JavaScript feature flags: Client-side vs server-side evaluation

JavaScript can evaluate flags in the browser or on the server. The useful question is which layer should own each decision.
A client-side flag can change a toolbar after the page loads. A server-side flag can choose the initial HTML or an application behavior before returning a response. Either can use downloaded configuration, and either needs a deliberate identity, fallback and refresh policy.
The trouble starts when those choices are mixed together: browser-visible rules are treated as private, a shared Node instance retains user attributes, or server HTML is immediately replaced by a browser decision using another identifier. The code can be syntactically correct while the application behaves inconsistently.
This guide compares the architectures and builds a small hybrid using the current GrowthBook JavaScript SDK. Node owns a project-board release flag. The browser owns a separate dense-toolbar flag. They receive the same anonymous identifier, but each behavior has one decision owner. The example uses ordinary rollout flags, without an experiment rule or analytics pipeline.
Separate evaluation from configuration delivery
Locate the code that makes the decision
Feature flag evaluation takes a feature definition and relevant attributes, then resolves a value. Configuration delivery supplies the definitions or selected values that process needs. Downloading configuration from a server does not automatically mean evaluation happens there.
| Architecture | Where the decision runs | What reaches the browser |
|---|---|---|
| Browser local evaluation | JavaScript in the page | Definitions and rules for its connection |
| Application-server local evaluation | Node.js in your application | Whatever resolved output you return |
| Browser Remote Evaluation mode | A private evaluation endpoint | Selected values and client-facing metadata |
The first two can both evaluate locally relative to their own runtime. In the browser, “local” means on the device. In Node, it means inside your application process. Neither description promises that acquiring configuration, refreshing it or reporting events requires no network activity.
Browser local evaluation fits public UI configuration and interactions that need values available in the page. Server evaluation fits initial rendering and behavior controlled by application services. The right choice depends on which context has trustworthy attributes and which output should be available before the response is delivered.
Compare the entire path rather than only the final lookup. A browser may need to download a payload before its first decision, then reuse it for local checks. A server may have a warm shared payload, or it may need configuration while preparing a cold response. A private evaluator adds another request boundary. Measure initialization, cache hits, refreshes and the rendered result under your workload; the location of one function call does not establish end-to-end latency.
Neither kind of flag replaces authorization. The server still enforces access to private data and privileged operations. A browser toolbar being hidden is not evidence that its API is protected, and a server release flag is not a substitute for checking the caller's permissions.
Treat browser configuration as observable
A browser downloads its local-evaluation payload. Users can inspect that payload and modify local execution. Scope the browser's SDK connection to an approved public project and inspect its actual delivered payload; do not send an unrestricted server connection to save setup time.
Encoding or encrypting a payload for a browser does not make its eventual readable values secret from that browser. If the page needs a value to render, assume the recipient can observe it. Keep credentials, private business rules and trusted account decisions in the appropriate server boundary.
GrowthBook Remote Evaluation is a distinct option for browser SDKs. The browser sends attributes to a maintained evaluation endpoint and receives resolved values instead of the full local rules. That changes network, infrastructure and cache requirements. It does not make the selected value invisible, and it is not the same integration as using the ordinary Node SDK inside your application.
The runnable example below uses local evaluation on each side. It does not enable remoteEval or configure a private evaluator. If you choose that mode later, use an enabled SDK connection and its supported backend rather than adding the option to the Node client.
Remote evaluation also changes where attributes travel. Ordinary local rule evaluation does not require sending those attributes to a flag evaluator for each lookup, but tracking plugins, application requests or other configured integrations can still transmit data. In remote mode, the evaluator needs the relevant attributes to compute its response. Review that data flow explicitly rather than assuming either architecture keeps every attribute on one machine.
Give the server and browser distinct decisions
Configure two scoped connections
Use Node 24.12 or another version compatible with the dependencies and commands you select. Create a small Express application:
Create two Boolean features with false defaults in your test GrowthBook environment. Put project-board in a server project and dense-toolbar in a separate browser-public project. Create project-scoped SDK connections for that environment and each respective project. This example requires an organization with capacity for both projects; check your organization's project limit. Use ordinary unencrypted local-evaluation connections for this walkthrough.
Different connection keys alone do not establish data isolation. Cross-project targeting and dependencies can affect delivery, so inspect the complete browser payload, including its rules and related data, before approving it. Do not assume that a key name or a feature's apparent project membership proves which definitions the browser receives.
Create this narrow inspection helper. It expects only the public feature used in this example and rejects a reference to the known server-owned flag:
Fetch your public test connection and run the inspection:
Passing this check establishes the expected key set and absence of that known server-key reference, not that every condition or related data field is safe to disclose. Review those fields explicitly. If you later add public prerequisites or other flags, approve their full delivery scope and update the expected set; do not delete dependencies from a downloaded payload just to make a check pass.
Add the connection settings:
The browser settings will be serialized intentionally. The server settings will remain in Node. These are SDK connection settings, not administrative API credentials. Keep the .env file outside anything your web server exposes as a public directory.
Separate the two connections even when they use the same environment. The browser should not receive private rules simply because both runtimes use one JavaScript package. Conversely, do not manually extract a single feature from a larger payload without checking prerequisites, saved groups and other dependencies. The desired unit of public delivery is a reviewed browser connection, not an arbitrary JSON slice that happens to work for today's default value.
Evaluate the server flag in a request scope
The GrowthBook Node SDK provides a shared GrowthBookClient and scoped instances for per-request attributes. Create this module:
The application shares configuration, not a mutable visitor. Each call receives a new user context; no request changes attributes on a shared instance. The strict comparison accepts an actual true value instead of enabling a branch for a truthy string such as "false".
Keep other targeting attributes equally deliberate. If a rule expects a subscription plan or organization ID, load the relevant value from trusted application context and include it in that request's attributes. The anonymous cookie in this tutorial supplies only id. It cannot establish a plan, account relationship or administrative role, and it should not be extended into an unverified bag of privileges.
Calling initialization at this boundary can reuse SDK caching and check for usable configuration. Do not memoize a single initialization promise forever and then assume a long-running process automatically receives edits. Configuration freshness remains a separate operational policy.
Inspect your flag dependencies
Open the graph workflow to inspect prerequisites, dependent flags and linked experiments before changing evaluation ownership. With configured API access, it produces a read-only relationship report for review.
View the Flag Graph WorkflowBootstrap one identity and serialize a small handoff
Use cookie-parser to read the incoming cookie, then validate it rather than trusting its presence. The anonymous UUID is a bucketing identifier for this example, not an authenticated user session.
Create the app below. A first request without a valid ID receives one cookie-setting redirect. If cookies remain unavailable on the marked follow-up, the page uses baseline behavior without evaluating either flag:
The server's selected view is already present in the HTML. The context contains the same opaque ID plus the browser's own connection settings. It contains no server definitions, administrative credentials or SDK instance.
JSON serialization inside a script element needs its own script-content escaping boundary. Replacing < prevents a string value from terminating that element with a closing script tag. The authored UI labels here are fixed text; if you add other dynamic HTML content, escape that separately rather than assuming script-data escaping protects every context.
The cookie lifetime uses milliseconds because this is Express. Use HTTPS and production cookie settings in a deployed application, and integrate the identifier with your consent and retention policy. Cookie blocking must preserve a usable baseline instead of generating a new temporary identity on every request.
An HttpOnly cookie cannot be read through ordinary page JavaScript. Here the server intentionally supplies the opaque identifier in its small context object because browser evaluation needs the same unit. That does not weaken an authentication session: this ID is not the authentication credential in the first place. Keep an actual session secret out of this object and supply only an approved targeting identifier.
Evaluate the browser flag without replacing the server result
Start from visible baseline UI
The server loads the pinned manual browser distribution and then your code through deferred scripts. This is not the automatic snippet plus a second SDK instance. The browser class is growthbook.GrowthBook; the Node class above is GrowthBookClient.
Add the runner:
The initial toolbar stays visible at its standard spacing. If configuration loads and the value is true, the runner changes that spacing. This policy can produce a visible transition; it does not promise a flicker-free initial interface. Choose a loading or server-owned alternative if changing this surface after load would disrupt the task.
The controls are a rendering probe: the sample does not implement project creation, filtering or a full board. Its purpose is to make ownership visible in HTML and DOM attributes. Connect actual application operations only after their own permissions and behavior are defined; the visual flag is not their implementation.
Consider when changing spacing is acceptable. A late layout update can move a control while somebody is reaching for it, even if the final screenshot looks correct. For a critical interaction, you might defer the new value to the next navigation or render the complete surface from one server decision. Test the chosen transition with keyboard and pointer use, not just with a fast development connection.
Update the UI when definitions change
The JavaScript SDK renderer callback connects a payload change to application rendering. Downloading a new definition alone does not update arbitrary DOM. Here the manual refresh button requests configuration, and setRenderer applies the selected toolbar state when the SDK updates.
The runner never reads project-board or replaces server-view. That server decision remains a snapshot until another page request. Refreshing the toolbar is deliberately not a request to change the server-rendered view.
The same ID does not make two different flags one coordinated rollout. Each key may have its own rules, conditions and allocation. If two behaviors must change atomically, model that dependency and decision owner explicitly rather than relying on coincidentally similar percentages.
This separation also determines how to debug mixed states. A board with a standard toolbar is valid when only the server flag is enabled. A list with a dense toolbar is valid when only the browser flag is enabled. If those combinations are invalid in your product, do not copy this independent-flag model unchanged; choose one resolved configuration or define an explicit dependency and verify every supported combination.
This example has fixed anonymous identity for the page lifetime. Authentication, logout or account switching needs a coordinated identity and rerender policy. Do not update one side's ID and assume the other side has followed it.
Verify each boundary and failure policy
Run the application and inspect the handoff
Add a local entry point:
Start with Node's environment-file support:
Inspect the cookie handoff from another terminal, then open the page in your browser:
Expect one 303 cookie-setting redirect, then a personalized response. The marked request without a cookie should render both baselines without another redirect. Inspect the source: the server view should already be selected, and flag-context should contain only the intended browser settings and opaque ID.
Enable each test flag independently and verify its own surface. Turning on the toolbar should not replace the server view. Turning on the server flag should change a new page response without requiring the browser SDK to select that value.
Use this application's failure contract when checking the integration:
| Condition | Server view | Browser toolbar |
|---|---|---|
| Cookie still absent after bootstrap | List baseline | Standard spacing, no evaluation |
| Server configuration unavailable | List baseline | Can still use its independent browser flag |
| Browser script or initialization unavailable | Existing server-selected view | Standard spacing |
| Wrong-type value for one flag | Baseline for that flag | Other decision retains its own policy |
| Browser payload refreshed | Existing server-selected view | Re-evaluated toolbar value |
These are properties of the example's two independent behaviors, not universal rules for a hybrid. Keep configuration errors visible to operators even when the user-facing default looks normal. Distinguish unavailable identity, failed initialization, missing definitions and intended false values in your monitoring rather than counting every baseline as an error.
Verify all four supported combinations: list with standard toolbar, list with dense toolbar, board with standard toolbar, and board with dense toolbar. Then disable browser JavaScript and confirm the server-selected view remains readable. These checks catch an accidental browser rewrite of the server decision and an accidental dependency between flags that were intended to be independent.
Repeat the check with one durable ID before introducing more audiences. Changing both identity and configuration at once makes a mismatch hard to explain. When you add targeting, record the exact test attributes, applicable rule and expected value at each owner. A result from a different cookie jar or cleared browser profile is a different input, not automatically a failed repeat.
Keep HTTP, SDK, and DOM caches separate
Personalized HTML caching is different from caching reusable feature definitions. The example marks its HTML private and no-store. Confirm upstream infrastructure honors that policy rather than caching one visitor's response for everyone.
SDK initialization can succeed with cached configuration. Browser refresh can complete without proving the latest dashboard edit reached the interface. Server processes can also hold different cached snapshots. Define freshness and recovery expectations for the actual application, and test an existing page as well as a newly loaded one.
A refresh failure can resolve while retaining the last usable toolbar value; it does not necessarily throw into the example's catch block. Keeping that value may be preferable to an abrupt UI change, but the enabled refresh button is not a freshness receipt. Add monitoring appropriate to your chosen recovery policy.
The no-store response policy does not purge an older response already held by an incorrectly configured upstream cache. When introducing personalization to an existing URL, review its prior caching behavior and invalidate old shared responses where necessary. Reusable SDK definitions and per-visitor HTML have different sharing rules; changing one cache's setting cannot repair the other automatically.
The manual browser refresh is not a background update subscription. For a different polling or streaming policy, use the appropriate documented SDK options and lifecycle. Do not assume a timer keeps a serverless process alive or that setting an initialization timeout guarantees every part of a request's latency.
Test isolation and malformed values
Test missing, malformed and JSON-decoded object cookies. Check absent connection keys, failed initialization, missing flags, nulls and strings in place of Booleans. Confirm the reviewed baseline remains usable, and verify diagnostics separately from normal false decisions.
Send concurrent server requests with different test IDs and rules. The shared client must reuse definitions while each scoped evaluation retains its own attributes. For the browser, test unavailable SDK scripts, malformed context JSON, failed loading and a payload update that changes the toolbar. Test actual layout and interactions in a browser; an HTTP or simulated-DOM fixture cannot establish what a user saw.
Inspect the actual browser payload in the network panel as part of that test. Confirm that the server connection key and server-only definitions are absent, while the expected public flag and any required dependencies are present. Inspect flag-context separately: a small, safely encoded handoff is different from a safe feature connection. Both boundaries need review, and neither is established solely by seeing the correct toolbar spacing.
If you later add experiments, assignment callbacks require a designed exposure boundary and outcome pipeline. A server response can be prefetched or abandoned, and a DOM change is not automatically a viewport impression. These examples intentionally omit experiment tracking rather than relabel diagnostics as analysis data.
Choose and maintain the evaluation architecture
Use the context that owns the behavior
Choose browser local evaluation for public UI behavior that the page can control with the attributes it legitimately has. Review what the connection exposes, establish a loading policy and wire refresh into rendering.
Choose application-server evaluation when the service owns the behavior or the initial response should include the decision. Keep attributes request-local, preserve a practical baseline and keep authorization independent. Return only the output the browser needs.
For example, an existing application can use a browser flag for a public panel without moving its data-access policy into the page. A service selecting a backend processing path should evaluate near that path and retain a compatible failure behavior. A page that only needs the server's chosen label can receive the label directly; it does not need another SDK merely because the page also contains JavaScript.
Choose browser Remote Evaluation when hiding full rules and unused values justifies its maintained evaluator and network boundary. Selected values remain visible, and the attributes sent to the endpoint need a deliberate privacy policy. It is an architecture decision rather than a universal performance improvement.
For a browser-heavy SPA, also define SDK ownership across navigation. The sample uses a full document lifetime. A component-based application may keep one instance above the routes and update its attributes deliberately, or create a scoped integration with matching cleanup. Repeatedly loading this runner for the same persistent document would add instances and listeners; adapt the lifecycle instead of treating page-script insertion as a refresh strategy.
Migrate ownership without running two authorities
When moving a flag between server and browser, specify which layer controls the first render, where identity comes from and when ownership changes. A hybrid may pass resolved values or deliberately initialize from a browser-approved payload, but it must coordinate attributes and definitions. Do not hydrate a browser by copying an arbitrary slice of the private server payload.
If the same decision must continue on the client after server rendering, start from the server-selected state and define when re-evaluation is allowed. An immediate second evaluation with another ID, stale payload or different attributes can replace the initial state during hydration. Passing a resolved value is often enough when no browser re-evaluation is needed; shipping a complete second engine is a larger choice with its own update policy.
Keep targeting rules, connection scope and fallback values in the same review as code changes. When a decision is durable, use code references to locate consumers and review removal of the unused branch. Check old clients before deleting configuration they still expect.
Record the key's owner, purpose, valid values and expected lifetime alongside its consumers. Review operational diagnostics without logging whole feature payloads or private attribute objects. A useful rollout readout distinguishes loading failures from normal decisions and measures the actual behavior it controls. Neither a successful payload request nor a high count of true lookups proves the release improved a business outcome.
The useful result is a clear owner for each behavior, not the maximum number of SDK instances. Start with one server decision and one public browser decision, verify their boundaries independently, and expand only where the next surface needs the additional evaluation context.
Build a clear flag boundary
Start with GrowthBook and one behavior whose identity, fallback and evaluation owner are explicit. Add browser or server evaluation where that application surface needs it.
Get StartedRelated Articles
T-statistic explained: What the t-value tells you about your results
An experiment report shows a t-value of 2.4. Is the product change large? Is it beneficial? Does it have a 95% chance of working? That single number cannot answer those questions.
A t-statistic compares an estimated quantity with a hypothesized value, using the estimate’s standard error as the measuring stick. It compresses the size of the observed departure and the precision of the estimate into one signed number. Understanding the two parts is more useful than memorizing a cutoff.
For experiment readers, the essential habit is to unpack the t-value before interpreting it. Establish what was subtracted, which uncertainty estimate sits underneath it, and what reference distribution applies. Then return to the effect in product units and the decision it needs to support.
Read t as a difference divided by uncertainty
Identify the numerator and its direction
The general structure is:
In an independent treatment–control comparison, the estimated quantity might be the treatment mean minus the control mean. With a null difference of zero, a positive t means the estimated treatment mean is higher. Reverse the subtraction order and the sign reverses.
“Higher” does not necessarily mean “better.” More purchases may be desirable; more errors usually are not. A negative latency difference can support an improvement, while a positive error-rate difference can indicate harm. Read the direction in the metric’s own terms before assigning a business meaning to the sign.
The hypothesized value also matters. The usual zero-difference comparison asks whether the population mean difference is zero. Testing another specified difference changes the numerator. Neither calculation independently establishes an equivalence or non-inferiority claim; those require the appropriate hypotheses, margins and decision procedure.
Distinguish standard error from standard deviation
An observation standard deviation describes spread among observed unit outcomes. A standard error describes uncertainty in an estimator, such as a sample mean or a difference between means. They can share units while answering different questions.
For an ordinary mean of independent, identically distributed observations with finite variance, the estimated standard error is the sample standard deviation divided by the square root of the sample size. The standard-error calculation therefore changes when the number of independent observations changes, even if their measured spread stays the same.
This is why dividing a mean difference by the observation standard deviation does not give the usual two-sample t-statistic. It omits how precisely the means were estimated. The distinction between standard error and standard deviation is the first denominator check to make when a t-value looks surprising.
Unit-free does not mean effect size
Both numerator and standard error have the same measurement units, which cancel in their ratio. Converting a revenue analysis from dollars to cents multiplies both by 100 and leaves t unchanged. That is a useful consistency check for a positive linear change of units.
However, t does not describe the size of the product effect independently of precision. An observed difference of one dollar can have a small or large t-value depending on the standard error. Comparing t-values across two experiments is therefore not a reliable way to rank their business impact.
The statistic measures departure from the null in estimated standard-error units. Keep the raw difference beside it so the unit-free summary does not displace the quantity the team actually needs.
The calculation also needs a usable, positive estimated standard error. A zero or undefined denominator is a reason to inspect the data and model, rather than manufacture a finite t-value or call the result certain. Software may reject such inputs; for example, the R t-test reference documents an error for effectively constant input data.
Match the denominator to the study design
Different designs estimate different uncertainty
The formula’s short form hides a substantial choice: how was the standard error calculated? Three common cases illustrate why the design must come first.
| Design | Estimated standard error | Reference degrees of freedom |
|---|---|---|
| Ordinary one-sample mean | Sample SD divided by √n | n − 1 under the classical model |
| Ordinary paired mean difference | SD of within-pair differences divided by √number of pairs | Number of independent complete pairs − 1 under the classical model |
| Independent Welch comparison | √(s² treatment / n treatment + s² control / n control) | Welch–Satterthwaite approximation using both group sizes and variances |
In the paired case, the relevant spread is the spread of differences within pairs. Substituting the two arms’ separate standard deviations discards the relationship the design uses. The paired t calculation treats the differences as its observations; the independent information comes from the pairs.
For two independent groups, Welch’s calculation does not require equal population variances. The pooled Student calculation adds that assumption and uses a different variance estimate. The Welch-versus-Student comparison helps distinguish those cases. A label reading only “t-test” can conceal the choice.
Separate a reference model from an arithmetic ratio
You can calculate a difference divided by an estimated standard error without proving that the resulting statistic has the reference distribution you intend to use. Exact classical one-sample and paired t inference relies on the corresponding independent normal-observation model; in the paired case, this concerns the differences.
Welch’s reference distribution is an approximation with estimated degrees of freedom. It is not generally exact even for unequal-variance normal populations. Non-normal outcomes can support useful large-sample mean inference under appropriate conditions, but there is no sample-size threshold that makes every heavy tail or dependence problem disappear.
If accounts are randomized, individual events from one account are not automatically independent units. A standard error that ignores the account structure may understate uncertainty. More rows in an exported file do not by themselves justify a smaller denominator or larger degrees of freedom.
Plan for more precise estimates
Explore variance-reduction methods and their tradeoffs before your next test. Match each method to the outcome you intend to estimate.
Explore Variance ReductionHold the observed effect fixed and change precision
Work through an independent mean comparison
Consider a hypothetical fixed-horizon experiment with one completed revenue outcome per independent user. Control has mean revenue of $10; treatment has $11. Each arm has sample standard deviation $4 and 100 users. The target is the arithmetic mean difference, treatment minus control, and the null difference is zero.
The estimated effect is $1 per user. Using the independent unequal-variance standard error, square each group’s standard deviation, divide by its sample size, add those terms and take the square root. Here that is √(16/100 + 16/100), or about $0.5657.
Divide the $1 difference by $0.5657 to get t ≈ 1.7678. The estimated difference lies about 1.77 estimated standard errors above the zero-difference null. The sample standard deviation is still $4; the $0.5657 figure measures uncertainty in the difference between the two estimated means.
For these equal sample sizes and equal observed variances, the Welch–Satterthwaite calculation gives 198 degrees of freedom. Under the conventional Welch approximation, the two-sided p-value is approximately 0.07864 and the 95% confidence interval for the mean difference is approximately −$0.1155 to $2.1155.
Compare two hypothetical studies
Now consider a separate hypothetical study with the same group means and standard deviations, but 400 independent users per arm. The observed $1 difference has not grown. Its estimated standard error is smaller.
| Summary | 100 users per arm | 400 users per arm |
|---|---|---|
| Observed mean difference | $1 | $1 |
| Sample SD in each arm | $4 | $4 |
| Estimated SE of the difference | $0.5657 | $0.2828 |
| t-statistic | 1.7678 | 3.5355 |
| Welch reference df | 198 | 798 |
| Conventional two-sided p-value | 0.07864 | 0.0004304 |
| Conventional 95% CI for the difference | [−$0.1155, $2.1155] | [$0.4448, $1.5552] |
Quadrupling each group’s sample size halves the estimated standard error when these observed standard deviations stay fixed. It doubles t for the same observed effect. The larger statistic reflects greater estimated precision here, not a larger observed revenue gain.
These are teaching summaries for separate studies. They are not an instruction to keep extending one experiment until significance appears. In a real extension, means and variances can change, and the stopping rule affects how the inference should be calibrated. An observed t-value also does not directly tell you the power of a future study.
Translate t using the appropriate reference distribution
Degrees of freedom and tails determine the threshold
A t-value needs a reference distribution before you can obtain a p-value or critical value. The Student t family includes a degrees-of-freedom parameter; smaller values generally give heavier tails. The same absolute statistic can therefore produce different tail probabilities at different df.
For a prespecified two-sided 5% test, the relevant critical value is the 97.5th percentile of the reference t distribution. It approaches about 1.96 as df becomes large, but 1.96 is not a universal t cutoff. Nor does every test use n − 1. Degrees of freedom depend on the calculation and may be fractional for Welch.
A one-sided alternative uses a specified tail and a different critical value. A negative statistic is not evidence in favor of an upper-tail alternative simply because its absolute value is large. Choose the direction before seeing outcomes; taking whichever tail looks favorable changes the decision rule.
A p-value conditions on the null model
For a conventional two-sided t-test, the p-value is the probability, under the null model and assumptions, of a statistic at least as extreme in either direction as the observed t. It is not the probability that the null is true, that chance caused the result, or that a future experiment will succeed.
The ASA’s p-value principles also distinguish statistical significance from effect magnitude and practical importance. A threshold crossing does not determine whether the expected benefit pays for the change. If you need the broader conditional-probability interpretation, the t-test p-value guide develops it separately.
Read the interval in the original units
The corresponding conventional interval is the estimated difference plus or minus the relevant critical t-value times its estimated standard error. For the same two-sided procedure at 5%, a 95% interval excluding zero agrees with rejecting a zero-difference null. That agreement requires compatible test, interval and adjustment settings.
Confidence describes the repeated-use coverage of the method under its assumptions. It does not assign a 95% posterior probability to the fixed true effect lying inside the particular computed interval. The confidence-interval construction makes the relationship between standard error, critical value and width explicit.
In the first study, the interval includes a small loss and a substantial gain. In the second, it excludes zero but still spans different levels of business value. A minimum worthwhile effect helps interpret that range. Neither a large p-value nor an interval containing zero proves equality; equivalence requires an analysis designed for that claim.
Audit the inputs before acting on a t-value
Trace an unexpected statistic to its components
Start with the estimated effect and null value. Confirm the subtraction order, currency or time units, eligible population and observation window. A swapped group order explains a sign change; it does not explain a change in a correctly calculated two-sided p-value.
Then inspect the denominator. Check group sizes and sample variances at the actual independent unit. Look for duplicated joins, repeated events counted as independent observations, incomplete follow-up and treatment-dependent missingness. The arithmetic can be internally consistent while the data fail to represent the intended experiment.
Inspect the analysis settings too. A ratio metric, regression adjustment or variance-reduction method can have a standard error different from the simple independent-mean formula above. Do not attempt to reproduce a dashboard by substituting top-line traffic counts into the wrong calculation. GrowthBook’s statistical effect calculations describe its estimators and frequentist reference methods, including distinct metric types.
Keep the decision rule attached to the evidence
The worked examples use conventional fixed-horizon inference. Repeatedly inspecting that calculation and stopping at a favorable threshold changes its false-positive behavior. A planned sequential analysis is a different inferential procedure; GrowthBook documents an asymptotic confidence-sequence approach with its own settings. It does not repair incorrect units or missing outcomes.
Likewise, selecting the largest t among many metrics, variants or segments creates a testing problem beyond a single comparison. Preserve the planned comparisons and applicable corrections rather than treating the most impressive number as an isolated result.
A useful readout names the estimated effect, interval, test, df, direction and sample sizes, followed by the measurement checks and decision threshold. The team can then assess what the evidence supports and whether that effect is worth shipping. Use the t-value to understand precision relative to the null; keep the launch decision grounded in effects on users and the business.
Review the result in context
Open the official experiment-analysis workflow to inspect saved results, health checks and analysis settings with configured API access. It can refresh a result snapshot when needed; stopping or rollout is separate.
View the Analysis WorkflowA published Strapi document is content. An A/B test also needs an assignment, a delivered experience, and an outcome you can attribute.
The integration becomes unreliable when those pieces disagree. An older example may read data.attributes, while your Strapi 5 endpoint returns flat fields. A numeric database ID may be mistaken for a document ID. An internal service call may return a draft even though a similar REST request defaults to published content.
This guide builds an explicit Strapi 5 content test using a small server-rendered Node application and GrowthBook. It validates the complete published content pair before enrolling a visitor, preserves a stable browser identifier, and records an exposure only after browser execution reaches the selected page.
The code also fingerprints the reviewed strings so a later edit can be rejected before assignment. The local event collector demonstrates the measurement boundary; a real experiment still needs your existing signup flow, durable analytics delivery and verified warehouse joins.
Define the Strapi 5 delivery contract
Pick a content model you can inspect
The existing GrowthBook Strapi integration guide demonstrates Dynamic Zone components selected through a frontend SDK. The reference here uses two ordinary string fields and server-side evaluation. It does not require a separately installed Strapi experimentation plugin or depend on an unverified plugin compatibility claim.
Start with a hypothesis about a meaningful outcome: an outcome-focused headline will increase completed trial signups per eligible offer-page visitor. Keep the offer and signup destination consistent if headline copy is the intervention. If the change includes a larger component package, describe that package as the treatment.
A plain-text pair makes the first integration easier to inspect. You can verify the exact response, render the selected value safely, and compare the approved strings. When extending the model to components or Dynamic Zones, validate their full structure and use a reviewed variation mapping. Reordering an array must not silently redefine which experience belongs to control.
For example, an editorial interface might place an alternative card first because it is easier to find, while the frontend still interprets index zero as control. The content can remain valid JSON and render perfectly while the experiment labels have effectively switched. Use explicit variation keys or a constrained, reviewed ordering contract when extending the simple pair. Check the rendered package against the experiment mapping, not just that every component has a title.
Use the document API that your code actually calls
Strapi 5 REST responses expose normal content fields directly under data. The document endpoint uses documentId; the numeric id in a response is not the identifier this guide puts in the URL. Some plugin endpoints have different response conventions, so this contract applies to an ordinary generated collection endpoint.
The production request explicitly sets REST status to published. This is also the REST default when Draft & Publish is enabled. The internal Document Service API defaults differently, to draft. Do not transfer that internal default—or an older v4 parameter—into this REST implementation.
Keep locale explicit as well. The requested document and locale together identify the content version being delivered. This guide checks the returned locale instead of borrowing another CMS's fallback behavior or assuming a valid document ID means every translation exists.
The document concept groups localized and draft/published variations around a document identifier. That does not make the document identifier an immutable snapshot of the text. Keep source identity, publication state and approved content as separate fields in the experiment record. Changing one is a deployment decision even when the others remain familiar.
Define assignment and measurement before launch
Use one stable assignment unit. The example creates an anonymous first-party browser ID; account-level interactions may require account assignment instead. Plan the anonymous-to-authenticated transition so signup measurement does not lose the identifier or enroll one person under several unrelated IDs.
Choose a primary business outcome, a small diagnostic funnel and relevant guardrails. A headline can increase button clicks without increasing completed or qualified signups. Estimate eligible traffic and the effect that would change your decision before starting a test that cannot resolve the question in a useful period.
Separate content ownership from experiment operation. An editor approves the published pair, an engineer owns delivery and fallback behavior, and the analytics owner verifies exposure and outcome data. Name who can change the tested content and how a change starts a new reviewed comparison.
Model, publish and inspect the content pair
Create a localized collection in development
Use an existing Strapi 5 test application with Draft & Publish enabled for the collection. Configure internationalization and the intended locale if the content is localized. The Content-type Builder belongs to the development workflow; deploy schema changes through your normal application process.
Create a collection named Experiment page with singular API ID experiment-page and plural API ID experiment-pages. A minimal schema at src/api/experiment-page/content-types/experiment-page/schema.json is:
For an existing model, merge the relevant changes rather than replacing its other fields. Restart or rebuild according to your development workflow, create a document, enter both approved headlines and publish the intended locale. Copy its actual documentId from the API response or your document inspection workflow.
Review model changes and content changes through their appropriate paths. Adding a field affects application schema and deployment; editing a headline changes a content entry. Give the editor clear field descriptions so control means the approved baseline and variant means the approved experimental alternative. A required field helps authoring, but the renderer still validates the actual response from the deployed endpoint.
With internationalized content, inspect the specific localized pair that will be eligible for the experiment. Translating control but leaving treatment in an unintended language changes more than the hypothesis describes. Keep locale targeting and content approval aligned, and decide in advance whether results across languages belong to one planned analysis or separate comparisons.
Because both alternatives are plain strings, this request needs no relation population. Components, relations and media require an explicit population strategy when added later; populate=* is not a substitute for defining the treatment's dependencies and permissions.
Use a read token on the rendering server
Create a scoped Content API token for the renderer. A read-only token supports find and findOne; a Custom token can narrow the relevant endpoint access. Keep it on the application server. You do not need to expose an admin token or broadly enable public-role access for this authenticated server request.
Read-only describes allowed operations, not a universal guarantee that a caller can never request drafts. Fix production status and locale in trusted code. Keep editor preview routes and any draft-capable configuration separate from live enrollment and analytics.
When extending the endpoint with a language switcher, select from an approved locale set rather than forwarding every browser query parameter to Strapi. A request parameter named status should not turn the production renderer into a draft-content proxy. Keep the collection path and publication choice fixed, and validate the returned document and locale even after constructing a valid request.
Check permissions on the actual route you use. A working admin interface does not prove the rendering token can read the generated Content API endpoint. Conversely, an access problem is not a reason to grant broad public access to the entire model. Confirm the token's scope and the endpoint response independently of experiment assignment.
Create a reference Node project, or integrate the files into your existing application. Use a modern Node runtime compatible with the locked dependencies, such as Node 22 or later:
Add server configuration. STRAPI_ORIGIN is the CMS origin, including a port when needed, without /admin or /api appended:
Use the appropriate HTTPS origin outside your local development setup. Store credentials through your deployment's configuration mechanism, not in public HTML or query parameters.
Validate the published response and reviewed fingerprint
The following reader fetches one document by documentId, with explicit locale selection. It rejects a missing, draft, wrong-locale or malformed pair before the SDK can assign anyone:
Leave the optional expected hash blank for the first read-only inspection, with Strapi running and the document published. Run:
Confirm the actual approved pair, source origin, document and locale. Set STRAPI_EXPECTED_SHA256 to the returned digest before enrollment. The SHA-256 calculation fingerprints those exact strings and source dimensions. It does not retrieve an older Strapi version or grant access to anything.
documentId, updatedAt and publishedAt are useful context, but they do not freeze a treatment. The fingerprint rejects changes to the reviewed pair before assignment. Identical strings can retain the same digest after republication; publication metadata is recorded separately. Extend the fingerprint when additional content becomes part of the treatment.
Treat updating the expected digest as a review step. First inspect the new delivered pair, confirm why it changed, and decide whether the current experiment can continue under a documented phase or needs a new experiment. Copying the latest digest merely to clear an unavailable response defeats the check. Preserve the approved strings alongside the digest so a reviewer can understand the change without reconstructing it from a hash.
The digest covers the two headline strings and source dimensions in this implementation. It does not cover an unrelated image, price, signup destination or application deployment. If those become part of the intervention, extend the reviewed package and its validation. A useful content contract is explicit about both its coverage and the changes that require another review.
Review the experiment before launch
Open the experiment-design workflow to turn your content hypothesis, assignment unit, and metrics into a reviewable specification. Configure API access as directed and review the plan before launch.
Open Experiment Design InstructionsRender the assigned page with request-local context
Link the content values to an experiment
Create a string feature strapi-offer-copy, defaulting to control. Link an experiment with tracking key strapi-offer-copy-v1, mapping the two string values to the intended variations. Set hash attribute id, the correct SDK environment, targeting and goals in the experiment configuration.
The content value chooses a string from the pair. The experiment and variation identifiers describe the assignment for analysis. Preserve experiment.key and result.key; do not substitute the headline or assume a variation key must equal the content value.
The GrowthBook Node SDK can share feature payload data while keeping each visitor's context in a scoped instance. The server initializes through the request path and checks success, using the SDK's caching and refresh policy rather than retaining one startup promise forever. Test stopping and refreshing on warm instances; a shared cache is not an instant global update guarantee.
Keep the browser identity stable and the output safe
Create the server below. The existing /signup application flow is a prerequisite: implement or replace that link before measuring completed signups. The collector is a local diagnostic endpoint until connected to durable analytics.
The cookie parser can produce non-string values, so identity is validated before assignment. Express cookie maxAge uses milliseconds. The redirect establishes the cookie before the offer request; its one-attempt marker avoids looping when persistence fails. Adapt identity and consent behavior to the application's existing policy.
The headline is escaped as HTML text. The JSON data element separately escapes less-than signs so a closing-script sequence cannot terminate it. These are different output contexts. Do not use this plain-string renderer as a general HTML or Rich Text renderer.
Keep failure behavior visible
A missing document, wrong locale, draft payload or changed fingerprint fails before enrollment. A failed feature-payload initialization can retain valid control content outside enrollment. Unexpected raw assignment values are logged and do not become misleading control-view events.
Fetching both alternatives in one validated payload makes this eligibility check independent of the assigned arm. A different architecture that first assigns treatment and then fetches only its content has another failure problem: a missing treatment resource can selectively remove treatment visitors from observed exposures. If larger packages require that architecture, preserve assignment and delivery failures explicitly and design the analysis population before launch. Do not infer that a missing view means the visitor never belonged to the comparison.
Preserve diagnostics and investigate the affected interval. Filtering failed treatment deliveries from analysis can select the comparison population after assignment. A safe display fallback does not make a broken experiment trustworthy; repair or halt it when delivery or configuration is unreliable.
The personalized response uses private, no-store. Shared caching of the complete validated pair is a separate choice from caching a visitor's selected HTML. Verify the actual reverse proxy and CDN behavior rather than assuming a response header alone describes every cache in the deployment.
When adding a shared content cache, include the CMS instance, document, locale, publication selection and reviewed fingerprint in the appropriate cache design. A cached English pair should not satisfy a request for a different locale, and a previously approved pair should not conceal a broken publication process indefinitely. Test invalidation and recovery using the same deployment configuration as the experiment.
The server selects one headline before producing HTML, so the browser does not need to rewrite that headline to implement this treatment. Measure the server response path under realistic load nonetheless: content fetching, payload initialization and application work still consume time. Keep loading and error guardrails appropriate to the page rather than promising that server-side evaluation has no performance cost.
Connect browser exposure to the outcome
Record a defined display boundary
The callback during evaluation queues assignments; it is not proof that HTML reached a browser. Use the deferred-tracking pattern to carry compact metadata into a client script that runs after parsing the page:
This boundary establishes that browser execution reached the rendered page. It does not prove viewport visibility or reading. Use a symmetric visibility condition if that is the intended exposure definition. A successful beacon queue operation also does not prove network acceptance or durable ingestion.
The data-element marker prevents duplicate execution within one document. Reloads can produce additional displays, and collector retries need their own event-ID policy. Define whether analysis uses first exposure or another consistent rule, then make the assignment query match that definition.
Replace local logging with verified ingestion
The diagnostic collector reads identity from the HttpOnly cookie, validates the event/source fields and discards unexpected fields. It still receives browser-supplied observations. Do not use that metadata as an authorization or billing decision, and do not treat a console row as a warehouse record.
Connect your supported analytics ingestion with appropriate validation, origin controls, delivery monitoring, retries and deduplication. Apply consent and tracker-readiness behavior consistently across both arms. Compare assignments, browser events, accepted requests and warehouse rows when investigating loss; those counts represent different boundaries.
Configure the GrowthBook data source to read experiment key, variation ID, timestamp and assignment unit from the actual stored schema. Preserve the document/locale/hash provenance in the event or a linked experiment record so the content behind a result remains identifiable.
Validate that mapping with one controlled event before increasing traffic. Compare the browser request, accepted collector fields and resulting warehouse row. Field names can change through an analytics transformation, timestamps may be stored in different zones, and a custom identifier can be dropped by a tag or export. Finding the row and joining its outcome is stronger evidence than seeing a successful HTTP response from the local collector.
Join the business outcome to the same identifier
Define completed signup or another chosen outcome with Fact Tables and metrics. Verify that it joins to the browser assignment identifier or a tested authenticated-identity mapping. A signup on another origin does not automatically carry this cookie.
Inspect an actual controlled exposure and outcome together, including timestamps, duplicate handling and follow-up window. Recent visitors may not yet have had enough time to complete the outcome. Keep diagnostic clicks and qualified conversions separate so faster but less useful signals do not replace the intended decision.
Verify publication and maintain a clean test
Check HTTP behavior before interpreting results
With the reviewed hash configured, run the application and establish a local cookie jar:
Expect a fixed redirect and consistent selected content for the same identifier while configuration remains stable. These requests inspect server HTML; they do not execute the browser exposure script. Use a real browser to inspect the beacon and collector, then verify the actual analytics and warehouse path separately.
Exercise both arms with controlled identities. A forced value can prove a rendering branch, but it does not prove ordinary randomized enrollment. Keep preview, forced and employee QA traffic outside production analysis.
Persistence is also narrower than sticky assignment across arbitrary configuration changes. The same browser ID can be used consistently while a changed experiment key, rule or allocation alters the experience. Freeze the intended configuration during the comparison, record deliberate changes, and use the platform's supported bucketing behavior when your requirements extend beyond a stable identifier.
Test the modern contract and its failure cases
| Test condition | Expected behavior | What to investigate |
|---|---|---|
| v4-style nested response | Reject missing flat fields | Migration compatibility and parser version |
| Numeric row ID used as document ID | No successful content display | Actual documentId and generated endpoint |
| Draft or wrong-locale payload | Reject before assignment | Status, localization and preview configuration |
| Reviewed string changes | Hash mismatch before enrollment | Publication approval and treatment version |
| SDK, cookie or event failure | Explicit baseline/unavailable/diagnostic path | Identity continuity and delivery loss |
Also test HTML-significant characters and a closing-script sequence in approved QA content. They should remain text, and the data element should remain intact. Repeat cache and identity checks on the actual deployment: local success does not establish proxy behavior, public rendering or warehouse attribution.
Include a schema deployment in the regression check when field names or content structure change. A correct new schema in development can coexist with older published documents or a renderer expecting the previous shape. Verify the deployed model, the delivered payload and the rendering contract together. Keep a clear unavailable/diagnostic result for incompatible content rather than disguising it as a valid control observation.
Close the loop with an owned content decision
Before interpreting lift, investigate allocation anomalies, missing data and variation-specific failures. Agree on duration, outcome maturity, meaningful effect and a decision framework before seeing the result. An imprecise estimate can leave the content choice unresolved, and nonsignificance does not prove equivalence.
When the test ends, stop enrollment, verify payload refresh in deployed instances and publish the chosen headline as ordinary content. Preserve the approved pair, source document/locale, fingerprint and metric definition with the learning. Remove retired assignment and tracking branches through review so future tests begin with a clear owner and a clean content contract.
Coordinate the content promotion with the fingerprint gate. Turning the winning headline into the new control changes the reviewed pair, so approve that resulting baseline and its new digest deliberately. Test the publication/configuration sequence in staging; do not clear a mismatch by copying a new hash without checking what it represents. The final baseline should work independently of the retired experiment rule and its exposure code.
Run a clear content comparison
Connect published Strapi content to consistent assignments and the business outcomes your team already measures. Start with a small test and verify the complete delivery and measurement path.
Get Started With GrowthBookYour randomizer can work correctly while the final experiment comparison becomes unreliable. People can enter different reporting paths, lose their identifiers, or disappear from outcome collection after assignment. The result still looks like an A/B test, but the rows left in the analysis may no longer support the question you meant to answer.
Random assignment is a powerful protection against selection into treatment. It does not guarantee representative recruitment, complete follow-up, or valid post-treatment filters. Understanding that boundary helps you distinguish an allocation problem from a problem observing the people you allocated.
The most consequential question is often what randomization lets you learn when some outcomes remain unknown. A larger sample or a more sophisticated statistical engine cannot supply those missing facts without additional evidence or assumptions.
Understand the protection randomization provides
Assignment should not depend on potential outcomes
Selection bias concerns systematic distortions introduced by how observations enter or remain in the comparison relevant to your question. The selection-bias definition emphasizes the relationship between the studied group and the population of interest. Terminology varies across fields, so name the actual selection process rather than assuming the label explains it.
With a common assignment probability, proper random assignment makes treatment allocation independent of the units’ potential outcomes. When assignment probabilities vary across prespecified strata, that independence is conditional on those strata. Product teams do not have to identify every motivation, prior preference, or unmeasured characteristic before assigning users. The allocation mechanism avoids deliberately routing more promising users into treatment and less promising users into control.
This protection concerns the units eligible for that randomization. It also relies on the implemented experiment matching the design: the assigned treatment must be well defined, identity must support the assignment, outcomes must be measured appropriately, and interference between units must be addressed where relevant.
Balance is an expectation, not a perfect copy
Randomization does not create two identical lists of people. A finite experiment can have chance differences in baseline behavior even when the mechanism is sound. The methodological basis of randomized comparisons is comparability on average under the design, rather than exact equality of every measured or unmeasured characteristic in every sample.
Investigate evidence of implementation failure, but do not repeatedly rerun allocation until the observed outcome looks attractive. Prespecified blocking or stratification can be legitimate design choices; changing allocation in response to favorable results is a different procedure.
Randomization also does not mean the observed difference equals the true effect exactly. Sampling or randomization uncertainty remains even with complete measurement. A confidence interval addresses uncertainty under its assumptions; it does not certify that recruitment, joins, filters, or outcome observation were valid.
There is another limit to what is observed. Each unit receives one assigned experience, so you do not observe that same unit’s outcome under the alternative at the same time. Randomization supports an aggregate comparison under the design; it does not reveal an individual causal effect for every user. Keeping this distinction clear prevents a clean experiment average from becoming an unsupported claim that every customer benefits.
Name the treatment effect you want to estimate. The effect of offering an experience and the effect among people who actually use it are different questions. Random allocation directly supports the first comparison under its assumptions; a user’s decision to adopt is not another random allocation.
Define the population and unit the design covers
Random assignment is different from random sampling
Assignment decides which experience an enrolled unit receives. Sampling concerns which units enter the study from a population. Randomizing a volunteer beta group can support a comparison inside that beta, while leaving a separate question about whether its result applies to customers who did not volunteer.
A narrow study is not automatically a failed study. An Android-only feature may appropriately be tested among eligible Android users. A launch claim about all platforms needs additional support. State the population, setting, and time period in the decision before debating whether the sample is representative enough.
The broader selection-versus-sampling comparison explains these categories and post-treatment subgroup examples. For an experiment review, the immediate task is to identify which population your assignment actually covers and whether the final analysis preserves it.
Choose identity and eligibility before treatment changes them
Eligibility should describe who can meaningfully receive the tested experience, using information that the treatment cannot already have changed. A supported device or an account state measured at entry can define a relevant population. Successfully using the new experience is usually an outcome, not an equivalent entry criterion.
Choose the randomization unit to match the intervention. Assigning individual employees to different collaboration settings may create interference when they work together. Assigning accounts can address some shared-experience requirements, but changes the number of independent units and the analysis needed. Randomization-unit design guidance makes identity stability and network effects part of the experiment design, not merely SDK configuration.
In implementation, preserve the assigned unit and variation through the outcome join. A device identifier that disappears after login can create missing outcomes rather than genuine non-conversions. A changing identity can also expose one person to multiple variations. Record the identity and timing assumptions with the experiment configuration, then verify the delivered data follows them.
Define the denominator first
Use the KPI playbook to distinguish binary, continuous, and ratio metrics, and connect each measurement to the experiment decision it is meant to support.
Read the KPI PlaybookMissing outcomes can defeat an apparently favorable result
Complete-case rates describe the people still observed
Suppose a hypothetical experiment assigns 100 units to each arm. The binary outcome window has ended for everyone, but a collection problem leaves some outcomes unknown. There are 18 known successes in each arm:
| Outcome availability | Control | Treatment |
|---|---|---|
| Assigned units | 100 | 100 |
| Units with observed binary outcomes | 90 | 60 |
| Known successes among observed outcomes | 18 | 18 |
| Units with unknown outcomes | 10 | 40 |
| Success rate among observed outcomes | 20% | 30% |
The complete-case comparison is 18/90 versus 18/60, an observed difference of +10 percentage points. It compares the units whose outcomes were available. It does not establish the success-rate difference for all 100 assigned units in each arm.
The larger missing fraction in treatment is a warning about observation. It is not proof of a particular direction or magnitude of bias. Equal missing fractions would not settle the issue either: equally many missing users could have different outcomes or different reasons for being unobserved.
Missingness does not inevitably create bias in every analysis. Purely random loss can mainly reduce precision; other missingness processes can select people with systematically different outcomes. The missing-value count alone does not identify which process occurred. Record the reasons for loss and what was known about affected units before treatment, rather than treating an arbitrary percentage threshold as a universal certificate of validity.
Intention-to-treat analysis with missing outcomes requires a defensible approach to those unknown outcomes. Keeping everyone’s original assignment preserves the target comparison, but the words “intention to treat” do not recover information that was never observed. Automatically assigning every missing value zero imposes an outcome assumption.
Ask what the unknown values could change
Because the example’s outcome is binary, we can calculate simple logical completion ranges. If none of control’s ten unknown outcomes were successes, its full-cohort sample rate would be 18%. If all ten were successes, it would be 28%.
Treatment has forty unknown outcomes, so its corresponding range is 18% to 58%. Across possible completions of these missing entries, the all-assigned sample rate difference could run from −10 percentage points to +40 points.
These are logical missing-outcome completion bounds, not a confidence interval and not exact bounds on a population causal effect. They hold the observed entries fixed and vary the unknown binary values. Even if every missing outcome were recovered, inference about an underlying effect would retain uncertainty from the experiment’s design.
The exercise reveals how little the complete-case +10-point result establishes without assumptions about observation. Filling all missing values with zero happens to produce 18% in both arms here. That zero difference is one completion, not a demonstrated absence of an effect.
Attrition analysis can use sensitivity checks to make missing-data assumptions visible. A useful next step is to recover outcomes from reliable records, investigate why collection failed, and assess conclusions under plausible alternatives. The most favorable completion is not a substitute for that investigation.
Preserve the comparison after assignment
Different exclusions require different repairs
Distinguish incomplete observation from a known zero. An account that was fully observed and did not purchase can legitimately contribute zero to a purchase metric. An account that no longer joins to the purchase system has an unknown outcome unless other evidence establishes what happened.
Telemetry-loss research shows why lost records can affect power and bias comparisons. Increasing traffic through the same broken collection path need not fix the problem. It may narrow the reported interval around a distorted complete-case comparison.
Inspect the metric and Fact Table definitions, including identifiers, join types, filters, timestamp alignment, and outcome availability. A left join followed by a filter on an outcome-table field can silently become an exclusion rule. A metric name does not reveal which people survived that rule.
Post-treatment filtering is another mechanism. Restricting analysis to purchasers or successful feature users can select different people in each arm when treatment changes those behaviors. Identical SQL applied to both arms does not make selected membership independent of treatment. Use such a view for a clearly limited descriptive purpose or an appropriate causal analysis; do not assume it inherits the original randomized comparison.
Triggering needs a causal and implementation argument
Analyzing a relevant triggered population can be defensible when the trigger identifies comparable eligibility across assignments. The criterion is not simply whether logging happens before or after the assignment timestamp. Ask whether the tested treatment can affect who satisfies the condition or whether that condition is recorded.
GrowthBook’s activation-metric settings explicitly warn that activation should be independent of assignment. Where a feature exists only in treatment, a carefully designed counterfactual logging approach can record who would have encountered it in control. That requires equivalent logic and validation, not just a treatment-only click event with a new name.
Passing a sample ratio mismatch check is useful evidence about allocation counts, not a guarantee that every inclusion rule is valid. Selection can change the composition of two equally sized groups. Review reasons for missingness and baseline characteristics alongside counts, rather than forcing arm sizes to match by discarding extra observations.
No statistical repair is universal. Weighting and imputation require assumptions and relevant observed information. A group absent from the data cannot be studied merely by assigning it a larger weight. A missingness mechanism related to unobserved outcomes cannot be declared harmless because a model returned a precise estimate.
For example, an imputation model may assume observed baseline information is sufficient to account for who has missing outcomes. That assumption needs a substantive argument; the unobserved values themselves cannot be inspected to verify it. Compare plausible departures from the assumption and report whether the practical decision survives. Do not choose the missing-data model because it produces the preferred launch recommendation.
Report the evidence randomization actually preserved
Track assignment separately from outcome observation
Maintain counts for eligible units, assigned units, recorded exposures, observed outcomes, and the final analysis population. Explain losses between stages by arm and by relevant baseline groups. The ledger should distinguish a known outcome from unavailable measurement, and an original exclusion from a later analytic restriction.
The data-quality requirements for online experiments include completeness, consistent assignment records, reliable joins, and timing. An allocation check alone covers only part of that path. A clean A/A test can reveal existing problems but cannot prove a future treatment will leave exposure and outcome collection unchanged.
Preserve the query or metric revision behind the result. If the team repairs a join or recovers delayed records, state whether the correction reconstructs the intended original population. If it does not, explain the resulting scope and whether a new run is needed. Avoid relabeling a selected survivor cohort as the original experiment.
Design outcome observation before increasing exposure. For an important transaction metric, a reliable backend record may allow recovery when a client event fails, provided identity and timing still join correctly. Verify that the recovery method is available in both arms and measures the intended outcome. A second event stream is useful evidence only when its own coverage and meaning are understood.
Make missing-data assumptions part of the launch decision
Alongside the experiment results, report outcome availability and the assumptions used for missing values. Transparent randomized-trial reporting treats participant flow, exclusions, and missing-data handling as information needed to judge a result. The same statistical discipline is useful for product experimentation when the participants are accounts rather than clinical-trial patients.
For the hypothetical experiment, the immediate conclusion is that the favorable complete-case result does not identify the all-assigned sample comparison without more information. The next action is to investigate observation, recover outcomes where possible, and evaluate the remaining uncertainty. It is not to declare a winner from 20% versus 30%, or a tie from zero-filled missing outcomes.
Randomization gives you a defensible starting contrast. Preserving it requires that the data and analysis continue to describe the units and outcomes the decision concerns. Before launching from a randomized result, be able to explain not only how people received a variation, but also whose outcomes are known, whose are missing, and what your conclusion assumes about the difference.
Inspect the metric definition
Open the read-only metric search workflow to find a fact metric and inspect its aggregation, filters, identifiers, and associated fact table. Running the workflow requires configured GrowthBook API access; the instructions do not execute an audit when opened.
View the Metric Search WorkflowReady to ship faster?
No credit card required. Start with feature flags, experimentation, and product analytics—free.





