How Should Retail Teams Plan a New Product Merchandising Test Brief?

Answer first

Retail teams should plan a new product merchandising test brief as a bounded learning document for one identified product, retail context, and decision—not as a promise of sales, a substitute for compliance review, or a permanent planogram. Define the exact SKU and sample, the shopper or staff question, the test location or channel, the comparison cells, the observation period, the measures, the evidence owner, and the stop or change rules. Keep product claims, package information, language, battery or power details, price, MOQ, lead time, and destination requirements tied to the reviewed model, sample, retailer, and market.

A useful brief identifies the model, variant, package revision, identifier, sample basis, decision, test-cell variable, observation method, and continue, revise, hold, or stop rule before a unit reaches the test location.

Define the test question before designing the display

Start with one decision sentence. A strong example is: “For the named drawing-board SKU in the stated retailer fixture, does the revised product-identification cue help staff place the correct variant without obscuring required package information?” Another might be: “In two comparable stores, does a single-SKU presentation make the approved talking-pen pack easier for staff to identify and replenish than the current mixed presentation?”

Avoid questions that contain their own conclusion. “Will this attractive display increase sales?” is too broad unless the retailer has an agreed measurement design, comparable locations, a defined period, and an approved commercial method. Even then, a small merchandising test may show an association rather than establish causation. “Does the test arrangement reduce observed variant mix-ups during replenishment?” is more controllable.

Record the retailer or channel, destination and language where known, review window, decision owner, exact SKU and sample or artwork reference, fixture or digital context, and variables that may change or must remain constant.

Separate the product from the merchandising treatment

Keep two objects separate: the product record and the retail treatment. The first identifies the unit; the second describes where and how it is presented.

For an LCD tablet, do not turn a nominal screen size into a claim about overall dimensions, writing area, erase method, lock function, battery, power source, stylus, or intended user. For a talking pen, a device reference does not establish the audio language, compatible content edition, download method, battery arrangement, or included book. Each field needs the current SKU or sample source.

GS1 describes a GTIN as an identifier that can uniquely identify a trade item that is priced, ordered, or invoiced in the supply chain [1]. Use an item identifier to keep the tested unit connected to the retailer record, but do not assume that a barcode alone proves product attributes, package contents, compliance, or current artwork.

Choose a test design that can answer the question

The simplest useful design compares a defined current arrangement with one proposed arrangement. A more complex test may use two or more cells, but additional cells increase the risk of confounding variables. The table below is a planning aid, not a universal retail method.

Where possible, change one variable. If the new cell changes the fixture, wording, image, number of facings, adjacent products, and price at once, the result cannot show which change mattered. If the retailer cannot hold variables constant, label the test as an exploratory observation rather than a controlled comparison.

Define the unit of observation

Decide what one observation means. It could be one staff replenishment, one documented variant error, one shopper question recorded under the retailer’s approved process, one product-information mismatch, or one scheduled shelf audit. Do not mix units such as “a lot of interest” and “three corrections” in the same result column.

Specify who records each observation, when, and whether it is factual or interpretive. “Pack was returned to the wrong slot during replenishment” is factual; “the display was confusing” needs a defined basis. Use aggregate counts and issue categories rather than unnecessary personal data, following the retailer’s privacy rules.

Build the evidence pack before the test starts

A test cannot correct an unknown product record. Assemble a small evidence pack and attach a source or owner to every important field. Use statuses such as verified for this test, provided but not physically verified, retailer assumption, open conflict, and not supplied.

The pack may include the exact SKU, variant, identifier, package revision, dated sample or artwork images, relevant measurements and method, documented contents, approved wording, fixture rules, retailer-owned price fields, destination and language, required-information owners, and the test-cell diagram, observation sheet, and version number.

Do not put a certification logo, safety badge, rating, customer result, or environmental statement into a test asset simply because the team wants the display to look complete. A merchandising experiment does not create evidence for a claim.

For EU routes, the European Commission’s Access2Markets guidance explains that the GPSR applies from 13 December 2024 and covers consumer products, including products sold online; it highlights risk assessment, responsible economic operators, and broader product-safety accountability [4]. A retail test should therefore name the destination and responsible review route. It should not say “EU approved,” “GPSR compliant,” or “ready for all Member States” unless the responsible party has made that exact, market-specific determination.

Select measures that match the decision

Use a small number of measures that the retailer can actually collect. A practical brief divides them into operational, information, and commercial measures.

Operational measures can include setup completion, time to restore a facing, number of wrong-variant placements found during scheduled audits, missing sample components, damaged packs, or whether the intended fixture was used. Define the counting rule. For example, “wrong-variant placement” should state whether a correction found and fixed during the same visit counts once or is logged by occurrence.

Information measures can include whether the item identifier matched the shelf record, whether the displayed pack matched the approved artwork revision, whether staff could locate the relevant instruction or product page, and which questions were escalated. These measures are especially useful when the test is meant to reduce content or handover errors.

Commercial measures require the greatest caution. Units sold, conversion, revenue, margin, and returns can be relevant if the retailer owns the data and approves the comparison. State the source, period, comparison basis, and known limitations. Do not present an unbalanced store comparison as proof that one display caused a result. Do not publish a percentage, uplift, ranking, or customer outcome unless the retailer has authorized the data and the method supports the statement.

Set the observation period and review cadence in advance. A single visit can identify a setup problem but rarely supports a broad conclusion. If the sample, fixture, or package changes, mark the test incomplete rather than as a failure or success.

At this point, teams planning a SKU-specific test can email Toywink at info@toywink.com with the model reference, variant, test market, channel, proposed fixture, current sample or artwork revision, and the decision the test should inform. Product configuration, documents, language, battery or power, compliance, MOQ, price, sample status, lead time, and destination requirements remain subject to review for the named case.

Write the setup and execution instructions

The setup section should let another approved team member reproduce the test. State the placement, cell code, orientation, facing order, neighbouring products, sign or image version, price owner, sample handling, and replenishment route.

Add a “must not change” list. It may include the tested SKU, package revision, item identifier, approved wording, language version, product image, safety or instruction visibility, or the retailer’s price process. If a store needs to make a local adjustment, require a deviation note with the date, reason, owner, and photograph or other evidence allowed by the retailer.

If a loose sample or interactive demonstration is proposed, define who supervises it, what parts are present, how it is reset, where the sellable packaged unit sits, and how the sample is distinguished from included accessories. Do not claim that every model supports a demonstration, that a battery is included, or that a device can be used in a particular way unless the exact record confirms it.

If the test uses translation or local language, freeze the source copy and language version before setup. A translation file from another market is not automatically suitable for the named destination. Record the reviewer and artwork revision. Keep the test brief separate from any final packaging, label, warning, instruction, or legal review.

Create decision rules before looking at results

• Continue as configured for the defined test context: the setup was reproducible, the product and information matched, and the stated operational measure met the pre-agreed criterion.
• Revise and retest: one or more issues were observed, but the issue can be addressed without changing the product’s approved identity or creating an unsupported claim.
• Hold: identity, sample, package, language, fixture, document, destination, or ownership is unresolved.
• Stop the test: a safety concern, product mismatch, damaged or incomplete sample, non-approved public wording, or retailer instruction conflict makes continued testing inappropriate.

Use conditional language in the final result: “In the named two-store test, the revised placement was easier for staff to restore under the recorded observation method.” Do not write: “This display increases sales,” “shoppers prefer this product,” or “the product is proven to perform better” unless the evidence and authorization genuinely support those statements.

Add change triggers

List events that invalidate or reopen the brief. Typical triggers include a new model or variant, package or artwork revision, changed pack contents, altered audio edition or language, different battery or power arrangement, changed retailer fixture, new destination, changed price programme, revised instruction, new identifier, or a material product or manufacturing change.

When a trigger occurs, preserve the old result with its original scope. Do not overwrite it as if the new configuration had been tested. Create a new brief or controlled revision and identify which evidence must be rechecked.

New toy merchandising test brief checklist

Before a retail team releases the brief, confirm:

• The question names one decision, one product context, and one intended use of the result.
• The exact SKU, model, variant, package or artwork revision, sample basis, identifier, and destination are recorded or clearly marked open.
• Product facts are separated from fixture, copy, price, adjacency, and other merchandising choices.
• The current and proposed test cells show what changes and what remains constant.
• Every measure has a definition, observation unit, owner, timing, and recording method.
• Commercial data, if used, has a retailer-approved source, comparison basis, period, and authorization.
• The test does not create or imply certifications, safety conclusions, developmental outcomes, customer results, guarantees, or universal suitability.
• Instructions, warnings, markings, traceability, language, and destination questions have responsible owners.
• Any battery, power, audio, accessory, MOQ, price, sample, lead-time, or configuration statement is tied to the named model and record.
• A loose sample or demonstration has a reset, supervision, identity, and pack-visibility method where applicable.
• Stop, hold, revise, and continue rules were agreed before results are reviewed.
• Change triggers and version history are defined.
• The final record states what the test did not prove.

FAQs

Is a merchandising test the same as a product test?

No. A merchandising test examines a defined retail presentation, information arrangement, handling process, or placement decision. It does not replace product safety testing, conformity assessment, quality inspection, battery review, language review, or destination-market analysis. A test may identify that a package is hard to identify or replenish, but it cannot certify the underlying product.

Can a retailer test several educational toy models in one brief?

Only when the brief defines a clear shared decision and preserves each model’s identity. If an LCD tablet, talking pen, and drawing board differ in operation, pack contents, language, power, instructions, or market treatment, use separate product records and test cells. A category-level observation can be recorded, but it should not turn one model’s evidence into a claim about all models.

Should the test use sales as its main success measure?

Not automatically. Sales may be useful when the retailer controls reliable data and approves a suitable comparison, but a short or uneven test may not isolate merchandising from stock, price, season, traffic, placement, or assortment effects. For an early test, operational measures such as correct placement, staff questions, pack matching, or setup consistency may answer the decision more directly.

Can a supplier render be used as the test display?

A render can communicate a proposed concept, but it should be labelled as a concept and tied to the intended SKU and revision. It does not prove that the actual package fits, that a fixture accepts the orientation, that a sample includes every depicted item, or that the retailer has approved the arrangement. Validate the proposal against the actual approved sample, packaging, fixture, and retailer instructions.

What should happen if the sample changes during the test?

Pause the affected test cell and record the change. Identify the old and new sample, package or artwork revision, product contents, language, power or battery arrangement, and date. Decide whether the new configuration is comparable. If not, close the original test with its original scope and issue a revised brief rather than combining the observations.

Can the test prove that a talking pen improves learning?

No. A shelf or presentation test can examine discovery, identification, staff explanation, or a defined interaction under the retailer’s method. It cannot establish a learning outcome, language suitability, audio quality, or educational effectiveness merely because a shopper engages with the product. Those claims require their own evidence and responsible review for the named product and market.

Who should approve the final brief?

The retailer should name owners for merchandising, product data, packaging or content, store operations, commercial data, and relevant destination or compliance review. One person may hold several roles in a small team, but the decision record should show who approved each part.

For a SKU-specific merchandising-test discussion, email Toywink at info@toywink.com with the current product reference, proposed test cells, destination, language, retailer context, and evidence gaps. Email discussion does not itself constitute retailer approval, certification, availability, or a market conclusion.

References