.png)
Frameworks, core principles and top case studies for SaaS pricing, learnt and refined over 28+ years of SaaS-monetization experience.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Mastering User Activation How to Track Activation Rate by User Segment
Most product teams can tell us how many users signed up last month. Far fewer can say which new users reached meaningful value, how quickly they did so, and whether the answer differs between a five-person startup, a 5,000-person enterprise, an administrator and an invited collaborator. The distinction matters because a rising company-wide activation rate can coexist with deteriorating activation inside every important customer group. Behavioural data can look healthier simply because the mix of new users changed, a statistical problem documented in research on disaggregated behavioural data.
The stakes extend well beyond a product dashboard. In a 2022 peer-reviewed study of SaaS free trials, more frequent product use was associated with higher conversion, while using a wider variety of features was associated with lower conversion. In other words, “more activity” was not synonymous with “more value”. A separate field experiment involving 2,673 cloud customers found that early customer education halved first-week churn and lifted cumulative usage over the following eight months by 46.57%, with the strongest effect among customers less familiar with the provider.
Monetizely's position is that activation rate should never be managed as one company-wide percentage. The operating metric should be a segment-specific cohort rate: the percentage of eligible new users in a defined segment who reach an evidence-backed value event within a fixed period. The company-wide figure is a roll-up for reporting, not a number from which to run the business.
Activation becomes useful only when it measures value rather than motion
A signup is an acquisition event. A login is an access event. Neither proves that the customer has received the benefit for which the product exists.
Slack made this distinction unusually visible in its 2019 S-1. It defined a daily active user as someone who created or consumed content during a 24-hour period, rather than somebody whose application happened to be open. During the three months ended January 2019, Slack reported more than 10 million daily active users; users at paying customers also spent more than 90 minutes actively using Slack during a typical workday. Slack treated depth of engagement separately from merely being connected.
Amplitude CEO Spenser Skates made the same point more sharply on the company's 6 May 2026 earnings call: “a weekly active user in Amplitude is a user who saves a chart”, rather than somebody who merely logs in. The precise action will differ by product, but the management principle is sound: define activity around behaviour that represents useful work.
A good activation event therefore has three properties. The user can clearly complete it, the event normally occurs near the beginning of the relationship, and historical data show that users who complete it subsequently retain, expand or convert at a meaningfully higher rate.
The distinction is easier to see when the common definitions are placed side by side.
| Activation definition | Example | Ease of measurement | Strength as an operating signal | Main failure |
|---|---|---|---|---|
| Account created | User completes registration | Very high | Low | Measures acquisition, not value |
| Setup completed | CRM imported; integration connected | High | Medium | Setup may precede any useful outcome |
| First-value event | Report saved; project completed; teammate successfully invited | High | High | Must be validated against later behaviour |
| Repeated-value event | User reaches first value several times within the activation window | Medium | Very high | Can delay the signal too long for fast products |
The table points to a practical standard: first value should normally anchor activation, while repeated value is better used as a confirmation metric. Neither should be selected by intuition alone.
The evidence argues against feature-count shortcuts. Hongshuang Li's 2022 Production and Operations Management study found that frequency and variety of SaaS free-trial use had different relationships with conversion: higher usage frequency encouraged subscription, while greater variety was associated with lower conversion. A user who deeply completes one core workflow can therefore be more activated than someone who samples ten menus.
Timing matters as well. A 2025 Journal of the Academy of Marketing Science study analysing 74,871 B2B SaaS subscription contracts found that selling more add-ons during onboarding could lower retention through higher perceived complexity, while the relationship changed after onboarding. The same behaviour can carry a different meaning at a different stage of the customer journey.
Segmentation before aggregation stops customer mix from masquerading as product progress
Monetizely's 5-Step Pricing Framework starts with Goals and Segmentation, then moves through Packaging, Pricing Metric, Rate Setting, and Operationalisation. The sequence matters. Management first decides what it is trying to achieve and which customer groups have materially different needs; it then determines what each group receives, what unit best reflects value, what rate should apply, and whether the company can quote, meter, bill and report the model reliably. As explored in Monetizing Agentic AI, the framework is designed for monetisation decisions, but its first step exposes the same error in activation measurement: combining unlike customers before understanding them causes every decision downstream to rest on an average that may describe nobody particularly well.
For activation, segmentation should not mean building hundreds of dashboard cuts. The goal is to identify groups whose path to value is genuinely different.
Five segmentation dimensions usually deserve testing first.
| Segment dimension | Typical B2B split | Why activation may differ | Priority |
|---|---|---|---|
| Customer size | SMB, mid-market, enterprise | Enterprise users often face procurement, integrations and administration before value | Very high |
| User role | Admin, creator, manager, collaborator | Each role receives a different benefit from the same product | Very high |
| Primary use case | Reporting, workflow, collaboration, automation | The first meaningful outcome changes with the job being done | Very high |
| Acquisition route | Self-serve, inbound sales, outbound sales, partner | Expectations and pre-sale education differ | High |
| Package | Free, entry paid, growth, enterprise | Entitlements and onboarding support change what users can accomplish | High |
The table means segmentation should follow differences in how value is reached, not whatever demographic fields happen to be available in the CRM.
For B2B products, we would also separate user-level and account-level activation. A five-user account in which one administrator completes setup but nobody else adopts the workflow is not equivalent to a five-user account in which four members repeatedly use the product. Dropbox's 2018 filing illustrates why the distinction is real: more than 30% of its paying users were on team plans, and its definition of a paying user counted licences rather than unique human beings, so one person with two paid licences could count twice.
At the user level, the question is: did this person experience value? At the account level, it becomes: did enough of the buying organisation experience enough value to make continued adoption likely?
Ignoring that distinction invites aggregation errors. Alipourfard, Fennell and Lerman's 2018 peer-reviewed work on Simpson's paradox showed how behavioural trends can reverse after data are disaggregated into meaningful subgroups, demonstrating the method on Stack Exchange, Khan Academy and Duolingo data. A product leader should assume the same risk exists in activation data until the major segments have been inspected.
Fixed cohort windows make activation comparable enough to manage
Once the value event and segments are defined, activation needs a denominator and a clock.
For segment (s), signup cohort (c), and activation window (t):
Activation rate = users in segment (s) who reach the activation event within (t) ÷ eligible new users in segment (s)
“Eligible” needs to be fixed before measurement. An invited collaborator who cannot configure an enterprise integration should not be judged on an administrator's activation event. By the same logic, users who fail to activate should not be removed from the denominator simply because their behaviour makes the metric look worse.
The clock solves another common reporting mistake. Suppose enterprise activation normally takes 30 days. Comparing users who signed up 35 days ago with users acquired last week quietly gives the older cohort four additional weeks in which to succeed. The newer cohort will appear worse even when nothing has changed.
A mature activation dashboard should therefore compare users after equal exposure periods: Day 7 against Day 7, Day 30 against Day 30. Leading windows may be useful for operational alerts, but management should not mix incomplete cohorts with mature ones.
Consider how customer mix alone can produce an apparently strong improvement:
| 30-day cohort | SMB self-serve users | SMB activation | Enterprise users | Enterprise activation | Overall activation |
|---|---|---|---|---|---|
| Earlier period | 6,000 | 60% | 4,000 | 30% | 48.0% |
| Later period | 8,500 | 58% | 1,500 | 28% | 53.5% |
| Change | +2,500 | -2 pts | -2,500 | -2 pts | +5.5 pts |
Every segment deteriorated, yet the headline activation rate increased from 48.0% to 53.5% because the business acquired more users from the higher-activating segment.
That is why Monetizely's position is stronger than “segment your dashboard”. The segment rate is the performance measure; the aggregate rate is mathematically downstream of both segment performance and acquisition mix. Treating the latter as the primary KPI confuses two different management questions.
The operating dashboard should retain the component data behind every percentage:
cohort start date and maturity date;
eligible users or accounts in the denominator;
number activated and activation rate;
median time to activation;
downstream retention, conversion or expansion for activated versus non-activated users.
Without the denominator, a rate hides scale. Without time-to-activation, two cohorts with the same eventual activation can have dramatically different onboarding experiences. Without a downstream outcome, the event remains an untested theory of value.
Public SaaS companies rarely report “activation rate” as such, but their filings reveal something more useful: sophisticated operators segment the business whenever customer groups differ economically.
Zoom provides an unusually clear case. In fiscal 2021, the company said customers with 10 or fewer employees had expanded sharply during the pandemic. Their share of Zoom's revenue doubled to 36% from 18% in the prior year, materially changing the company's customer mix. A company-wide product metric could have moved during that period even if behaviour inside each customer-size cohort had remained unchanged.
The pattern repeats across SaaS.
Company and disclosure date What management separated Evidence Lesson for activation Slack, 2019 Active users, paid organisations, large customers >10m DAU; >88,000 paid customers; paid users averaged >90 active minutes per workday Behavioural depth and commercial status are different dimensions Zoom, FY2021 Customer size ≤10-employee cohort rose from 18% to 36% of revenue Mix shifts can radically alter aggregate statistics ZoomInfo, 2019 Enterprise, mid-market, SMB Net annual retention was 127%, 112% and 87%, respectively One retention average would conceal radically different customer economics monday.com, 2025 Enterprise versus smaller accounts 29% of >$50k ARR customers had multiple products versus 6% below $50k ARR Expansion behaviour changes with account scale Dropbox, 2018 Team and individual licences >30% of paying users subscribed to a team plan User type changes what “successful adoption” should mean The recurring message is straightforward: customer structure is not dashboard decoration. It determines what good behaviour looks like.
Executives themselves increasingly talk in these terms. On 7 May 2026, Dropbox CEO Drew Houston pointed to “funnel and product improvements in Teams”, separating the team motion from work on individual retention. In February 2026, monday.com co-CEOs Roy Mann and Eran Zinman said “larger customers increasingly adopt more solutions”; the company's 2025 filing quantified that difference through multi-product penetration.
Even broad usage metrics need context. Atlassian CEO Mike Cannon-Brookes reported “reaching 2.3 million AI monthly active users” for fiscal 2025. MAU is valuable for measuring reach, but a product team managing activation would still need to ask what those users accomplished, how first value differed by product and role, and whether the behaviour predicted continued adoption.
Amplitude gets the measurement plumbing right but its commercial meter still needs one final simplification
Amplitude is a useful single-brand case because measurement is not peripheral to what it sells. Its platform ingests behavioural events, resolves identities, builds cohorts and connects customer behaviour to engagement, growth and loyalty. As of 31 December 2025, Amplitude reported 4,797 paying customers; management described a land-and-expand model in which customers begin with a use case and add data, users, products and teams as value becomes visible.
That makes its own pricing architecture particularly instructive. A company that tells customers to measure behaviour precisely also has to decide which behaviour should determine the customer's bill.
Against the three relevant steps of Monetizely's 5-Step Pricing Framework, our scorecard is:
Framework step Grade Monetizely's assessment Packaging A- Free, Plus, Growth and Enterprise create a clear maturity path, while the current platform page gives every tier access to the broader product suite subject to limits and advanced features. Pricing metric B Event volume is measurable and visible on the live self-serve page, but current billing documentation still discusses MTU or event-volume charging even after management said in August 2026 that Amplitude had moved to a single meter. Operationalisation B+ Pay-as-you-go rules, usage entitlements and overage treatment are documented, and management says quoting has become simpler; public documentation has not yet fully caught up with the commercial reset. Amplitude gets the most important design choice right: package capability and sophistication separately from raw user seats. Where it falls short is public consistency around the meter.
The evolution explains why.
The structure is moving in the right direction. Amplitude's 5 August 2026 earnings call is particularly telling: CEO Spenser Skates said the company had “reduced down to a single meter”, while CFO Andrew Casey said the new model gives customers greater cost predictability and makes quoting easier.
What does Amplitude get wrong? Its public materials still expose remnants of the old dual-meter system. The August 2026 pricing page leads with events, yet the current billing guide continues to explain MTU-based quotas and an event-per-MTU guardrail. For a small buyer, that creates exactly the question good monetisation design should remove: which unit should we actually use to predict our bill?
Monetizely's recommended next reset is therefore one committed move: finish standardising new business on event volume as the published primary meter, retire MTU as a choice for newly sold plans, and make included event bands, overage rules and enterprise commitments consistent across the pricing page, documentation and quoting systems.
We would not price Amplitude on “activated users”. Activation is the right management outcome but a poor universal billing unit because every customer can legitimately define value differently. Events are observable, auditable and native to the data architecture; activation should sit above them as the customer's business KPI.
The distinction is central to this article. What we meter for billing and what we manage for customer success do not need to be the same unit.
Management discipline turns segment activation from analytics into growth
A segmentation dashboard earns its place only when teams make different decisions because of it.
Suppose enterprise administrators activate at 42%, SMB creators at 68%, and invited collaborators at 74%. The wrong response is to launch one onboarding redesign aimed at moving the company-wide average. Enterprise administrators may be blocked by data integration and permissions, while invited collaborators may need nothing more than a clear invitation and a useful shared object.
The 2016 cloud-service field experiment is instructive here. Providing early guidance cut first-week churn by half, but the effect was strongest for customers with less previous experience with the provider. A uniform intervention would have hidden where the treatment created the most value.
Before teams promote an activation KPI into the operating cadence, we would require five checks:
The event corresponds to a real customer outcome, not registration or passive access.
Activated users show stronger subsequent retention, conversion, expansion or recurring usage than non-activated users.
Each major segment has enough observations to produce a stable rate.
Cohorts are compared at the same maturity, with late-arriving events handled consistently.
Product analytics, CRM and billing identities reconcile well enough to connect individual behaviour with account economics.
Once those conditions hold, activation becomes a bridge between product and monetisation rather than another product-team metric. Product can see where customers stall. Customer success can prioritise intervention. Marketing can compare acquisition sources on the quality of customers they create rather than signup volume alone. Finance can distinguish growth caused by better onboarding from growth caused by acquiring more customers from naturally high-activation segments.
The public data reinforce that discipline. ZoomInfo's 2019 net annual retention ranged from 87% in small business to 127% in enterprise; monday.com's 2025 multi-product penetration was almost five times as high among >$50,000 ARR customers as among smaller accounts. A single product KPI sitting above customer groups with economics that different is too blunt for resource allocation.
The final move is organisational rather than analytical. Management must decide which activation gaps deserve capital.
Choose one customer segment whose activation gap matters enough to change the annual plan. A five-point enterprise improvement may be worth far more than a ten-point gain among low-retention free users.
Allocate product and customer-success capacity using the economic value of closing that gap. Rank opportunities by the customers, retention and expansion affected, rather than by which onboarding problem generates the most support tickets.
Put segment mix into forecasts. When the sales plan moves upmarket, enters a new geography or shifts from self-serve to sales-led acquisition, the activation forecast should change with it rather than assuming last year's blended rate survives.
Require operating reviews to explain both the segment rate and the mix effect. An executive should be able to see whether a movement in headline activation came from a better product, a different customer mix, or both.
Retire activation events that stop predicting durable value. Products change, packages change and customers learn new workflows. A metric that was predictive two years ago should not receive permanent status simply because the dashboard already exists.
Monetizely's position is ultimately a management one. User activation is not mastered by finding the perfect percentage. It is mastered by identifying who the customer is, defining the earliest behaviour that reliably signals value for that customer, giving every cohort the same clock, and refusing to let an aggregate conceal what is happening underneath.
Assumptions
The cohort arithmetic in the scenario table is modelled to demonstrate the effect of changing segment mix and is not an industry benchmark. Amplitude pricing and packaging reflect public information available through 13 August 2026; negotiated enterprise terms may differ. Historical Amplitude pricing claims rely on dated SEC disclosures and official company materials where individual Wayback captures could not be reliably parsed; the Wayback pricing-page archive is included below for independent historical inspection.
Footnotes
Monetizing Agentic AI: https://www.amazon.com/Monetizing-Agentic-AI-Handbook-Transformation/dp/B0H7Z13VKJ/
Li, H. A., “Converting free users to paid subscribers in the SaaS context,” Production and Operations Management, 2022: https://doi.org/10.1111/poms.13672
Retana, G. F., Forman, C. and Wu, D. J., “Proactive Customer Education, Customer Retention, and Demand for Technology Support,” Manufacturing & Service Operations Management: https://doi.org/10.1287/msom.2015.0547
Steinhoff, L., Kim, J. J., Kanuri, V. K. and Palmatier, R. W., “Unintended consequences of selling B2B digital subscription add-ons for customer onboarding,” 2025: https://doi.org/10.1007/s11747-025-01088-3
Alipourfard, N., Fennell, P. and Lerman, K., “Using Simpson’s Paradox to Discover Interesting Patterns in Behavioral Data,” 2018: https://ojs.aaai.org/index.php/ICWSM/article/view/15017
Slack Technologies, Form S-1/A, 2019: https://www.sec.gov/Archives/edgar/data/1764925/000162828019006616/slacks-1a1.htm
Zoom Video Communications, Form 10-K, fiscal 2021: https://www.sec.gov/Archives/edgar/data/1585521/000158552121000048/zm-20210131.htm
ZoomInfo Technologies, Form S-1/A, 2020, reporting 2019 retention by customer segment: https://www.sec.gov/Archives/edgar/data/1794515/000162828020007502/zoominfo-s1a1.htm
monday.com, Form 20-F, fiscal 2025: https://www.sec.gov/Archives/edgar/data/1845338/000117891326000870/zk2634436.htm
Dropbox, Form 10-K, fiscal 2018: https://www.sec.gov/Archives/edgar/data/1467623/000146762319000005/a12311810-k.htm
Amplitude, Form 10-K, fiscal 2022: https://www.sec.gov/Archives/edgar/data/1866692/000095017023003045/ampl-20221231.htm
Amplitude, Form 10-K, fiscal 2025: https://www.sec.gov/Archives/edgar/data/1866692/000119312526057847/ampl-20251231.htm
Amplitude, Q1 2026 earnings-call transcript, 6 May 2026: https://investors.amplitude.com/static-files/25d7b80b-f022-4b5e-94fe-4f95ff2d2990
Amplitude, Q2 2026 earnings-call transcript, 5 August 2026: https://investors.amplitude.com/static-files/c501e2aa-38c5-4399-b395-0ef5f993325e
Amplitude, official pricing and billing pages, accessed August 2026: https://www.amplitude.com/pricing; https://amplitude.com/docs/faq/billing-and-plans; historical pricing archive: https://web.archive.org/web/*/https://amplitude.com/pricing
Dropbox, Q1 2026 results, 7 May 2026: https://investors.dropbox.com/news-releases/news-release-details/dropbox-announces-first-quarter-2026-results
monday.com, Q4 and FY2025 results, 9 February 2026: https://monday.com/p/press-release/monday-com-announces-fourth-quarter-and-fiscal-year-2025-results/
Atlassian, Q4 and FY2025 results, SEC exhibit filed 7 August 2025: https://www.sec.gov/Archives/edgar/data/1650372/000165037225000028/ex991q4fy25.htm