers building real momentum are the ones that get good at executing the simple things at speed.

They're the ones who turned what they noticed on Monday into something they learned by next Monday.

And then they did it again.

And that could be you, too.

Regardless of whether you are a marketer, a marketing leader or an entrepreneurship doing all your own marketing, this methodology is available. And if you're looking at your marketing results and are wishing something looked different about them, you can start practicing today!

And if you'd like some support, you can find it today with the Authentic Marketing Fellowship! In particular, a conversation with Stuart Webb will probably help 🎉