Merit · Product
Working documents shared for review. Each page is republished from its source, so what you see here is the current version rather than a copy taken at some point in the past.
Clickable concept demo of the PIF employee rewards journey on Merit mobile web: SMS invite with auto-filled OTP, the PIF membership with missions and activity, redeeming an offer from a PIF company, buying a digital gift card and merchandise with points, inviting family inside a PIF ceiling, and the PIF rewards admin on desktop. Guided path plays the standard journey; everything is clickable. Illustrative data.
Three sliders for the September 2026 investor presentation, in Fred's order, plus the roadmap as a separate slide. Slider 1: loyalty value trapped behind identity, the one verified Loyalty Passport that answers it, four tiers from Starter to Enterprise with the reason each one buys, the AI core (mTrust behavioural scoring to GA in October, best-method routing at checkout, and the cross-programme data moat), and the differentiator that no competitor holds verified identity, KYC reuse and trust scoring together. Slider 2: markets, the B2C persona, the five issuer types in the B2B ICP, the partners already connected, and the live pipeline. Slider 3: the 2027 revenue model by driver, with revenue YTD and planned 2026 left as Finance gaps rather than estimated.
Three sliders, with the roadmap inside slider 3. Slider 1 frames the product as one loop rather than a wallet: earning is locked to one brand so balances stay small and expire, and spending needs the issuer's own store at the issuer's own rate. Merit opens both halves, and the AI core carries the checkout agent that finds value across programmes plus the coach that surfaces one thing to do on the earn side. Slider 2 is the four Year-1 markets with their GMV split, the personas, the priority issuers per market and the three growth paths. Slider 3 stays on the Year-1 target rather than a monthly run-rate, states plainly that there is no public user base yet, and carries the roadmap.
The product Merit sells, not a migration. A white-label, multi-tenant, loyalty-native commerce engine: a bank, telco or airline launches its own branded marketplace on Merit's supply chain and loyalty infrastructure without building a commerce stack. Slider 1 puts the pain on the buyer's side of the table, one row per person who feels it, from the head of loyalty who cannot run logistics to the member forced to pay all points or all cash. Then the four tiers with their subscription and both take rates, the matrix against generic commerce and loyalty engines, and the AI core led by Arabic search as a launch-critical capability. Slider 2 is the clients already trading and the named targets per market. Slider 3 is the tenant ramp to $869.2K a month, what that annualises to, and the roadmap.
How the products actually connect, drawn rather than listed. The four-sided network from the approved Merit ID artifact, with identity sources and issuers feeding in on the left, Merit surfaces and merchants drawing out on the right, and Merit ID in the middle holding none of the value itself. Then the five layers with an owner against each one, including the value graph that the issuer keeps, which is the reason an issuer joins at all. Closes on the five step loop and why it compounds: three issuers make three connections, five make ten, eight make twenty-eight, so settlement scales with the connections between customers rather than with the count of them.
Three demos Fred can run without a product person present, four minutes each, with the exact line to say at every step. Demo 1 leads with the live Al Fursan marketplace, where the miles and cash slider is in production today under MP-7605 across around 1,500 SKUs. Demo 2 is the B2C Super App on Android, with the pre-flight checklist and the recording fallback. Demo 3 is the E-Commerce platform, and it names the constraint plainly: it has to run on a prepared environment with a seeded tenant and cannot be created live in the room. Ends with the seven open items and their owners, three of which are Finance.
Everything the Al Fursan migration team needs on STC, per 11 September 2026. Why STC is a gap to build rather than data to migrate, the four properties of STC's stock model that cause every complication downstream, the six API calls with their required fields, the single delivery address shape and the six problems it deletes, the nine step customer journey with the two deliberately manual steps, the verified Jira state of all eight tickets, the rewritten checkout check with its fail open rule, the four things still blocking (three of them STC's), the six items the new Seller Portal has to build and has never been sized, and the six specification documents that exist only as repo markdown.
The whole solution on one page, per 11 September 2026. The map first, read downward and drawn box to box across eleven numbered seams: the green band is what we own with Core on its own row as the single front door, the purple band is Akshay's Platform services that we call, the amber band is MGC and the Online Catalogue from the old world, and the vendors sit outside. Then the five rules that decide where a thing lives (PIM owns the product and Core owns the offer, Core is the only front door, only OMS moves an order once it exists, everything after the order is the Fulfillment Service, and we connect to a rule engine rather than keep a copy of it). One row per service saying what it owns, who it calls, who calls it, and its work-tree node. Three flows traced end to end: an order, the catalog and stock bridge, and the money. Closes with what we depend on and do not own, and what on this page goes stale first.
Two pages of text and five diagrams, written for Muneeb Meer before the technical deep dive of 12 September, so the twenty-page PRD is not the first thing he reads. What the work actually is (storage behind a LiveOps screen that already exists), the three phases and why only Phase 0 blocks anyone, who sets which commercial number and at what level, why Ecommerce Core blocks the Rewards Portal rather than the reverse, the two VAT streams the ZATCA invoice has to print, and the six open questions. Names the one contradiction in the record: the grain was settled per product on 8 September and recorded as per variant on 7 September.
v1.1, corrected after Julie Barbier's note of 9 September. Seventeen slides for the internal team, plain language, no technical detail on the slide faces. Merit ID members are Merit's own and Merit owns their data, so screening them is a product decision with no counterparty. Partner programme members are the separate case that is a partnership question. Why sellers already work, the three routes with the owned route first, and the two things the business is being asked for. Arrows or space to advance, Esc for all slides, F for fullscreen, N for speaker notes, T for light or dark.
v1.1, corrected after Julie Barbier's note of 9 September. Merit ID is Merit's own product and Merit owns the data in it, so the blocker there is scope of collection, not data rights. Partner programme members are the separate case where data rights do bind. Which LSEG World-Check product fits, the three routes to the data with the owned route first, and where the work belongs on the roadmap. Answers angle 1 and angle 2 of Julie's LEAP note. Section 7 maps the five counterparty touch points and the two that are unguarded.
Clickable walkthrough of the promo code MVP agreed on the 2 September reset call. Simulates a real session: the merchant onboards through the invite link, pins the outlet on Google Maps and creates a promo code, then the customer picks the offer, reads the code to the cashier and confirms it. Every action plays and stops on its own. Since 14 September it also carries the SSO handoff, with SAIB's outstanding half stated on the screen, and a full Arabic right to left build behind the toggle in the header. Space plays, arrows step, shift with arrows changes screen.
The customer side only, built for SAIB to walk SAMA through it. Starts at sign in to the SAIB app, then opens Aseel offers already signed in, finds an in-store offer, reads the terms, slides to claim, shows the code at the counter and confirms it was used. No merchant or technical steps. Full Arabic right to left build behind the toggle. Space plays, arrows step, shift with arrows changes screen.
A working prototype of the Merit ID app, five journeys each shown end to end: earn with a survey, get a discount, get cashback, use a member benefit, and buy online. Built to be screen recorded for the VISA demo video. Press 1 to 5 to pick a flow, M for the menu, H hides the presenter bar.
Three decks in one presenter: the E-Commerce Solution, the B2C Super App and Merit ID. Each opens with the pitch, then the working detail. Arrow keys or the edges to move, Index for all 50 slides, n for speaker notes.
Diagram-first engineering onboarding map: the twelve seams between the Marketplace and Platform stacks, six process diagrams, the Old vs New feature parity matrix, and the decisions it forces. For engineers joining Merit commerce.
Why we cannot list the build yet, the three discovery lenses, the phase plan with dates, and a dated go/no-go on the October cutover. For Fred, Aditya, Akshay, Muneeb, Marthino and Ops.
How a digital gift card reaches the catalogue, how a discount is set for one product and one tenant, and what happens to the money between the order and the settlement report. Two swimlanes, the setup flow, twelve gaps and eight open questions. For Marthino and Habeeb.
What PIM owns, the category-to-variant data model, how a product gets in, the collision rule, and every PIM feature with its rules, services and spec.
The order state service: what it owns, the twelve states from the code, item roll-up, Seller Portal mapping, cancel and refund, where PRD and code differ, and every feature.
The client store and the Storefront Admin Portal: three surfaces, how a product reaches the store, checkout, roles, and every portal and store feature.
The commerce backbone: Old World vs New World, Product vs Offer, Buy Box, order lifecycle.
The consumer side: the points wallet, earn and burn, paying with points, the four markets.
The supply side: how sellers list, price and fulfil, screen by screen for each user.
Is this uploaded product already in the catalogue? Why a high score is not enough, shadow mode, and the region trap.
Market as a first-class entity, why Tier 1 needs no multi-currency at all, and the six phases to size.
What happens when STC has no stock in the buyer city, the scenario simulator, and five ways to fix the rejection experience.
The concept behind MSP-392 and DEC-0172, in diagrams. What makes an attribute an axis, how axis values multiply into SKUs, the cap of five per category, who may create one, and the iPad WiFi case shown before and after.
The requirement behind MSP-392. What makes a Variant distinct, the four paths through Product Upload, and what happens when two real items collide on the same axis values. Requirements B, C, D and G with acceptance criteria.
One PDP for the UAE, International and US phone. The live version selector, why the axis route beats display grouping, and the five things engineering has to confirm.
Answers to the technical and commercial clarification list, 14 points.
The commercial proposal, five open business decisions and the phase map.
Who scans what, and what changed on 5 August 2026. The customer scans the outlet QR on their own phone and the store scans nothing. Built for the SAIB compliance re-check.
The recommended in-store journey, screen by screen on the shipped Aseel design. STALE: still shows the cashier handover removed on 5 Aug 2026.
Why SAIB is accepting a downgrade, and the six ways back. QRIS and Grab Dine Out as the mental models, five journeys animated end to end, and the two questions that decide it.
How a rate change, an order and a redemption differ, and which phase owns each step.
The five phases, what each delivers, and where the accounting integration attaches.
The week point by point, then the whole portfolio as one explorable tree, with what moved in the reporting week.
Real screenshots and a walkthrough of the actual Linear product, for the JIRA/Linear decision.
The identity layer under everything: why loyalty fails, the architecture rule, the issuer network, the checkout agent.
How an order becomes a parcel: the routing rule, what is built on the board, and where that sits against the Q3 plan.
The two decks for the Amman on-site: B2C Super App in six slides, then the E-Commerce Solution in fifteen. That deck maps the six things inside the E-Commerce Solution, then gives each one a slide: Seller Portal, Storefront, Storefront Builder, E-Commerce Core and PIM, Fulfilment Service, OMS. Diagrams, the tenant ramp, and real product journeys screen by screen.
The same two Amman decks, one slide on screen at a time. Arrow keys or the edges to move, Index for the grid of all 23 surfaces, n for speaker notes, f for fullscreen. Use this one in the room; the review page above is for reading.
What happens to an order that has not finished by day 30: the release condition that cannot be evaluated without a delivery event, the five proposed rules, and what Partnership and Operations each have to do before the 21 August sign-off.
Fayez's 16 tests all pass, and twelve further probes found five defects. The three blockers are one problem: a code matched by nothing but itself, no rate limit on guesses, and a 65,536 code space. With the attack arithmetic and what has to be decided before Tintash starts.
The store has one door today, the rotating "Shop with Miles" card in the Exclusive offers carousel on Home, so it disappears when the campaign does. Two placements, drawn before and after over the live Saudia app screens: a "Shop with Miles" row second in the AlFursan tab Miles list, and an "AlFursan Store" row with the external marker in the side menu after Miles Calculator. Both are one list row. No fifth bottom tab is being asked for.
How Merit's fulfillment works today through OTO, what phase 1 of the proposed Omniful integration actually moves (Merit's seller and product data outward, one way, no return channel), how that inverts the synergy the two companies aligned on back in August, and the gap in a possible phase 2: Omniful has no product-upload path for its own sellers today.
The plan for the next PIF demo: Merit as the engine PIF runs its programmes on, scene by scene in presenting order, what each scene shows and why, today's demo next to the new one, and what is real today versus labelled "coming" in the prototype.
The full PIF session in one file, scenes 0 to 7: opening slides, the clickable demo of Super Platform, LMS programme setup, Ceer onboarding in the Seller Portal, the employee app, recognition slides, revenue through Al Fursan and the Seller Portal, then phases and close. Arabic and English.