To be used by Got Users clients & community.

At a high level, you need to understand your customer and value props before you jump into brainstorming growth experiments.

Otherwise, you risk running the wrong experiments and drawing the wrong conclusions.

Here’s the high-level growth process we recommend.

Phase 1: Set North Star Metric (NSM)

The goal of growth is to get more people to experience the value of your product.

Growth strategy means coming up with experiments that help people ”do the thing” — experience that value — more.

To know if an experiment worked, a metric should improve. That metric should flow into your North Star Metric: the measurement of when people “do the thing”.

Examples

• Rides completed per week (Uber)
• Monthly active users (Facebook)
• Documents sent per month (DocuSign)
• Daily active users (Slack)

Bad Examples

• Signups. What if people don't actually take a ride, order food, etc.?
• Revenue. Revenue happens because of your NSM.

We have serious horror stories about teams picking the wrong NSM. Choose wisely.

Phase 2: Clarify User Personas

Growth depends on users. Products don’t grow if people don’t use them.

So, to come up with a growth strategy, you first have to understand your ideal user, how you solve their problems, and how they hear about you. Then, you can prioritize experiments to get more of these people.

Who are your ideal users? They’re the ones that retain — they experience your core value over and over.

This means you need to answer the question: who is doing your NSM the most?

For example, if your NSM is # of monthly events thrown:

1. Pull a list of the people who throw events the most.
2. Interview them. Paint a picture of who they are and how they ended up using you.
3. After your interviews, fill out this value props sheet and these persona slides:

You'll generally need to interview 10 to 20 users. Stop when you hear the same things again and again.

In many cases, your product person will have some answers already.

Without answers, you'll likely have the wrong value props, you risk running the wrong experiments, and you may write off channels that would otherwise work.

Phase 3: Monitor Main Growth Funnel

Measure every major step of the funnel, along with CAC.

For example:

• Traffic
• Sign Up
• Activation Event
• North Star Metric
• Revenue
• Referral

It should look something like this. Here's a template you can copy.

This usually starts in a spreadsheet, but you’ll eventually move to a data tool like Improvado, Databox or Google Data Studio so you don’t have to manually update it every week.

Phase 4: Brainstorm Growth Loops

Most high-growth products feed outputs from the product as inputs into acquisition channels.

Examples

Figure out your growth loop if one applies and start measuring the major steps.

(Your growth loop is not always obvious! You may have to research & experiment before you find it.)

Phase 5: Brainstorm Experiments

Example experiments:

• A/B test welcome email copy
• Launch Facebook ads targeting teachers offering an academic discount
• Internationalize in-product copy for Spanish-speakers

Some of these depend on the CAC you’re willing to spend. Here's a good way to think about channel-CAC fit.

If you haven't monetized your product, you may need to rule out paid channels and focus more on in-product changes, content, & viral marketing.

Every experiment should have a metric that determines whether it succeeded or failed — so you can build on the learnings from it. Otherwise you’ll be spinning your wheels.

As a reference: here's an Airtable template we frequently use to prioritize growth experiments. Feel free to copy.

Phase 6: Prioritize

Phase 7: Run Experiments

May require resources from other teams. If growth is actually a priority, you should be able to get dev time, sales time, etc.

Process: Growth Meetings

Here’s a general framework for growth meetings. Most teams do this every week.

1. Review relevant learnings
• We’ll double-click on the learnings you sent us through Slack. (E.g., “Only VPs responded to your user interview email; tell us more about that.”)
2. Post-mortem-ing & debugging
• We’ll look back at why experiments were (or weren’t) run effectively and debug any blockers getting in the way. (E.g., ”Why did it take so long to design ads?”)
3. Strategizing & prioritizing new experiments
• After the first couple weeks, we’ll brainstorm experiments that build on learnings. (E.g., ”The ‘7-day trial’ messaging is converting better on the pricing page, so lets A/B test it on the homepage too.”)
• They should usually improve a growth loop — so the results compound over time. (E.g., ”Let’s add a popup after the magic moment to ask for a referral” is usually better than “Let’s do a PR push.”)
• Once we brainstorm experiments, we’ll prioritize them using RICE, learnings, and patterns we’ve seen across other clients.

We recommend setting a timer to timebox each part.