How to improve activation rate

People sign up for your SaaS, look around once and leave, and with only a trickle of signups, every one who leaves is a big share of your future paying users. To improve your activation rate, the share of signups who get real value from your product, you write down the one action that shows it, watch where each new user stops, and remove those steps one at a time. If you skip this and only chase more signups, most of the new ones leave the same way.

Do these three things

About 4 hours today, then 2 to 3 hours a day. Judge each fix after the next 20 signups reach that step, not after a few days.

Show the three actions in fullHide the full list

1. 1Today: write down the one action that shows a new user got value, mark every signup so far yes or no, and set up a free funnel in PostHog from signup to that action.
2. 2Every day: watch the first session of every new signup, offer every one a 15-minute setup call, and ship one fix to the step where most people stop. AI drafts the notes and helps write the fix; you watch, talk and decide.
3. 3Every Sunday: count your activation rate for signups who have had 7 full days, split by where they came from, and pick the one step to fix next week.

Get help doing these

ChatGPT or Claude reads this guide and works through it with you.

Do this with ChatGPT(opens in a new tab)Do this with Claude(opens in a new tab)

1. Pick the one action that means a new user is activated

A new user is activated when they do the one action that shows your SaaS helped them, within a set time after signup. Write that action down as one sentence before you change anything. In a product-led SaaS, where people try it before they pay, people who never do it rarely pay. Your activation rate is the number of signups who did it, divided by all signups, times 100.

You do not need an analytics team for this. With 2 paying users and a few who came back, a sheet is enough:

1. List every paying user. If you have none, list everyone who came back on 3 or more different days.
2. Next to each one, write what they did in their first 3 days, from your database or admin page.
3. List 5 people who signed up and never came back, and what they did.
4. Pick the one action that every person in the first list did and most people in the second list did not. It must be something your app can record.
5. Write it as one sentence: "A new user is activated when they [action] within [7 days] of signing up."

If nobody has come back yet, use the moment you would show in a two-minute product demo, the screen where people say "oh, nice". Check it in your next five customer interviews: ask "What would you need to see in the first 10 minutes to keep using this?"

Then add two columns to a sheet of every signup so far: "activated, yes or no" and "where they came from". Count the yeses. That is today's activation rate. This same action is the "first win" your onboarding emails point to, so write it once and use it in both places.

• Tool: your database or admin page, and Google Sheets. Free.
• Time: 45 minutes.
• You will have: one sentence, and a sheet of every signup with "activated: yes or no" and their source.
• It worked if: every paying user did the action, and fewer than half of the people who left did.
• Common mistake: picking a setup step, such as "verified email" or "finished the tour". It makes the number look good and tells you nothing. The other mistake is an action people do once and never again. In September 2026, a founder on Indie Hackers wrote that all 7 users who made a report made "exactly one report in a single sitting, and none have returned". If people do the action and still leave, pick a later one, such as the second time they use it.

2. Measure your activation rate and find the step where signups stop

To find where new users stop, record each step between signup and your activation action, and put them in a funnel: a chart of how many people reach each step. The step with the biggest drop is where you start.

Use PostHog. Its free plan includes 1 million events and 5,000 session recordings a month, needs no card, and stops at the free limit, so it cannot charge you by surprise. You need one project, which the free plan includes.

1. Install it. Sign up, then run PostHog's setup wizard from your app's folder: npx -y @posthog/wizard@latest. It adds the code for you and turns on session recording, which you use in step 5. If you are a developer who would rather build than sell, this part takes minutes.
2. Send one event per step. Write down the 3 to 5 steps a new user must take before the activation action, such as "confirmed email", "created a project" and "connected their data". Send one event at each, with posthog.capture('created_project'), plus one for the activation action itself.
3. Tag the source. Save where each signup came from, from the UTM tags on your links or a "How did you hear about us?" question after signup. Step 6 uses it.
4. Build the funnel. Click Product Analytics in the left sidebar, then + New Insights, then Funnel. Add your steps in order, from signup to the activation action. Set the conversion window limit to 7 days.
5. See who stopped. Click a step in the funnel to list the people who dropped off there. Save the list. These are the people you write to and watch in step 5.

• Tool: PostHog, free under 1 million events a month.
• Time: about 2 hours, most of it adding the events.
• You will have: a funnel with every step, and the number who stop at each.
• It worked if: a test signup you make yourself shows up at every step within 5 minutes.
• Common mistake: tracking 40 events, or only page views. You end up with charts nobody reads. Track the 3 to 5 steps and the activation action, nothing else.

3. Sign up as a stranger and cut every step before the first win

The fastest way to raise activation is to remove steps. Sign up to your own product with a new email address, on your phone and on a laptop, time every step until you reach the activation action, and then cut every step you can.

Make a new Gmail address. Open your home page and start a timer. Write down every screen, every field and every wait until you reach the activation action. Do it again on your phone, and note any screen that is slow to load: website speed counts inside your app too. Then cut, in this order:

1. The email check before people can start. Let people use the product first and confirm their email later. In May and July 2026, people on Hacker News wrote that a confirmation email "never arrives". If you use Supabase, two settings matter: hosted projects require a confirmed email by default (the "Confirm email" setting, under the email provider in Auth Providers), and Supabase's built-in email only sends to your own team, 2 emails an hour. Turn the check off until after the first win, or set up your own email sender before anyone signs up.
2. Form fields. Ask for an email address and nothing else at signup. Add Sign in with Google; in Supabase it is under Authentication, then Providers, then Google.
3. Questions before value. Move "company size", "role" and "how did you hear about us?" to after the activation action. Ask one at a time.
4. Waits and dead ends. Every link in your emails and every button on the first screen opens the exact next step, not a dashboard or your docs.
5. The card. If your free trial asks for a card, count it as a step. Whether to keep it is a separate decision: free trial or freemium.

Use ChatGPT or Claude to check your list: paste in your steps and ask "Which of these could a new user skip and still get [the activation action]?" A coding assistant can make most of these changes with you in an afternoon.

Date | Device | Step number | What I did | Seconds | Could this step go? (yes / later / no)

• Tool: a new Gmail address, your phone's stopwatch, and your sign-in settings. Free.
• Time: 1 hour for the test. Ship the cuts as your daily fixes in step 5.
• You will have: a log of every step and second, and a list of steps to cut, in order.
• It worked if: you reach the activation action in under 10 minutes, with no email check and no question before it.
• Common mistake: adding a product tour to explain the extra steps. Remove the step instead. A tour is one more thing to click past.

4. Fill the empty first screen with sample data and one clear button

When a new user first logs in, they should see a finished example of what your product makes and one button for the next step, never a blank page. A blank screen asks them to do the work before they know if it is worth it.

Pick one of these for your first screen:

• Sample data. Fill the first screen with a realistic example, clearly marked "Sample", so people see the result before they add their own. Ask ChatGPT or Claude to write 10 rows of realistic sample data for your kind of user, and check every row makes sense.
• A test action inside setup. Let them make the result happen on something you control. Sentry, the error-tracking tool, ends its setup guide with "Verify Your Setup": add a button that throws an error, so the first error report shows up straight away.
• Do it for them. If setup needs their files or another person's help, offer "Send it to us and we will set it up for you", or a link they can forward to a teammate or client. If your product needs data from a tool they already use, make connecting that tool the first button.

Then put one button on the screen: the next step, with how long it takes. Every other link is smaller. Each extra call to action pulls people away from the activation action, the same rule as on a landing page that converts.

• Tool: your app, and ChatGPT or Claude for sample data. Free.
• Time: about half a day.
• You will have: a first screen with an example and one button, never a blank page.
• It worked if: 4 of the next 5 people you watch in step 5 press the button without asking what to do.
• Common mistake: building a seven-item onboarding checklist first. In Userpilot's 2025 report on 188 companies, the median share of users who finished their onboarding checklist was 10.1%. One button that does the next step beats a list of seven.

5. Watch every new user's first session and offer every one a 15-minute setup call

Every day, watch the recording of every new signup's first session, write down the step where they stopped, and offer every one a 15-minute call to set it up with them. This is how you find out why people stop, which no chart tells you.

Every morning:

1. Watch every first session. In PostHog, open the list of people who dropped off that you saved in step 2, plus every new signup since yesterday, and watch each one's first recording. PostHog hides what people type into form fields by default. Add a line to your privacy policy saying you record how people use the app.
2. Write down the step and the reason. One row per person in a stop log: the step where they stopped, and what you saw, such as "clicked Settings 4 times" or "left at the install screen". A sheet is fine until it gets slow; then move to a CRM.
3. Offer every new signup a setup call. Send each one a short personal note offering 15 minutes to set it up together, with a free Cal.com booking link. AI drafts each note from your log; you add the one true line about what they did. Run each call like a discovery call: they share their screen and do it; you only ask questions. The full set of notes and follow-ups to stuck signups is in the onboarding emails guide; follow up up to 3 times.
4. Ship one fix every day to the step where most people in the log stopped: a cut from step 3, a better first screen from step 4, or one clearer sentence on the screen where they got lost.

If you get fewer than 5 signups a day, book 5 people a day who fit your ideal customer to try your product for the first time while you watch on a call. Find them with the 100 personal messages a day you send for signups; ask for 15 minutes "to watch you try it and see where it is confusing". Five first-time users a day show you more than a month of charts, and if they will not use it even with your help, you may have a product problem, not an activation one. Many of them become your first beta users.

If you get fewer than 10 signups a week even after this, your bigger problem is signups, not activation. If people visit but do not sign up, fix the site first.

Subject: [product] + [their company or project]Hi [first name],I saw you [one true line: what they did, such as "created a project"] yesterday and stopped at [the step]. That step trips up a lot of people, and it's on me to make it easier.Could I set it up with you on a 15-minute call? You share your screen, and we get you to [the activation action] together: [booking link]If a call isn't your thing, a one-line reply on where it got stuck helps me a lot.[Your first name]

• Tool: PostHog session replay (5,000 recordings a month free), Google Sheets, Gmail, and Cal.com's free plan for bookings.
• Time: 2 to 3 hours a day: 5 minutes a recording, 3 minutes a note, 15 minutes a call, and the rest on the day's fix.
• You will have: a stop log with a step and a reason for every signup, and one fix shipped every day.
• It worked if: at the end of each day, every signup older than 24 hours has a row in the log, and every one who has not activated has had a call offer.
• Common mistake: asking "Any feedback?" on the call, or watching recordings without writing anything down. Ask them to do the task, stay quiet, and write down the exact step where they hesitate.

6. Count activation every Sunday and fix one step at a time

Every Sunday, count how many people who signed up 8 to 14 days ago did the activation action within 7 days, split by where they came from, and pick the one step you will fix next week. This is your activation rate, and the one number this guide moves.

Wait until people have had 7 full days, or the number looks worse than it is. Fill in one row:

Week | Signups (8 to 14 days ago) | Activated within 7 days | Activation rate | Rate by source | Top stop step from the log | Fixes shipped | New paying users

Then use these rules:

• Fewer than 10 signups a week: activation is not your main problem. Spend most of your hours on getting your first 100 SaaS users, and keep watching every session.
• Activation under 40%: fix the step with the biggest drop in your funnel. Read that step's rows in the stop log first.
• One source activates far less than the others, such as 1 in 20 from a Product Hunt launch against 1 in 3 from your messages: those signups are the wrong people, not a broken product. Move your hours to the source that activates, and check your value proposition matches what that source was told.
• Most people activate, but few pay: activation is working. The next job is turning free trials into paying customers, your pricing and your pricing page.
• People activate, pay, then leave within 2 months: your activation action is set too low. Move it to a later step, work on reducing churn, and ask the people who left why, with a win-back email.

Judge a fix after the next 20 signups reach that step, not after a few days. With so few people, a normal A/B test cannot help you yet: to tell a 30% activation rate from 40% with the usual test, you need about 350 signups in each version. Until then, change one step at a time, watch the sessions, and count.

Once a month, put the rate in your monthly marketing review, next to your cost per paying user and your MRR. If the rate has not moved for a month, find out why growth stalled before you change more steps. People who activated quickly are the right ones to ask for a testimonial or a short case study, and their words are the best feedback from your first users.

• Tool: Google Sheets, and your PostHog funnel for the counts.
• Time: 30 minutes each Sunday.
• You will have: one row a week, and the one step to fix next week.
• It worked if: you can name this week's step and the rule that picked it.
• Common mistake: changing three steps at once. When the number moves, you cannot tell which change did it.

Your activation rate checklist

0 of 14 done

Work through this list with an AI assistant

It drafts the work and keeps track of what is done.

Do this with ChatGPT(opens in a new tab)Do this with Claude(opens in a new tab)

Step 1 · Pick the activation action

• Write the one action that means a new user is activated, and within how many days.
• Mark every signup so far activated yes or no, with where they came from.

Step 2 · Measure the rate and find the drop

• Install PostHog and send one event for each step and for the activation action.
• Build the funnel and find the step with the biggest drop.

Step 3 · Cut steps before the first win

• Sign up as a stranger on your phone and laptop, and time every step.
• Move the email check and the questions to after the activation action, and add Sign in with Google.

Step 4 · Fill the first screen

• Add sample data or a test action and one button for the next step.

Step 5 · Watch every first session

• Every dayWatch the first session of every new signup and log where they stopped.
• Every dayOffer every new signup a 15-minute setup call and hold every call booked.
• Every dayIf you get fewer than 5 signups a day, watch 5 people who fit try it on a call.
• Every dayShip one fix to the step where most people stop.

Step 6 · Count every Sunday

• Every SundayFill in the Sunday row.
• Every SundayPick the one step to fix next week using the rules.
• After 20 signupsJudge the last fix after the next 20 signups reach that step.

Copy listDownloadReset