Skip to content

Activate Measure

Measure what product information changes

Connect the exact product truth and passport version people see to the actions and outcomes that follow. Define the denominator, comparison, time window and costs before calling a result impact or ROI. Available now for configured deployments.

Measure one product intervention See how the engine decides
One governed identity, one defined treatmentfive steps
01

Define

One intervention, one eligible population, one primary outcome, one decision threshold.

02

Instrument

Product identity, passport version, treatment and exposure rule bound together.

03

Observe

Machine requests, human interactions and verified outcomes kept apart.

04

Compare

A holdout, phased rollout or pre/post design agreed before results are read.

05

Account

Benefit and cost for a named actor, period and scope, with the uncertainty beside it.

Direct answer

A scan tells you something happened. It does not tell you what changed. Activate Measure separates the observation from the exposure, the exposure from the outcome and the outcome from the conclusion, then joins only the signals the measurement design allows and reports the result at the strength the comparison supports. One governed product identity, one defined treatment, and a result you can explain.

Last verified
A packaging carton beside a laptop showing a compliance dashboard with product rows and status marks.

A result you can explain, or no result

The number is the easy part. What makes it worth acting on is the population it was measured over, the comparison it was measured against and the costs that were counted. Activate Measure keeps those beside the figure.

A scan tells you something happened. It does not tell you what changed.

Activate Measure separates observation, exposure, action and outcome. It joins only the signals the measurement design allows, then makes the comparison and its limits visible.

SignalA resolver request, page view or inspection event
ExposureAn eligible unit received the defined product-information treatment
OutcomeA pre-set operational or commercial result occurred within the measurement window
ConclusionThe result is reported at the strength the comparison supports, from descriptive to causal

The method behind this is written down rather than asserted, and you can read it before you buy anything: explore the measurement method.

How to test whether product information changes business outcomes

From an observation to a conclusion, and what each step has to earn

Every step admits fewer things than the one before it. That is the point, and it is why an exposure count is not an outcome.

SIGNAL

Observed

A request reached a resolver or an endpoint. That is all it establishes.

request
EXPOSURE

Eligible

An eligible unit received the defined treatment under the agreed rule.

eligiblerulebots excluded
ACTION

Attributed

A human interaction the design allows to be joined to that exposure.

actionunjoinable
OUTCOME

Verified

A pre-set operational or commercial event inside the measurement window.

outcomeincomplete
RESULT

Stated

Directional or causal, with the comparison, period and uncertainty named.

directionalcausalinconclusive

One system, three outcome views

Digital journeysQualified actions after a defined information exposure: discovery, basket action, enquiry or completion where lawful joining is possible
Operational workflowsCompletion, exceptions, handling time, inspection outcomes, repair routing or returns against a known denominator
Commercial effectIncremental revenue, contribution or avoided cost, only where the design supports attribution

ROI is kept separate from ROAS and from raw activity.

Measure the version that was actually shown

An outcome is only interpretable when it is tied to the exact product, claim, evidence and passport version in treatment. Activate Measure uses the same governed record as the rest of your passport estate. It does not create a parallel analytics identity.

Product identityGS1 identifiers and Digital Link syntax stay standards-aligned where they are used
Version identityWhich content and evidence state formed the treatment
Treatment identityWhat changed, who was eligible, and when it ran
Outcome identityThe event defined once, then applied consistently

Where the identity itself needs watching rather than measuring, that is a different question with a different answer: Activate Protect asks whether a use is consistent with the issued unit, and can protect the same identity this measurement runs on.

Build a GS1 Digital Link and see the URI it produces

No denominator, no rate. No comparison, no uplift.

Every result shows its population, exposure rule, comparison, period and limitations, and the state it is in is a word rather than a colour.

Setup requiredDefinitions or instrumentation are incomplete, so nothing is measurable yet
Collecting dataValid data is arriving and the threshold or window is not complete
Ready for analysisDenominator, exposure and outcome data pass the defined checks
Directional associationAn association is observed without sufficient causal identification
Estimated causal effectA pre-agreed design and analysis support it, with confidence and scope
InconclusiveThe data or the uncertainty does not support a decision
Results cannot be comparedComparison integrity failed

Nothing becomes proven because a chart moved. Caused is used in one state only; everywhere else the words are measured, observed or associated with.

Value is different for every actor, so ROI names its owner

A brand may value conversion or return reduction. A retailer may value catalogue quality or handling efficiency. A repair partner may value diagnosis and routing. The actor, the benefit and the cost scope stay visible in the calculation.

Illustrative equation: ROI = (attributed benefit minus scoped cost) divided by scoped cost.

Beside every figure sits the actor, population, period, currency, benefit type, the cost components counted and excluded, the attribution method, a confidence or scenario range, and the version and timestamp it was produced from. If a mandatory field is missing, the calculation is shown as incomplete rather than as one authoritative number.

The equation is simple. Defensible inputs are the product.

Plan, monitor and review in one governed workspace

Configure the intervention, monitor instrumentation and denominator health, then review results with their evidence and limitations. The same identity and event layer serves both products without either one writing the other's conclusion back as a product fact.

  • the intervention and its eligible population
  • exposure health, and what was excluded as bot, duplicate or technical traffic
  • the primary outcome and comparison health
  • the result state, with benefit and cost scope
  • evidence and limitations, disclosed beside the result

Activate Protect looks for identity and use anomalies. Activate Measure evaluates defined outcomes. An anomaly signal never becomes a commercial conclusion.

Product identity protection, which asks a different question of the same event

Join only what the question requires

Aggregate or pseudonymous data where possible, with consent, purpose and retention controls explicit. Personal data, raw evidence uploads and customer-sensitive product content do not go into general web analytics.

Questions this page is asked

Does a passport scan prove engagement?No. It proves a resolver or endpoint received a request. Human engagement needs an agreed rule and bot, duplicate and technical traffic controls.
Can you calculate ROI from scans alone?No. ROI requires a defined benefit, a scoped cost and an attribution method, for a named actor and period.
Do we need an A/B test?Not always. The right comparison depends on the decision and the deployment. Where a causal claim matters, a randomised or otherwise credible controlled design is preferred.
Does this replace our analytics platform?No. Activate Measure governs the product, treatment and measurement logic, then integrates with the analytics, commerce and operational systems you already run.
Does it work with Activate Protect?Yes. Both use the same governed item identity and event layer. Protect looks for identity and use anomalies; Measure evaluates defined outcomes.
Is it available now?Yes, for configured deployments. Scope depends on the intervention, data sources, permissions and comparison design.

A question this page does not answer?

Ask it directly. A real person reads these, and the reply will tell you what a measurement design can establish for your products and what it cannot.

hello@activatedigital.aiOpen it in your mail app

Where the underlying work is written down

MethodTesting whether product information changes business outcomesThe accepted method, with its comparisons and its limits.EvidenceWhat a scan actually tells youThe difference between a request, an interaction and an outcome.ProtectProtect every product identityThe same identity and event layer, asked a different question.CategoryA Battery Passport tool for the EU Batteries RegulationOne of the category products whose records a measurement runs on.IdentityWhat GS1 Digital Link resolves and how it carries a passportThe identity a measurement binds its treatment to.EngineHow evidence becomes a governed answerResolution, provenance and what a blank means.

One intervention. One outcome. One defensible answer.

Tell us what product information you want to change and the decision you need the result to support. We will define the identity, exposure, comparison, outcome and cost scope before implementation.

Measure one product intervention