VyForge is the platform businesses use to create, manage and optimise promotions, vouchers, loyalty and referral programs. Plug in your stores, apps and POS, then let Claude-powered AI draft campaigns, choose the right offer for each customer, and explain what worked in plain language.
API-firstMulti-tenantBuilt on the Claude API
Win back customers who haven't ordered in 45 days. Keep cost under 8% of basket.
Promotions, vouchers, loyalty and referrals usually live in four different tools that disagree about the customer. VyForge runs them on shared rules, shared budgets and a shared customer record, so a voucher knows about the loyalty tier and a referral reward knows about the budget cap.
Define conditions, priorities and stacking order once. Schedule by channel, store, product set or segment, and put a hard budget cap on every campaign.
Generate single-use or multi-use codes in bulk, hand pools to partners, and validate every redemption in real time at the till and online, with the same answer in both places.
An auditable ledger for earn, burn, expiry and adjustments. Tiers with their own rules, and points that can be spent as a voucher at checkout.
Unique links and codes per customer, two-sided rewards that pay out only after the referred order clears, and checks for self-referral and duplicate devices.
VyForge uses the Claude API where a model actually earns its place: drafting, ranking, explaining and answering. The rule engine and ledger stay deterministic, so every discount is still a decision you can audit.
Pick the offer each customer is most likely to redeem, within the margin you allow.
Describe the goal in a sentence. Get a complete, rule-valid campaign draft to review and launch.
Segments that explain themselves: who is lapsing, who only buys on discount, who would upgrade a tier.
Suggested changes to discount depth, caps and timing, each with the expected effect on cost and uplift.
Answers for "why didn't my code work" with the actual rule that blocked it, for staff and for customers.
Ask questions about campaigns in plain language and get answers computed from the ledger, not guessed.
Which voucher campaign last month had the best uplift net of cost, and why?
Weekend bundle, 2-for-1 coffee had the best net result: +18.4% incremental revenue against a cost of 5.1% of basket. Redemptions clustered on Saturday mornings, and 61% came from customers who had not bought a bundle before.
Flash 20% storewide drove more redemptions but most went to customers who were already buying that week, so net uplift was only +2.3% at nearly three times the cost.
Suggestion: rerun the bundle with a 50,000₫ cap and a Saturday-only schedule. Want me to draft it?
One tenant per business, one API for every channel. Your POS, storefront and app call the same validation endpoint and get the same answer.
Every campaign reports incremental revenue against a holdout, cost per redemption and budget burn, so the question is never "did people use it" but "was it worth it".
| Campaign | Program | Redemptions | Redemption rate | Incremental uplift | Cost / redemption | Budget used |
|---|---|---|---|---|---|---|
| Weekend bundle, 2-for-1 coffee | VOUCHER | 1,862 | 12.9% | +18.4% | 21,400₫ | 64% |
| Come back, we kept your table | PROMOTION | 311 | 12.9% | +9.7% | 38,900₫ | 41% |
| Bring a friend, both get 30k | REFERRAL | 540 | 6.1% | +11.2% | 60,000₫ | 77% |
| Flash 20% storewide | PROMOTION | 3,050 | 24.6% | +2.3% | 57,200₫ | 100% |
Sample data for illustration. Uplift is measured against a randomised holdout group per campaign.
Bring one real campaign you ran last quarter. We will load it, draft the AI version next to it, and walk through what the numbers would look like.