Open Loyalty receipt scanning: verified receipts become points with no review team in the loop
Open Loyalty receipt scanning: extraction, SKU validation, duplicate and fraud checks in Steve, then campaign rules and points in Open Loyalty.
Receipt scanning is one of the most requested earning mechanics in loyalty programs and one of the most often postponed. The reason is rarely the scanning. It is everything after: deciding whether the receipt qualifies, catching the same receipt twice, and doing both fast enough that the member sees points before they leave the store car park. Teams that build this in-house end up with an OCR contract, a rules engine, a duplicate index, a review tool, and a queue of edge cases.
The Steve integration for Open Loyalty exists so that none of that has to be built. Steve was built by the Open Loyalty team, and the integration is native rather than a connector: a verified receipt posts to Open Loyalty as a transaction, and the program’s own earning rules award points.
What happens to a receipt
A member photographs a receipt in the program’s app, on a web page, or in a chat channel. From there the path is fixed.
Quality gate. The photo is scored for sharpness, exposure, and framing at capture time. A blurry or cropped photo is refused immediately with an instruction to retake. It never enters the pipeline and is never billed.
Extraction. Merchant, store, date, time, line items with quantities and prices, total, currency, receipt number. Every field carries a confidence score. No per-retailer templates. The program’s SKUs are matched against the line items.
SKU validation. Steve validates the extracted products against the product catalogue. Recognised SKUs give Open Loyalty consistent product data to use in its campaigns. Matching a product does not decide whether it earns a reward.
Duplicate checks. A perceptual hash of the image against every receipt the workflow has processed catches the same photo again, even cropped or recompressed. An identity fingerprint from merchant, date, time, and total catches the same purchase photographed again. Both run across the full history of the program.
Fraud checks. Steve checks for signs of altered or fabricated evidence. Suspect submissions can be flagged for review or blocked before they reach Open Loyalty.
Verdict. Approved, flagged, or blocked. Approved receipts post to Open Loyalty as transactions immediately. Flagged ones wait in a review queue with the receipt, the fields, and the reason side by side. A reviewer confirms or clears each with one click, and confirmed ones post.
Campaign rules and points. Open Loyalty receives the transaction with its line items and applies the campaign and earning rules: qualifying products, participating retailers, purchase windows, member limits, and how many points to award. Steve verifies the evidence and supplies the purchase data. Open Loyalty decides whether that purchase qualifies for a campaign and what it earns.
Why the split matters
Keeping verification and earning separate is deliberate. The loyalty team owns the campaign and earning rules in Open Loyalty, where they already manage tiers, campaigns, and rewards. Steve handles extraction, duplicate and fraud checks, and SKU validation against the product catalogue. Changing “two points per bottle” to “three points per bottle”, or adding a participating retailer to a campaign, is an Open Loyalty change. Tightening the duplicate checks is a Steve change.
It also means the transaction record in Open Loyalty is a real transaction: merchant, date, line items, total. Segmentation, reporting, and downstream campaigns work on it exactly as they would on a POS-fed transaction.
What the member experiences
Photograph the receipt. If the photo is bad, a message within a second saying what to fix. If the photo is good and the purchase qualifies, points can follow automatically. Steve returns reasons for evidence issues, such as a receipt already submitted. Open Loyalty determines campaign eligibility, such as whether the purchase date is inside the campaign window or a product qualifies for points.
What the program team experiences
A review queue with only the exceptions. Reporting that shows submitted, gated, approved, flagged, and blocked volumes per campaign, alongside the billed count, which excludes anything the quality gate refused. Verification settings stay in Steve; campaign and earning rules stay in Open Loyalty. There is no separate OCR vendor or duplicate database to run.
Setting it up
The Steve workflow is configured around the program: which fields to extract, which product catalogue to validate against, which fraud checks to use, which identity fields define a duplicate, and which quality preset to apply. Campaign and earning rules are configured in Open Loyalty. Submissions flow in through the program’s existing app or through Steve’s web upload pages. Verified transactions flow out to Open Loyalty. The whole thing sits behind the Open Loyalty account the program already has.
Agencies running programs for several brands on Open Loyalty get one Steve engine with per-brand workflows, per-brand duplicate history, and one queue their reviewers can staff.
If you run a program on Open Loyalty and have receipt scanning on the roadmap, book a demo and bring a few real receipts. The integration is live and we can show the full path from photo to points.
