Pet-care loyalty across retailers: the receipt evidence workflow

Pet-care loyalty has an attractive pattern: food, treats, litter, and supplements are bought repeatedly, but often across supermarkets, pet specialists, marketplaces, and local stores. A brand that does not receive every retailer’s transaction feed can ask members for receipts. The hard part is turning inconsistent retail text into reliable purchase evidence without making the member type every product by hand.

Steve and Open Loyalty can divide that job in a scoped pilot. Steve extracts configured receipt fields and applies configured evidence checks. Open Loyalty runs the member program: campaigns, product earning rules, challenges, points, badges, and limits. Product matching across pet-care variants must be validated on the brand’s catalogue and representative receipts before the combined workflow is presented as deployed.

Why pet-care receipts are difficult

Retail descriptions are compact and retailer-specific. A receipt might say PUR DOG SEN 12K, CAT LTTR CLMP 8LB, or only a shortened brand and price. The loyalty catalogue may distinguish adult and senior formulas, dog size, flavor, pack weight, and multipack quantity.

Those distinctions matter. A “buy any food” campaign may tolerate a broad category match. A veterinary-diet challenge may need an exact product family and size. The extraction workflow should return what the evidence supports and route uncertain catalogue matches to review rather than silently choosing the nearest product.

Synthetic examples:

Receipt line	Possible interpretation	Safe treatment
PAWPLUS ADLT CHK 3KG	Adult chicken food, 3 kg	Match if brand catalogue has one unambiguous item
PP CAT PCH 12	12-pouch cat multipack	Review if several flavors share the abbreviation
CARE SUPP JOINT	Joint supplement	Category match may be enough for a broad campaign
PET ITEM 19.99	Unknown pet product	Keep unmatched; do not invent a SKU

These examples are illustrative. A real pilot should use the brand’s product catalogue and receipts from its priority retailers.

Build a product catalogue for decisions

A useful catalogue contains the identifiers the campaign needs: brand, product family, variant, size, unit count, category, and any retailer-specific aliases the team has confirmed. It does not need to recreate every global product database.

Start with the smallest set that supports the first campaign. During pilot evaluation, label sample receipt lines as exact match, family match, ambiguous, or no match, then compare Steve’s configured output with those expected labels. Do not assume these categories or arbitrary catalogue candidates are native output fields until the workflow schema and checks have been implemented and tested. Approved product facts can then drive the appropriate Open Loyalty rule.

Do not let the loyalty campaign compensate for weak evidence by guessing. Campaign rules should operate on approved facts.

Model repeat purchase in Open Loyalty

Steve verifies each purchase independently. Open Loyalty owns cumulative behavior across the member’s history. That is where a program can express mechanics such as:

• earn one point per currency unit on qualifying pet food;
• complete a challenge after three purchases in eight weeks;
• give a bonus for buying food and treats in one basket;
• reward a first purchase in a new category;
• cap a promotional bonus per member.

The receipt becomes a normal transaction with line items. Open Loyalty can evaluate it alongside other member activity. Steve should not decide that a member has completed a three-purchase challenge; it supplies the approved transactions that make the challenge measurable.

Handle cross-retailer differences explicitly

Collect test receipts from supermarkets, pet specialists, e-commerce order receipts, and independent stores. For each channel, check merchant recognition, dates, returns, discounts, quantity notation, and how online orders show fulfillment or cancellation.

A retailer list can be a campaign rule, but retailer identity first has to be extracted with enough confidence. If two banners share a legal entity or online marketplace receipts name a payment processor, define what the program accepts. These are policy choices, not OCR bugs.

Returns need their own rule. A negative line or return receipt should not create positive earning. If the available evidence cannot link a later return to an earlier purchase, document that limitation and decide how operations will handle it.

Plan review around ambiguity

Route cases to review when a configured confidence or attention rule fires, a qualifying section is cropped, a duplicate is flagged, or required fields are missing. The current review queue can show the receipt, extracted fields, confidence, rule reason, and matched duplicate. A candidate-product comparison view is an additional requirement unless the pilot confirms it is represented in the configured output.

Measure review by cause. “8% reviewed” is less useful than knowing that most reviews came from one retailer’s abbreviation or one new package size. That finding can improve the catalogue and reduce future work.

A practical pilot

Choose one country, two product families, and three to five retailers. Gather representative receipts, including poor but usable examples, discounts, returns, multipacks, and online orders. Agree an expected result for every sample. Then test:

1. extraction of merchant, date, total, and lines;
2. product matching at the required granularity;
3. duplicate handling;
4. reviewer decisions and any separately scoped correction path;
5. member matching;
6. transaction delivery;
7. Open Loyalty earning and challenge progress.

The pilot should report field accuracy, product-match outcomes, review reasons, delivery failures, and campaign results separately. Do not turn a small test into a universal accuracy claim.

Cross-retailer loyalty is valuable because it recognizes behavior the brand cannot see in one POS feed. It works when uncertainty is visible and the campaign only acts on approved evidence.

See receipt verification, the proposed pet-care program page, and the Open Loyalty integration. If you are planning a pet-care pilot, book a demo and bring a product catalogue plus difficult receipts from the retailers that matter most.