ecomcloser.com

Shopper holding a phone and a card beside shopping bags

Seven conversational commerce examples for online stores

Seven conversational commerce examples for online stores are checking whether a sofa fits a wall, comparing the cut of two jackets, filtering lipsticks by finish, matching a laptop to a sleeve, finding a replacement handle by hole spacing, choosing a dog harness by chest measurement, and finding a gift that fits a budget and delivery date. In each case, the shopper asks in ordinary language and the store turns the question into a check it can actually perform.

The useful answer is not always a product recommendation. A furniture shopper may need a measurement check and a room preview; a beauty shopper may need a swatch tied to the right shade; a gift buyer may need a delivery check before seeing any products at all. The examples below show the question, the catalog facts behind the answer, and the next action that moves the decision forward. The products and measurements used to illustrate those decisions are examples, not live offers.

1. Furniture: Check the wall before previewing the sofa

Shopper question: “Could this sofa work on a 235 cm wall?”

Catalog facts needed: The overall width of the exact configuration, the orientation of any chaise or return, and a dimension drawing that makes clear what is included in the measurement. For a corner sofa, the shopper also needs the depth of the projecting section. If the store is going to help check the delivery route, it needs the dimensions of the relevant packages or pieces, not just the assembled sofa.

A useful answer starts with the arithmetic. If an example sofa is 228 cm wide, it leaves 7 cm across a 235 cm wall before allowing for anything that reduces usable space. Ask whether the shopper measured between the actual obstructions: skirting, a radiator, trim, or a door that swings across the wall. A sofa can fit against the wall and still interfere with the door. And whether it fits once assembled is a different question from whether the pieces can get through the entry.

Next action: offer a measurement check, then a preview. Show the dimension drawing for the selected configuration and ask the shopper to mark the wall ends and any door swing. If a room-preview tool is available, open it with the same orientation the shopper selected. IKEA Kreativ is a reference point for the room-design approach. A mock preview is illustrative: it can help a shopper picture the arrangement, but it does not verify clearance, scale, or delivery access.

For Senoji Baldinė, keep the owner-described workflow separate from the public product listing. The owner’s account describes a workflow; it is not evidence that a particular sofa fits or that every described step is available for every product. Use the public listing to verify only the facts it actually publishes for the configuration in question. If a needed dimension is missing, flag it for confirmation rather than inferring it from a product image or the owner’s description.

What to measure: Track how often a wall-fit question ends with a checked configuration, a request for a missing dimension, or a deliberate decision not to proceed. A corrected orientation before checkout is useful even if the shopper does not order in that session.

2. Fashion: Compare the cut, not the size labels

Shopper question: “Which of these jackets is cut more loosely?”

Catalog facts needed: Garment measurements for the sizes the shopper would actually consider in both jackets, plus the measurement method. Chest width is a useful starting point; shoulder width and room at the hem may matter depending on where the shopper wants extra space. Keep garment measurements separate from a size chart that shows the body measurements a brand uses to assign a size. They answer different questions.

Ask one question that clarifies what “more loosely” means: “Do you want room for a sweater across the chest, or a looser shape at the hem?” Then compare the relevant measurements. In an example comparison, jacket A in the shopper’s likely size might measure 56 cm across the chest laid flat, while jacket B in their likely size measures 59 cm using the same method. That supports saying B is wider across the chest in those sizes. It does not establish that B is roomier at the shoulders or that the shopper should order it in the same letter size.

Next action: show a side-by-side measurement comparison. Label the sizes and where each measurement was taken, then offer the size charts. A shopper who is between sizes can measure a jacket they already like in the same way and compare it with both options. If one supplier gives flat garment widths and the other gives only body measurements, show the difference in the source data instead of ranking the jackets by numbers that are not directly comparable.

Where photo try-on is available, offer it alongside the charts, not in place of them. Google Shopping’s photo-based clothing try-on is an example of a way to preview appearance. An illustrative image may help the shopper compare silhouettes, but it does not tell them whether a sweater will fit comfortably under the jacket or whether they can move freely in it.

What to measure: Count comparisons that can be answered with measurements taken on the same basis. Also record which dimensions people ask for after seeing the comparison. If the same shoulder question keeps coming up and the catalog cannot answer it, the fix is to add a measurement, not to make the chat response more confident.

3. Beauty: Make the satin filter lead to the right shade

Shopper question: “Which of these lipsticks has a satin finish?”

Catalog facts needed: The finish wording used on the product label, the shade name and stock-keeping unit (SKU), the swatch assigned to that shade, and availability for that exact variant. Check whether the finish is recorded for the whole line or separately by shade. If the product page groups multiple finishes together, relying on its title alone can send the shopper to a shade that does not meet the request.

Start with products or variants explicitly labeled “satin.” If another label says “cream,” show it as a separately labeled option only if the shopper wants to broaden the search. Do not silently treat the two terms as interchangeable. Likewise, if the only evidence is a product image with a soft sheen, say the finish is not confirmed. Filtering should still work when merchandising copy uses a different description from the actual finish field.

Next action: show a labeled swatch shortlist. For every swatch, display its shade name and finish in text, and make the selection resolve to that shade’s SKU. In an illustrative example, the shopper picks an example shade called Rosewood from the satin results. The next view should show Rosewood selected, not the default shade on the lipstick page. Keep the swatch illustrative rather than promising an exact color match, and make sure the name and finish are usable without relying on color alone.

The most revealing test is to change the availability of one shade while leaving the rest of the lipstick line in stock. Selecting that shade should produce an honest availability message, not an available product-level status or a quiet substitution. If the shopper wants another shade, keep the satin filter in place. Let them choose whether to broaden the finish or color rather than relaxing both conditions on their behalf.

What to measure: Check whether the swatch, selected shade, and cart variant agree. Track the share of finish questions that return at least one confirmed match, but keep “no satin shades available” separate from “finish data missing.” The first describes the assortment at that moment; the second identifies a catalog gap.

4. Electronics accessories: Verify the laptop and the sleeve opening

Shopper question: “Will this sleeve fit my computer?”

Catalog facts needed: The external width, depth, and thickness of the exact laptop model; the sleeve’s usable internal dimensions; and any published limit on the opening or entry point. Get the model identifier if the product family includes more than one chassis, and ask about a hard shell or other cover if the shopper plans to leave it on. The advertised screen diagonal may help them navigate the store, but it does not supply the three dimensions needed for this check.

Use a concrete comparison. If an example laptop is 31.5 cm wide, a sleeve with a listed internal width of 32 cm looks plausible on that axis. But if the documented usable opening is only 31 cm in that same direction, the interior figure alone does not support calling it a match in the intended insertion orientation. If the opening is not documented, the answer should say what checks out on paper and what still needs confirmation. Avoid promising that an undocumented amount of stretch will make up the difference.

Next action: give a compatibility verdict with conditions. A useful result might be “Internal dimensions match the bare laptop; the opening is not confirmed.” Show the closest matching sizes with the reason each is included or ruled out. If the manufacturer publishes an exact-model compatibility entry, distinguish that from a fit inferred by comparing measurements. If the shopper only knows the screen size, ask for the model identifier or explain which outside dimensions to measure instead of guessing a sleeve size.

Where the customer has an existing cover, check the combined dimensions or explicitly limit the answer to the bare laptop. Also confirm that the stated sleeve measurements belong to the selected size, not to another size on the same product page. Those two checks matter most when the margin is small and a shopper expects a snug fit.

What to measure: Track outcomes separately for documented matches, dimension-based matches, and unresolved checks. Review fit-related support questions or return reasons against the original verdict. If shoppers repeatedly come back with the same unknown opening dimension, publish that dimension rather than adding more similarly labeled sleeves to the shortlist.

5. Home improvement: Match the screw holes before discussing style

Shopper question: “I need a replacement handle with holes 128 mm apart.”

Catalog facts needed: The center-to-center mounting distance for each handle variant, not the overall handle length. Once that matches, check the overall dimensions and footprint if the shopper wants to cover an old mark, along with any published fixing specifications and information about included hardware. On listings that offer one design in several lengths, verify the spacing at variant level rather than using the first measurement in the description.

The first reply can resolve a common ambiguity: “Is 128 mm measured from the center of one screw hole to the center of the other?” If the shopper measured from end to end across the old handle, have them measure the hole centers on the door or drawer instead. An example handle that is 128 mm long overall is not a match unless its mounting centers are also 128 mm. A handle with the correct spacing can be longer overall and still line up with the holes, but its ends may sit differently near a trim edge.

Next action: produce a mounting checklist and a filtered set. Show only options with confirmed 128 mm centers for a shopper who wants to reuse the holes. Beside each option, display the overall length and the hardware information the catalog supplies. If covering the old footprint matters, have the shopper compare the base dimensions with the marks left by the existing handle. An unknown footprint should be called out, not presented as a match on every requirement.

Do not include a nearby spacing as an equivalent when the shopper wants to avoid drilling. If no match is available, the useful outcome is a no-match result or a handoff for measurement confirmation. If style or finish matters after that, ask the preference once the fixed mounting constraint is satisfied. This keeps a visually appealing but incompatible handle from floating to the top.

What to measure: Track how often shoppers clarify that they had supplied overall length rather than hole spacing. Audit whether every option shown matches the confirmed spacing, including variants that share a product page. Classify misses as unavailable matches or missing mounting data so the store knows whether to fix stock coverage or product information.

6. Pet supplies: Put a 47 cm chest measurement on the size chart

Shopper question: “My dog measures 47 cm around the chest. Which harness size should I look at?”

Catalog facts needed: The manufacturer’s published chest range for each size of the specific harness and its instructions for where to measure. If its chart also requires neck circumference or another measurement, collect that before calling a size a match. Stock belongs at the end of the check: it tells the shopper which suitable sizes can be ordered, not which size is suitable.

First confirm the shopper measured in the place specified by the chart. Then show where 47 cm lands, including whether it is near the end of a range. In an illustrative chart, one size might cover 40–47 cm and the next 46–54 cm. Both published ranges include 47 cm, but that does not make the two sizes identical choices. Follow the manufacturer’s guidance for overlapping sizes if it gives any. Otherwise, present the overlap and ask for any second measurement the chart uses, or hand off if the catalog still cannot distinguish them.

Next action: return a sizing check the shopper can complete. Put the shopper’s measurement next to the applicable ranges, identify the size or sizes still in consideration, and show the relevant measuring instruction. Use the chart for that exact harness instead of transferring a size from another model or guessing from breed. If the shopper remeasures and gets a different number, recalculate the shortlist rather than treating the first result as fixed.

If the only size supported by the chart is unavailable, say so. The next action could be checking a different model against its own chart, not moving the shopper up or down a size because that variant happens to be in stock. Treat the chart answer as a way to narrow the choice; it is not proof of how the harness will fit when adjusted on the dog.

What to measure: Record boundary and overlap cases separately from clear matches. Check whether shoppers can resolve them using the measurement instructions or need help. If sizing questions persist after purchase, review those conversations alongside the chart version and harness model used, so an answer for one model is not credited or blamed for another.

7. Gifts: Check Friday delivery and the total before shortlisting

Shopper question: “Can I get a gift delivered on Friday for under €50?”

Catalog and order facts needed: The destination postal code and the intended Friday date, stock for each exact variant, the available delivery services and their windows or cutoffs, and the total for the chosen option. Check whether the shopper means the whole order, including delivery and any requested extras, or the gift alone. “Delivered on Friday” may mean arrival that day, not just arrival by then; ask if early delivery would also work when that changes the options.

Check feasibility before asking about interests or suggesting products. A gift that is under the product budget can stop qualifying once delivery is included. A product in stock can also miss Friday if the available service arrives later for that destination. Use the service window that applies to the address and time of the check, and label an estimate as an estimate. If the store cannot verify Friday for the destination, do not turn a dispatch date or a general delivery description into an arrival promise.

Next action: show a shortlist with the total and the delivery window. Once the hard constraints are satisfied, ask a preference question if it will help narrow the choices. If nothing qualifies, say whether the limiting factor is the budget, the date, or both. Offer a different window or a different option for the shopper to consider without changing the original constraints silently.

If the shopper uses voice to select a gift, distinguish that selection from an accepted cart action and from payment. “The second one” identifies the item under discussion. Before adding it, recap the exact variant, quantity, current total and service window, and ask the shopper to accept the cart action. Report the item as added only after the store confirms it. That confirmation is not payment: payment is a separate checkout step, and an item in the cart is not evidence of a paid order or a secured delivery slot.

What to measure: Check whether shortlisted gifts still meet both constraints when the shopper reaches checkout. Log changes in stock, delivery window, and total separately. Test a changed service window after selection and before cart acceptance: the conversation should show the updated terms and let the shopper decide again.

Choose one decision to pilot

Read recent pre-purchase questions and group them by the decision shoppers are trying to make. “Will it fit?” is too broad a group for implementation: wall fit needs overall dimensions and orientation, whereas laptop sleeve fit needs the dimensions of two items and, where documented, an opening check. Start where the same decision comes up repeatedly and the required facts are already present for most of the variants people ask about.

Pick one category and three frequent questions. For each, write down the minimum facts needed, the action that counts as resolving the decision, and what to do if a fact is missing. A good fallback names the missing fact and keeps the shopper’s context. For example, a handoff should carry forward the 128 mm measurement and the shopper’s need to reuse the holes so they do not have to explain the entire job again.

  1. Run a variant-change test. Switch sofa orientation, jacket size, or lipstick shade halfway through a live test conversation. Check that dimensions, preview, swatch, and cart selection, as applicable, now refer to the new variant. Details from the original selection should not survive as if they were still current.
  2. Run a stock-change test. Make the selected variant unavailable after it is shown but before the shopper accepts the cart action. Check that the conversation reports the change, preserves the original constraints, and asks before offering to act on a substitute. It should never report an addition that the store did not confirm.

Measure decision resolution, not just clicks on suggested products. For each eligible question, record whether the shopper received a checked answer and a usable next step, needed a handoff, or left with a fact still missing. Keep cart acceptance, confirmed addition, and payment separate in reporting. A swatch that opens the right shade, a size chart that rules out an option, and a delivery check that prevents a bad shortlist can all resolve the shopper’s question without producing an immediate sale.

Try the EcomCloser demo

Related guides

Sources