The number of scans tells you that people attempted to enter an AR campaign. It does not tell you whether the experience loaded, whether they understood it or whether it helped them take a useful next step. A measurement plan should distinguish those stages before the campaign goes live.
Start with the campaign's purpose and define a small number of meaningful observations. Then confirm what the implementation can actually measure. A modest report with clear definitions is more useful than a large dashboard whose numbers mix different actions and technical states.
Define the outcome before choosing events
The desired outcome might be understanding a product feature, reaching a specification page or starting a relevant enquiry. Choose an outcome that fits the campaign and can be evaluated honestly.
Avoid treating a business ambition as an automatically measurable event. "Improve brand trust" is not the same as a button click. It may require separate research, while a link opening can be observed more directly.
Write a simple chain from invitation to outcome. For example: reader opens the campaign, reaches the product explanation, explores a relevant feature and follows the enquiry route. This gives the team a basis for deciding where measurement is useful and where interpretation requires caution.
Separate entry from successful delivery
A visit to the landing page is different from the AR content becoming usable. If both are reported as engagement, loading problems can remain hidden behind an apparently healthy total.
Ask the developer which milestones are observable in the selected implementation. Possible distinctions include page opened, experience requested and content ready. These are proposed reporting concepts, not a claim that every AR platform exposes identical events.
Define each milestone in writing. "Content ready" should correspond to a specific state, not an arbitrary timer. If a state cannot be measured reliably, say so and use another method rather than giving a precise label to an uncertain signal.
Treat repeated attempts carefully
One person may open the campaign several times, retry after an interruption or use more than one device. Do not label every event as a different person unless the measurement method genuinely supports that interpretation.
Distinguish visits, sessions and actions in the report. Explain the definitions used by the selected tools and any limitations. A stakeholder should not have to guess whether a figure counts people, browser sessions or repeated button presses.
Avoid collecting persistent identifiers simply to make a small report appear more sophisticated. Decide what information is necessary for the campaign question and review the proposed collection with the appropriate organisational advisers.
Identify what a meaningful interaction looks like
Not every movement or tap deserves its own success metric. Choose actions connected to the message: viewing an explanatory sequence, comparing a relevant option or opening approved product information.
Be cautious about duration. A long session might indicate interest, confusion or an unattended device. A short session might mean the person found the answer quickly or left because the experience failed. Time alone does not establish quality.
Combine activity data with observation. A short usability session can help explain why people repeat a control or abandon a stage. Reporting becomes more useful when the numbers are interpreted alongside the actual journey.
Keep the next action connected to the campaign
Check what happens when the visitor leaves AR. Does the destination relate to the product or option they were viewing? Can they find the specification or enquiry form without starting again?
If the campaign links to another system, agree what can be observed across that boundary. Do not assume a click establishes that the destination loaded or that an enquiry was completed.
Discuss this handover when planning custom augmented reality development. The visible experience and its onward journey should be scoped together. A campaign can be entertaining while still making the useful next action unnecessarily difficult.
Include failures in the reporting plan
Successful activity is only part of the picture. Identify which technical or usability failures can be observed appropriately, such as content failing to load or a visitor reaching an unsupported route.
Keep error categories understandable and avoid collecting unnecessary detail from the user's environment. A report about loading failures does not automatically need camera images, precise location or personal information.
Provide a way to investigate important issues without exposing raw technical messages to visitors. The public experience should offer recovery or an alternative. Internal reporting can then help the team understand which part of the journey needs attention.
Test the numbers before launch
Run known test journeys and compare them with the recorded events. Complete the experience once, abandon it at a known stage and repeat an action. Confirm that the report reflects those journeys according to the agreed definitions.
Check how refreshes, interrupted sessions and return visits are handled. Exclude or identify internal testing where the chosen tools permit. A launch report should not unknowingly include repeated rehearsals by the project team.
Record the tracking version with the application version. If the event definitions change midway through the campaign, note the change. Two periods may not be directly comparable when the way actions are counted has changed.
Ask a few questions the data cannot answer
Invite representative participants to describe what they understood and what they expected next. Avoid questions that simply invite praise for the technology. Ask where the explanation was clear and where something was missing.
Use illustrative observations carefully. One person's comment can identify a useful problem, but it does not establish how all visitors feel. Keep qualitative findings and numerical counts distinct in the report.
Do not invent a percentage improvement from a small informal test. If the organisation needs a formal comparison, design it with suitable research expertise. A straightforward account of limitations is more valuable than an impressive claim the evidence cannot support.
Compare like with like
If a campaign is compared with an earlier version, identify changes in audience, placement, promotion and duration. A different invitation or distribution channel may affect results independently of the AR content.
Avoid attributing all subsequent enquiries or sales to the immersive feature. Attribution depends on the available evidence and the wider customer journey. A visitor may encounter several channels before acting.
Use comparisons to generate questions and prioritise improvements. A weak transition from landing page to experience might call for clearer instructions or faster delivery. A successful experience with few onward actions might require a better connection to the relevant product information.
Keep collection proportionate and explained
Map the services involved in reporting and decide who can access the information. Confirm retention and deletion arrangements with the responsible team. The measurement plan should reflect actual implementation rather than a generic privacy statement copied from another project.
Explain relevant collection to users through the appropriate interface and policy arrangements. Requirements depend on the implementation and context, so involve qualified advisers where needed. This is a planning process, not legal advice.
Do not assume that using AR requires recording the camera feed for analytics. Ask what is necessary for the experience and what, if anything, is separately collected for reporting. Distinguishing those functions makes the conversation clearer for both the organisation and the audience.
Report what you learned and what changes next
A useful campaign summary states the objective, definitions, observed activity, known limitations and recommended changes. Keep the conclusion connected to the evidence. If loading reliability is uncertain, resolve that before celebrating a large number of attempted launches.
For a new project, contact Augmented Reality South Africa with the desired outcome and reporting needs alongside the creative brief. Ask which measurements are feasible before committing to a dashboard.
The strongest measurement plan does not count everything. It distinguishes people reaching the campaign, people receiving a usable explanation and people taking a relevant next step. That makes it possible to improve the experience without confusing novelty, technical activity and genuine usefulness.