ecomcloser.com

Brown dog wearing a blue harness outdoors

How to guide pet accessory selection using published size charts

An ecommerce pet supply product finder helps shoppers choose a product or variant by comparing what they need with published catalog information. For a dog harness, it should ask for chest girth and, when the manufacturer requires it, neck circumference. It should then compare those measurements with the chart for each model. A dog with a 47 cm chest can be in range for one model’s Small and outside another’s.

Build the finder to report a chart match, not promise a fit. If a required measurement is missing, ask for it. If a model falls outside the chart, explain which measurement ruled it out and check an alternative against its own chart. If there’s no documented match, offer store follow-up rather than steering the shopper toward the nearest size.

Define what the product finder can recommend

Start with a narrow decision: which stocked harness variants have published ranges that cover the dog’s measurements? That’s more useful than asking the finder to name the “best harness,” which could mean anything from a preferred style to a particular use. Keep those preferences separate from the size check. A shopper’s favorite color can narrow the choices, but it can’t make an out-of-range size a match.

For each model, identify the measurements its maker uses and record the instructions that go with them. Then map each row in the chart to the sellable variant it describes. A product feed can identify a variant as Small or group related variants under one item group, but those fields don’t tell you where to measure or what chest range Small covers. Keep the manufacturer’s chart attached to the model it describes, even when a neighboring model has identical size labels.

Give each variant one of three outcomes: within the published range, outside the published range, or not enough information to compare. The third outcome matters. A missing neck measurement isn’t evidence that the neck is out of range. A missing neck range on the store’s copy of a chart isn’t evidence that the maker doesn’t use neck circumference.

Check availability separately. A chart match may be unavailable in the shopper’s chosen variant, while an in-stock harness may fail the measurement check. That distinction determines what you tell the shopper: “This size is in range, but the selected variant isn’t available” describes a different problem from “This model’s chart doesn’t cover the chest measurement you gave.”

When the manufacturer updates a chart, verify that the change applies to the version the store sells before replacing the recorded ranges. If you can’t connect a chart to the stocked version, hold that model out of automatic recommendations. The rest of the catalog can still be compared using its verified charts.

Ask for measurements the shopper can take correctly

Ask for chest girth as a circumference, not a width. Tell the shopper where to place the tape using the instructions for the model under consideration. If the maker’s diagram marks the widest part of the chest, say to measure around that point. Don’t reuse that wording for another harness unless its instructions agree. The same principle applies to the neck: the point indicated on a harness diagram may not be where the shopper normally measures for a collar.

Make the voice prompt short enough to follow while holding a tape: “What’s the chest measurement around the point shown in this model’s size guide? Please include centimeters or inches.” If the shopper can’t see the diagram, describe the location from the maker’s written instructions when those instructions are clear. If the location can’t be explained reliably without the illustration, direct the shopper to the size guide before comparing numbers.

Ask for neck circumference before declaring a match whenever the chart requires it. You don’t necessarily need both answers up front. With a 47 cm chest reading, you can eliminate a variant whose chest range ends at 44 cm without asking for the neck. For a variant that passes the chest check, explain why you need another number: “The chest is in range for this size. Its chart also checks the neck.”

Read back an ambiguous measurement rather than guessing what was said. “Was that 47 centimeters around the chest?” resolves the value, unit, and measurement in one question. Convert units consistently for the comparison, and keep enough precision that rounding doesn’t turn a number just outside a range into one inside it. Show the shopper the values you actually used.

If the shopper only knows the current collar size, a previous harness label, or the dog’s weight, don’t substitute that information for a required circumference. Ask for the missing measurement. If they can’t take it now, leave the comparison open rather than choosing a size on their behalf.

Use two model-specific charts to show why Small isn’t universal

The charts below are illustrative examples, not specifications for products sold by EcomCloser or any merchant. Think of Harness A and Harness B as two catalog entries, each with separate sellable Small and Medium SKUs. Both example charts require chest and neck circumference. For this exercise, assume the makers use the same measuring points and explicitly include the endpoints shown. For real products, check those details rather than carrying the assumptions over.

Illustrative chart for Harness A
Size and example SKU Chest girth (cm) Neck circumference (cm)
Small (A-S) 42–49 30–36
Medium (A-M) 48–56 35–42
Illustrative chart for Harness B
Size and example SKU Chest girth (cm) Neck circumference (cm)
Small (B-S) 37–44 27–33
Medium (B-M) 45–52 32–38

At 47 cm around the chest, Harness A Small passes the chest check and Harness B Small does not. Harness A Medium also fails that check because its range starts at 48 cm. Harness B Medium remains a candidate. With no neck measurement yet, neither of the remaining candidates is a confirmed chart match.

Compare the measurements against the same row. The finder must not use the chest range from one size and the neck range from another, even though both rows belong to the same model. If it does, it can report a match that no sellable variant supports. Keep model, size, required measurements, and variant identifier together when storing or reviewing the chart.

Preserve qualifiers that affect the comparison. If the maker publishes different measuring instructions or a note for shoppers between sizes, keep that information with the relevant chart. If a number is presented as a garment dimension rather than a body measurement, don’t assume it can be used as a chest or neck range. If its meaning is unclear, ask the store to verify it.

Before publishing, check that the size labels on the product page, in the finder, and in the cart all refer to the same variants. Read the range back with its measurement name and unit, too. A correct number attached to the wrong measurement can create a recommendation as misleading as selecting the wrong SKU.

Walk through a 47 cm chest match without overpromising

Suppose a shopper asks, “My dog is 47 centimeters around the chest. Can I get a Small?” The useful response isn’t a blanket yes or no. If both examples are being considered, say: “Harness A Small covers a 42 to 49 centimeter chest. Harness B Small stops at 44. To finish checking Harness A Small, I need the neck circumference at the point shown in its size guide.”

Now suppose the shopper gives a neck measurement of 34 cm. Harness A Small covers both numbers: 47 cm is within 42–49 cm for the chest, and 34 cm is within 30–36 cm for the neck. A clear recommendation is: “Both measurements are within the published range for Harness A, Small.” The statement tells the shopper what was checked without promising how the harness will sit when tried on.

Harness B Medium also covers 47 cm at the chest and 34 cm at the neck. If the shopper wanted Harness A first and that variant is available, there’s no need to list every other match. If they prefer Harness B or the original choice is unavailable, name Medium as a separate option and explain that the maker’s sizing differs by model. Don’t present it as a substitution the shopper has already accepted.

Change only the neck measurement to 37 cm and the decision changes. Harness A Small tops out at 36 cm at the neck. Harness A Medium covers the neck but starts at 48 cm at the chest, so it isn’t a documented match either. Harness B Medium covers both 47 cm and 37 cm. That makes it a chart-supported alternative, not simply the next size up.

Keep those details in the recommendation even if the shopper asks for a quicker answer. “For 47 chest and 37 neck, Harness B Medium is in range; Harness A Small stops at 36 at the neck” is short enough to say aloud and specific enough to explain why the answer changed.

Handle boundaries and overlapping ranges without guessing

Using the example charts, a dog with a 49 cm chest and a 35 cm neck falls within both Harness A Small and Harness A Medium. Small reaches 49 cm at the chest and covers 35 cm at the neck. Medium starts at 48 cm at the chest and includes 35 cm at the neck. Under the example’s endpoint rule, both are legitimate chart matches; the figures alone don’t tell you which one the maker prefers.

For a case like this, present the two relevant options and look for guidance attached to that model. The maker may publish instructions for overlapping ranges or for readings at an endpoint. Relay those instructions as written rather than turning them into a general “always size up” rule. If the maker gives no preference, say so. The finder can establish that both sizes are listed without claiming that one will feel better.

Near an endpoint, ask whether the shopper took the measurement at the specified point. If the shopper says “about 49,” invite them to measure again before treating that value as exact. If two readings would lead to different results, explain the difference: “At 49 cm both chest ranges apply. If the repeat reading is above 49 cm, Small is outside this chart’s chest range.” Keep the neck condition visible in the final comparison.

An overlap and a gap call for different answers. In an overlap, more than one row can satisfy all the required dimensions. In a gap, none does. Don’t average numbers, add a tolerance of your own, or borrow an endpoint from a neighboring size. If the chart’s wording leaves endpoint inclusion unclear and it would change the outcome, flag it for verification.

Where both sizes are in range and the shopper doesn’t know which to pick, offer the maker’s published fitting instructions or store help with that choice. Leave the cart unchanged until the shopper chooses the specific size. Reporting two matches is a useful answer, even without resolving the final preference.

Offer an alternative that passes its own chart

An alternative should solve the problem that ruled out the original choice. If the shopper’s 47 cm chest rules out Harness B Small, Harness B Medium can stay under consideration because its chest range is 45–52 cm. Confirm the neck is within 32–38 cm before recommending it. If the alternative uses a different measurement point, ask for the reading that its maker describes instead of reusing the old value automatically.

Keep a size mismatch separate from an availability issue. If Harness A Small covers both readings but the desired variant isn’t available, say that the measurements match the chart and the selected variant isn’t available. If another available variant is covered by the same chart, you can present it without implying the dog needs a different size. If the alternative is another model, give its own comparison and make the model change explicit.

For a catalog gap, try an illustrative request with a 58 cm chest and a 38 cm neck. Harness A’s largest listed chest range stops at 56 cm, and Harness B’s stops at 52 cm. If these are the only models in scope, say: “I don’t have a chart match among these harnesses for a 58 cm chest. The largest published chest range here ends at 56 cm.” There’s no need to call the neck measurement a problem when the chest already rules out both models.

If the store has another appropriately sized item, check every required reading against that item’s published chart before naming it. Otherwise, offer follow-up. The question for staff should be specific: is there a stocked model with a published neck and chest range that covers these measurements, using the maker’s measuring points? Don’t promise the store can source a suitable item.

Avoid presenting the nearest size as a fallback. A short explanation of what fell short is more useful than “no results”: it helps the shopper spot a misheard number or decide whether to ask the store to check another model.

Hand off cases that a size chart can’t settle

Separate missing shopper information from missing product information. If Harness B Medium’s chart is complete but the shopper hasn’t supplied the neck reading, ask for it. If the shopper supplies both readings but the store cannot verify a required neck range for that variant, repeating the question won’t help. Say that you can compare the chest but cannot verify the neck for that model, and offer store follow-up or a documented alternative.

Escalate chart conflicts instead of picking the version that produces a match. A product description might show one range while a chart supplied for the stocked version appears to show another. Identify the model and variant, and ask the store to verify which range applies. Until then, mark that variant as unverified. A discrepancy on one model doesn’t make another model’s documented chart unusable.

Other handoffs are about the shopper’s question rather than the data. Someone might ask whether a harness will sit correctly in an area they’re concerned about, or request a guarantee that it won’t slip off. A pair of circumferences cannot answer those questions. Explain what the chart confirms, point to the maker’s fitting instructions where relevant, and offer to pass the unresolved question to the store. Don’t turn an in-range result into a guarantee about the harness in use.

Send staff a concise summary when a handoff is available: the model and size considered, the measurements and units, the chart finding, and the exact question still open. For example: “Harness B Medium, chest 47 cm and neck 34 cm. Chest is in range. Please verify the published neck range for the stocked variant.” Present this as a request, not as confirmation that someone has reviewed it.

Make it possible for the shopper to stop without a cart action. A good handoff preserves the comparison already completed so they don’t have to reconstruct it, but it doesn’t convert an unresolved size question into an accepted selection.

Keep voice selection, cart confirmation, and payment separate

Use distinct language for the recommendation and the action. “Harness A Small is within the published ranges for both measurements” is a recommendation. “Would you like to add one Harness A, Small?” asks permission to act on a specific variant. A shopper who says “that sounds right” while discussing measurements may be acknowledging the comparison, not accepting an add-to-cart prompt. Ask directly if the intent is unclear.

In a voice flow, say the model and size together when requesting acceptance. Include the quantity and any other choice needed to identify the variant. This is particularly important when discussing Small in two models or changing the selection from Harness A Small to Harness B Medium. “Add the Medium?” is too vague if both models have a Medium. Make the spoken confirmation as specific as the cart line the shopper should see.

After the shopper accepts the cart action, use the selected product or variation identifier and quantity for the cart request. In a WooCommerce implementation, use the cart response to verify what the cart actually contains. Report the model, size, and resulting quantity from that state. If the request fails or the intended variant can’t be verified, say the addition hasn’t been confirmed. Don’t announce success based only on what the shopper asked for.

Keep payment language for checkout. Selecting a size by voice is not a cart update. Accepting an add-to-cart action is not payment. An appropriate post-cart message is: “Your cart now contains one Harness A, Small. You can review it at checkout.” Avoid “purchased” or “order complete” at this step.

If you use a mock preview in training or merchandising, label it as illustrative. Showing an example cart line can demonstrate how model and size should appear together, but it does not establish that a real cart changed or that a payment occurred. Test those outcomes separately from the preview and from the accuracy of the size recommendation.

Test the finder with decisions a shopper can verify

Build a small test sheet around real chart rows before expanding the finder to more models. List the shopper’s inputs, the range that controls the outcome, the expected reply, and whether a cart action should be offered. Test the spoken answer as well as the underlying comparison. A correct match calculation still leaves the shopper guessing if the reply names a size but not its model.

  • In range: Enter a 47 cm chest and a 34 cm neck. For the example Harness A, expect Small to be described as within both published ranges. Remove the neck answer and verify that the finder asks for it instead of repeating the recommendation.
  • Boundary and overlap: Enter 49 cm at the chest and 35 cm at the neck for Harness A. Expect both Small and Medium. Move the chest reading just above 49 cm and verify that Small drops out without the reading being rounded back into range.
  • Missing data: In a test copy, remove the required neck range from Harness B Medium. Verify that the finder reports an incomplete chart rather than approving the size on chest alone or claiming the neck is outside the range.
  • Alternative and gap: Use 47 cm and 37 cm to check the switch to Harness B Medium. Then use 58 cm and 38 cm to verify that neither example model is offered as a documented match.

Finish by testing the action separately. Hear an exact variant confirmation, decline it, and verify that nothing is added. Accept it in another run and compare the returned cart line with the spoken selection. Try a rejected request as well: the response should say the addition wasn’t confirmed, not imply that the item was bought.

Repeat the relevant tests when a chart, variant mapping, or stock offering changes. The test passes when the shopper can tell which numbers were checked, why a size stayed in or dropped out, and whether a cart action actually happened.

Try the EcomCloser demo

Related guides

Sources