Software Growth

Power user curve

The power user curve is a histogram of how many days per month each user is active, showing whether you have a core of heavy users, also called the L28 or L30.

The power user curve is a chart that shows, for a month, how many users were active on 1 day, 2 days, 3 days, all the way to every day. Where an average hides the spread, this chart shows it. Do you have a group of heavy users who come back daily, or does everyone dabble? That tells you what kind of product you have.

Where it comes from

The framework was written up by Li Jin and Andrew Chen at Andreessen Horowitz and is commonly used there when evaluating companies. It is also known as the activity histogram or the L30 (the name comes from Facebook's growth team). Variants use 28 days (L28) so every window is the same length and includes four of each weekday.

How to build it

  1. Pick a window: the last 28 or 30 days.
  2. For each user active at least once, count the distinct days they were active. Define "active" as a core action.
  3. Plot a bar chart: x axis is days active (1 to 28), y axis is the number of users.

Example: a team tool has 1,000 users active in the last 28 days. 400 were active on 1 or 2 days, 300 on 3 to 9 days, 150 on 10 to 19 days, and 150 on 20 or more days. The last group is 15% of users (150 / 1,000), and that is your power user segment.

How to read it

Chen and Jin describe a "smile" shape as healthy for many products: a peak of occasional users on the left, a dip in the middle, and a second rise at the right from daily users. The right-hand bump suggests a hardcore segment. A curve piled entirely on the left means weak habit. A curve tilted right means a sticky product, though it may also be a product people must use daily.

They also note that not every product needs high frequency. A product used once a month for a high-value job may do well with a left-heavy curve. What matters is the right shape for your use case, and whether it improves over time.

Relation to DAU/MAU

The DAU/MAU ratio collapses the histogram into one number, the average of days active divided by the month. The curve keeps the distribution, so you see the share of power users, not just the mean.

Ways to use it

  • Compare the curve by signup cohort to see whether newer users are more engaged.
  • Compare power users with the rest to find which features they use, then guide others toward them.
  • Cut the curve by plan or channel to see where heavy users come from.

For small SaaS

You can build it from a single SQL query on your events table. With a few hundred users, the shape is still readable. Pair it with revenue, since your power users should be your highest-retention customers.

Related terms

Sources

Back to the SaaS glossary