How to measure growth loops
What makes growth loop metrics different from funnel metrics
Funnel metrics measure a one-way journey. A prospect enters at the top, moves through awareness, consideration, and conversion, and exits at the bottom as a customer. Each stage has a conversion rate, and the whole thing is linear: what comes out the bottom is done. To grow, you pour more in at the top.
Growth loops work differently because the output of the loop becomes the input for the next cycle. A user signs up, takes an action that produces something (an invite, a piece of content, a shared link), and that output brings in new users who then do the same thing. Because of this circularity, the questions you ask about performance change. With a funnel, you ask: how many people made it through? With a loop, you ask: does each cycle produce more input than it consumed, and how fast does that happen?
This distinction matters because loop metrics are multiplicative rather than additive. A funnel improvement adds users. A loop improvement compounds them. If your loop produces slightly more than it takes in, growth accelerates over time without extra spend. If it produces slightly less, growth stalls no matter how much you push at the entry point. Measuring loops means tracking the relationship between cycles, not just counting outcomes at fixed stages. You need three core measurements: how long a cycle takes, how efficiently each stage converts, and whether the loop nets positive or negative per turn.
The loop cycle time: how fast one iteration completes
Cycle time is the elapsed time from when a new user enters the loop to when their activity produces a new entrant. It is the clock speed of your growth engine. Two loops with identical conversion rates will grow at wildly different speeds if one completes in two days and the other takes two months.
To measure it, define the start and end events precisely. For a referral loop, the start might be account activation and the end might be the moment an invited friend activates. For a content loop, the start could be a user publishing a post and the end could be a new visitor arriving from that post and creating their own account. Track the median time between these events, not the average, because a few very long tails will distort an average and hide the typical experience.
Cycle time is often the most overlooked lever. Teams obsess over conversion rates while ignoring that shortening the loop from 30 days to 15 days effectively doubles the number of cycles per year. Practical ways to reduce cycle time include removing friction between the value moment and the sharing moment, prompting the output action sooner, and reducing the delay before a new user reaches their own output-producing action. Even a modest reduction here compounds meaningfully across many turns.
Loop conversion rate: measuring output per input at each stage
A loop is a chain of steps, and each step has a conversion rate. Breaking the loop into stages lets you find where users leak out before they can contribute to the next cycle. Consider a viral loop with four stages: a user becomes active, the active user shares an invite, the recipient clicks the invite, and the clicker converts into a new active user.
Measure each transition separately. If 100 users become active, how many send at least one invite? Of the invites sent, how many are clicked? Of those clicks, how many convert? Multiplying these stage rates together gives you the overall loop conversion, but the individual numbers tell you where to focus. A loop might have a healthy sharing rate but a weak invite-to-click rate, pointing you toward the invite's messaging or channel rather than the sharing prompt.
Be careful to measure per input rather than in aggregate. The useful question is not how many total invites were sent, but how many invites the average active user generates, and how many new active users each of those produces. This per-user framing keeps your metrics comparable as your base grows and prevents you from mistaking raw volume for loop efficiency. A loop can send more invites in total simply because you have more users, while the efficiency per user quietly declines.
The amplification factor: are you gaining or losing users per cycle
The amplification factor is the single number that tells you whether a loop is alive or dying. It answers: for every user who enters the loop, how many new users does their activity produce? If the number is above one, each cycle grows the base and the loop is self-sustaining. If it is below one, the loop shrinks with each turn and needs constant external fuel to keep running.
Calculate it by dividing the number of new users produced by a cohort's loop activity by the size of that original cohort. If 1,000 users generated activity that brought in 400 new users, the amplification factor is 0.4. That loop is sub-viral: it amplifies acquisition from other channels but cannot grow on its own. A factor of 1.2 means the base grows 20 percent per cycle before accounting for churn.
Most real loops sit below one, and that is fine. A loop with a factor of 0.5 still doubles the effectiveness of every user you acquire elsewhere, because each paid or organic signup triggers a chain that produces additional users. What matters is knowing the number and understanding what it does to your economics. Combine amplification with cycle time to project growth: a factor of 0.8 with a five-day cycle can out-produce a factor of 1.0 with a ninety-day cycle over a meaningful horizon.
Retention and reinvestment: tracking whether outputs feed back in
A loop only continues if users stick around long enough to complete their part and if the outputs actually re-enter the system. Retention is therefore not a separate concern from your loop metrics; it is a structural input to them. A user who churns before producing an output is a broken link in the chain.
Track retention specifically against the loop-relevant action, not just general activity. For a content loop, the question is not whether users log in, but whether they keep publishing, because publishing is what feeds the loop. For a referral loop tied to ongoing usage, you want to know whether referred users become referrers themselves, and at what rate compared to organically acquired users. This second-generation behavior determines whether the loop sustains across generations or fizzles after the first wave.
Reinvestment measures whether the value a user generates flows back into the loop rather than leaking away. If your loop depends on users reinvesting time, content, or connections, track the proportion who take that reinvestment step and how it changes over their lifetime. A loop that produces strong first-cycle outputs but where users rarely return to produce a second output will grow in a burst and then flatten. Watching cohort behavior across multiple cycles reveals this pattern early, while a single-cycle snapshot hides it entirely.
Connecting metrics to loop health with a simple measurement framework
Individual metrics are only useful when assembled into a picture of loop health. A workable framework tracks four things together: cycle time, stage conversion rates, amplification factor, and loop-specific retention. Read as a set, these tell you not just whether the loop is working but why, and what to change.
Start by mapping your loop's stages explicitly, then instrument the event that marks each transition. Establish a baseline for each metric so you can detect movement. When growth slows, this framework lets you diagnose the cause: a rising cycle time, a specific stage that has begun leaking, a falling amplification factor, or retention decay. Each points to a different fix, and without the breakdown you would be guessing.
The practical rhythm is to review these metrics per cohort on a regular cadence. Cohort analysis is essential because loop performance changes as the user base and market evolve. An amplification factor that was healthy when your users were early adopters may drop as you reach a broader, less-motivated audience. Watching cohorts over time turns your measurement into an early warning system rather than a post-mortem. The framework does not need to be elaborate; a small dashboard with these four dimensions, segmented by cohort, is enough to guide most decisions.
Common measurement mistakes and how to avoid them
The most frequent mistake is measuring loop outputs in aggregate rather than per user. Total invites, total shares, and total signups all rise as your base grows, which can mask a loop that is becoming less efficient per person. Always normalize to a per-user or per-cohort basis so growth in the base does not disguise decline in the mechanism.
A second mistake is ignoring cycle time and focusing only on conversion. Teams celebrate a higher share rate while a lengthening cycle quietly erodes the compounding effect. Track both, because the interaction between them determines actual growth. A related error is using averages where medians belong; long tails in cycle time and skewed distributions in output per user routinely mislead teams who rely on means.
Third, many teams attribute all resulting growth to the loop when other channels are also feeding it, inflating the perceived amplification factor. Be disciplined about isolating loop-driven acquisition from paid and organic sources. Finally, avoid measuring the loop only once and assuming the number holds. Loops decay as markets saturate and as your audience broadens beyond early enthusiasts. Re-measure on a cadence, watch cohorts age, and treat any single reading as a snapshot rather than a permanent truth. Correcting these mistakes usually costs nothing beyond attention and clearer definitions, and it prevents you from optimizing the wrong thing.
Example
Core growth loop metrics and what each one reveals
| Metric | What it measures | Diagnostic signal |
|---|---|---|
| Cycle time | Elapsed time for one loop iteration to complete | Slowing cycles reduce compounding even at steady conversion |
| Stage conversion rate | Output per input at each transition in the loop | Pinpoints where users leak before contributing |
| Amplification factor | New users produced per user entering the loop | Above one means self-sustaining; below one means sub-viral |
| Loop-specific retention | Whether users survive to produce and reproduce outputs | Second-generation drop-off predicts a flattening loop |
| Reinvestment rate | Share of value that flows back into the loop | Low reinvestment causes burst-then-plateau growth |
FAQ
What is a good amplification factor for a growth loop? There is no universal target. A factor above one means the loop grows on its own, but many successful products run loops below one that still multiply the impact of other acquisition channels. What matters is knowing your number, pairing it with cycle time, and understanding how it affects your economics rather than chasing an arbitrary threshold.
How is measuring a growth loop different from measuring a funnel? A funnel is linear and additive: you count how many people pass through each stage. A loop is circular and multiplicative: its output becomes its next input, so you measure the relationship between cycles. The key loop questions are how fast a cycle completes and whether each cycle produces more input than it consumed.
Why should I track cycle time if my conversion rates are strong? Cycle time is the clock speed of your loop. Two loops with identical conversion rates grow at very different speeds if one completes in days and the other in months. Shortening the cycle increases how many times the loop turns in a given period, which compounds growth even without any change in conversion.
How often should I re-measure my loop metrics? Review them on a regular cadence and analyze by cohort. Loop performance shifts as your audience broadens and the market saturates, so an amplification factor that was healthy with early adopters can fall later. Treating any single reading as a snapshot and watching cohorts age turns measurement into an early warning system.
Read next
Learn more