YundaDeskWhen a customer asks, “Can you match this price?”, they are not always just bargaining. Often they already like the product and need one more reason to buy. If an agent drops the price too quickly, margin gets chipped away; if the agent replies with a blunt “no”, the order may disappear.
The job of a price match script is to separate three things: what you can explain, what you can save, and what you must decline. The operating rule matters even more: price changes are high-risk actions that require human approval. AI can explain the policy and collect evidence, but it should never discount on its own.
Diagnose first: price check, negotiation, or complaint
The same line, “I found it cheaper elsewhere,” can mean different things. The script should follow the intent.
| Customer intent | Common wording | Agent goal |
|---|---|---|
| Normal price check | “I found a lower price elsewhere.” | Explain the value gap and try to keep the order at current price |
| Discount testing | “Can you do better?” | Offer controlled options without changing the price directly |
| Post-purchase frustration | “I just bought it and now it is cheaper.” | Check order and policy, then escalate if needed |
| Churn threat | “Match it or I leave.” | Stay polite while holding the boundary |
Do not start with “I can request the lowest price for you.” First acknowledge the concern, then collect the facts: link, screenshot, currency, shipping cost, ship-from country, and whether the product is the same model. Many lower prices stop being lower once shipping, duties, delivery time, and return coverage are included.
Price-Match Request Scripts That Protect Margin Without Losing: split handling permissions by risk first
Ready-to-use price match scripts
Use these as support macros and adjust by market.
When the customer sends a competitor link:
I understand you want to make sure you are getting the right price. We can help review the other page, but we need to compare the product model, specification, ship-from location, shipping cost, after-sales terms, and stock status. If it matches under the same conditions, we will submit it for human approval under our store policy. Before approval, support cannot promise a price change.
When your price is higher but your service is stronger:
The price you found is lower on the product page. Our offer also includes official after-sales support, trackable shipping, and help after purchase. Some lower-price listings do not include the same delivery or service terms, so it is worth comparing total cost rather than the item price alone.
When the customer is simply asking for a better deal:
I understand you are looking for the best available price. This item is already priced under the current promotion, and support does not have direct price-change permission. I can help check whether any coupon, bundle, or eligible offer is available without affecting your after-sales coverage.
The logic is simple: do not dismiss the customer, attack competitors, or make discounting the only exit.
Save the sale without making price the only lever
The most common mistake is treating retention as automatic discounting. Many shoppers are really asking, “Will I regret buying now?”
You can offer alternatives that protect margin:
- Explain total cost: item price, shipping, duties, and delivery time together
- Emphasize certainty: stock, dispatch time, after-sales response, trackable shipping
- Use non-price incentives: coupon, gift, bundle, or next-order voucher
- Give the customer control: if they prefer waiting for a sale, offer a reminder path
When you must decline: soften the wording, keep the line clear
Not every price match is worth accepting. Some links are not the same product, some prices are abnormal, and some requests push you below cost. Decline when needed, but do it without sounding cold.
Different product or different terms:
We reviewed the link you shared. Because the specification, ship-from location, or after-sales terms are not the same as this item, we cannot adjust the order to that price. We can still help check any current eligible offers to make the purchase more favorable.
Below your policy range:
This price is outside the range we are allowed to approve, so we cannot match it directly. To protect product quality, shipping reliability, and after-sales service, we do not adjust orders based on unstable or abnormal prices.
Customer keeps pushing:
I understand the price difference makes the decision harder. What we can do is review the request under the published policy and submit it for approval when it qualifies. We cannot bypass approval or change the price directly. You can keep the current order, or wait for a future promotion if that works better for you.
Avoid three weak lines: “the system does not allow it”, “my manager said no”, and “that other seller is unreliable”. Stronger language uses policy, matching conditions, and approval limits.
Price changes are high-risk: AI explains, humans approve
The risky part of price-match handling is permission. Price changes, compensation, and refunds affect margin and billing, so they are high-risk actions. YundaDesk’s rule is straightforward: AI answers first, humans back up; money-impacting decisions need human approval.
AI support can:
- Detect whether the customer is price checking or complaining after purchase
- Ask for the competitor link, screenshot, and order number
- Explain the price-match policy and approval flow
- Summarize the request before handing off to a human
AI should not:
- Promise “we will definitely match it”
- Change the order price automatically
- Issue discounts beyond policy
- Invent an exception just to calm the customer down
If you want AI to handle more repetitive conversations, write the boundary first. See the AI-first, human-backed boundary: low-risk questions can be automated, high-risk actions go to humans, and every decision remains traceable.
Turn scripts into a knowledge base, not scattered chat history
A useful price match script should become knowledge and rules, not tribal knowledge in one senior agent’s head.
Create a knowledge base section for “pricing and promotion policy” that covers at least:
- Conditions for eligible price matching: same item, same specification, same currency or conversion rule, same delivery terms
- Exclusions: clearance prices, member-only prices, used or refurbished items, region-limited offers, suspicious screenshots
- Retention options agents can use: coupons, gifts, bundles, next-order vouchers
- Actions that require approval: price change, refund, compensation, oversized discount
- Standard decline language: prepared by market language, not only in English
When AI misses an answer, an agent fills the gap, or an agent corrects the AI, YundaDesk creates learning suggestions you confirm. Only owner-approved suggestions become skills, knowledge, or customer memory. Every item is traceable, testable, and revertible. That is what gets smarter over time means; it is not a system quietly learning to discount.
Review metrics: do not only count saved orders
Reviewing price-match handling should not stop at “orders saved”. The better question is whether margin and boundaries held.
- How many price-match requests converted at the original price?
- How many converted through non-price incentives?
- How many went to human approval, and what was the approval rate?
- Which products get price-compared most often? Do their product pages need clearer value explanation?
- Which agents made promises beyond policy? Do scripts or permissions need tightening?
If the same product keeps attracting price-match requests, it may not be a support-script problem. The product page may need clearer material, warranty, shipping, or bundle value. Support can recover part of the sale, but merchandising needs to help.
A price-match customer is not a bad customer. If they ask, they are still giving you a chance to win the sale. Support’s job is to catch that chance, keep discounts inside approval, and turn each conversation into reusable capability.