How should you brief a product tester without exaggerating the result?

01

How do you identify exactly what is being tested?

Tie every test and media file to the physical item the tester received. Footage from one model, formulation, or supplier should not silently become evidence for another.

Start the brief with a project ID, tester name, test date, location, commissioning company, and approval owner. Record whether the tester was paid, received a free product, has an affiliate arrangement, or received another incentive. If the footage becomes an endorsement, assess the applicable disclosure requirements. The FTC advises that material connections should be disclosed clearly and that video disclosures should appear where viewers are likely to notice and understand them.

Record the brand, product name, supplier, seller, SKU, model, size, color, formulation, and variant. Add the batch or lot number when available, plus the packaging version, instruction version, included accessories, product condition, delivery date, and assembly state. Ask for clear images of the unopened parcel, shipping label, packaging, identifiers, and contents.

For dropshipped products, this step is especially important operationally because listings and suppliers can change. If the received unit cannot be matched to the advertised item, label the result “product identity unconfirmed” and do not treat it as support for that listing.

02

How do you turn a proposed statement into an observable test?

Rewrite the desired statement as a neutral question that can be answered from recorded measurements or visible conditions. Do not ask the tester to “prove” the conclusion.

First classify the statement. Is it a subjective preference, directly observable attribute, performance claim, speed claim, durability claim, comparison, or health or safety claim? This distinction helps the reviewer identify what the test can realistically establish.

For “cleans instantly,” ask: “How much of the marked stain remains after one application and 30 seconds of wiping?” For “leakproof,” ask: “Does visible leakage occur when this filled container is inverted for five minutes?” For “works faster than Product B,” define both products, quantities, procedures, starting conditions, measurement method, and comparison outcome.

Do not turn a casual creator test into supposed proof of a health or safety claim. The FTC explains that anecdotal experiences are not a substitute for appropriate scientific support for causal health claims. A tester may accurately report a personal observation, but copy such as “treats chronic pain” or “cures eczema” is a broader claim requiring separate review and support.

1. Proposed statement: Write the exact sentence someone wants to use.
2. Claim type: Mark preference, attribute, performance, speed, durability, comparison, health or safety.
3. Observable question: Rewrite the statement without assuming a successful result.
4. Permitted conclusion: Define the narrow statement the setup could support.
5. Excluded conclusions: List broader statements the test cannot establish.

03

What test method should be agreed before filming?

Predefine the method so the result cannot be reverse-engineered after the team sees the footage. The brief should make a repeatable attempt possible without forcing a positive outcome.

Name one primary outcome: the measurement or visible change that answers the test question. Record the baseline before product use. If there is a pass/fail threshold, define it in advance rather than moving it after the result. State whether the comparison uses an untreated sample, another product, or no control.

Specify tools such as a timer, scale, ruler, thermometer, measuring vessel, or fixed visual reference. Document charging, assembly, cleaning, shaking, preheating, quantities, settings, distance, pressure, duration, surface, lighting, temperature, water type, and any relevant device configuration.

Set the number of attempts and a stop rule before testing. Record known limitations, such as a small sample, uncontrolled room conditions, subjective scoring, or inability to verify the batch. Google’s campaign experiment guidance similarly recommends defining a hypothesis and success metrics before a test and limiting changes so results are interpretable; that guidance concerns ad experiments, but the planning discipline is useful here too.

1. Primary outcome: What will be measured or observed?
2. Baseline: What is true immediately before use?
3. Method: What exact actions, quantities, settings, and durations will be used?
4. Conditions: What environmental or equipment factors could affect the outcome?
5. Attempts: How many complete attempts will be recorded?
6. Stop rule: When does testing finish?
7. Limitations: What conclusions will this setup not support?

04

What instructions should the tester receive?

Require complete, traceable documentation and explicitly give the tester permission to return an unfavorable or inconclusive result.

Ask the tester to record the product identifiers and baseline before the first attempt. Each attempt should run continuously from setup through outcome whenever practical. If filming stops, a product is swapped, the setup changes, or an unexpected event occurs, the tester should say so on camera and enter it in the log.

The tester must not replace a tested sample, measurement, or finished result with a cleaner-looking version. They should retain failed, interrupted, weak, and mixed attempts. An unusable camera angle can be marked as a production failure, but it should not disappear from the attempt history.

Ask for a neutral recap: what happened, what was measured, what deviated from the method, and what remains uncertain. Discourage scripted conclusions that go beyond the observation. “A mark remained after 30 seconds” is an observation. “It reduced the mark in this attempt” is a limited interpretation. “It removes every stain instantly” is an unsupported leap from that result.

05

What should the evidence log contain?

Create one log entry for every attempt, including attempts that fail, are interrupted, or cannot be interpreted. The log should connect the method, raw files, result, limitations, and review decision.

Use a stable attempt ID such as TEST-001-A01 in the slate, file names, folder names, and log. Never overwrite originals. Keep raw files separate from selects, edited exports, still images, and disclosure takes.

The most important separation is between observation and interpretation. “The timer showed 00:30 and a visible mark remained” is directly observable. “The product reduced but did not remove the mark in this attempt” is an interpretation limited to the recorded event. A general statement about all stains, all users, or typical results is not established by that single attempt.

1. Attempt ID, date, time, tester, and location
2. Product, SKU, variant, batch, packaging, and instruction version
3. Raw file names and storage location
4. Setup, baseline, conditions, and measurement tools
5. Procedure followed and any deviation from the brief
6. Direct measurement or observation
7. Tester interpretation, entered separately
8. Limitations and possible confounding factors
9. Result: pass, fail, mixed, inconclusive, or interrupted
10. Statements this attempt may support
11. Statements this attempt does not support
12. Retest requirement and reviewer decision

06

What footage package should the tester deliver?

Request enough material to reconstruct the test, not merely enough polished footage to build a favorable edit.

The delivery should include product identification, unboxing and condition footage, instructions, setup, baseline, every complete attempt, failed or interrupted attempts, measurement close-ups, final-state stills, and a neutral recap. Include the evidence log and raw-file index in the same delivery.

If the tester will appear as an endorser, request a separate disclosure take based on the actual relationship. Do not rely on a disclosure to correct a false demonstration or an unsupported product claim. FTC guidance treats endorsements as part of the advertising message: they must reflect honest experience and cannot communicate a claim that would be deceptive if made directly by the advertiser.

1. Folder 01: Product identification and unboxing
2. Folder 02: Setup and baseline
3. Folder 03: Raw attempts, organized by attempt ID
4. Folder 04: Measurements and final-state stills
5. Folder 05: Neutral recap and applicable disclosure takes
6. Folder 06: Evidence log, file index, instructions, and notes

07

How should the edit be reviewed for implied exaggeration?

Review what a reasonable viewer is likely to take away from the whole ad—not only whether the tester’s spoken words are literally accurate. Images, sequencing, overlays, reactions, timing, and omitted context can change the apparent claim.

The FTC says it considers both express and implied claims and examines the ad as a whole. Its Endorsement Guides include an example in which multiple roaster demonstrations took 45 minutes instead of the advertised 20, after which a chicken from a different roaster was substituted for the final commercial. The problem was the false demonstration created by the substitution and presentation, not merely one inaccurate sentence.

Check whether the edit uses the same product and variant, hides slower or failed attempts, or speeds footage up without making the time treatment clear. Compare before-and-after shots for lighting, angle, distance, camera settings, elapsed time, and product quantity. Examine whether captions, music, reaction shots, or a landing page turn a limited result into a broad promise.

A qualification should not contradict the central demonstration. If the footage appears to show instant removal, a small caption saying “results vary” does not document what actually happened. Change or remove the misleading sequence instead of expecting fine print to repair it.

1. Does the final edit show the tested SKU and the actual recorded result?
2. Were omitted attempts excluded for production reasons or because they weakened the claim?
3. Do overlays and voice-over stay within the documented observation?
4. Could speed changes, cuts, framing, or reactions distort elapsed time or effectiveness?
5. Are before-and-after conditions genuinely comparable?
6. Does the destination page broaden the claim beyond the footage?
7. Are material connections disclosed clearly where applicable?
8. Are important limitations visible at the point where viewers need them?

08

When should a statement be sent back for review?

Stop publication when the test did not measure the proposed claim, the product identity is uncertain, material setup changes went undocumented, or the final presentation communicates more than the evidence supports.

Use consistent review statuses: “supported for this documented attempt,” “potentially usable with qualification,” “inconclusive,” “contradicted by the test,” “requires additional substantiation,” and “do not publish.” Avoid the vague status “approved” unless it identifies the exact approved statement, footage, conditions, and version.

Send health, safety, superiority, typical-results, scientific, and guaranteed-outcome statements for appropriate substantiation review. Strong advertising performance does not validate the underlying product claim; click-through rate or sales can show how an audience responded, not whether the demonstration was truthful.

09

How can ATIYO preserve the test context?

ATIYO can keep the brief, brand context, assets, iterations, evidence-log decisions, and reusable learnings together. It does not independently verify a claim or know the product’s performance unless a user records the evidence and review outcome.

Create the product-test method in the brief, attach or reference the raw-delivery index, and record which statement each edit was intended to support. Preserve reviewer decisions and limitations alongside the creative version so a later iteration does not accidentally restore rejected language or omit a qualification.

Media performance remains in the ad platform, and ATIYO preserves creative context and learnings. ATIYO does not connect to ad accounts, buy media, calculate ROAS, or establish that an objective advertising claim is substantiated.