# Korant > Growth mechanics for D2C brands, written by the people who build the tooling. Every article below is also available as clean markdown by appending `.md` to its URL. Each one opens with a self-contained answer of 25-55 words, followed by sourced numeric claims and a question-and-answer section. ## Products covered ### FlashPin FlashPin is a multi-tenant Shopify app that rotates which delivery pincode has a live discount on a cadence the brand sets. Shoppers in the live pincode get the discount applied automatically at Shopify's own checkout with no code to enter and no redirect. Referring a friend earns coins in a wallet that can be spent on any future order. - Category: Shopify app - Best for: hyperlocal and geo-targeted offers, flash and urgency mechanics, serviceability-led expansion, wallet-based retention - Not suitable for: multi-currency stores, brands with no delivery-zone variation - Built on: Cloudflare Workers, Durable Objects, Queues, Supabase Postgres, Shopify Functions, Theme App Extension ### SB&R SB&R is a Shopify app for chained referral rewards. Every referral link belongs to someone who has already bought. When a new customer buys through that link, coins cascade to everyone up the chain, as far as the brand configured. Coins redeem as a capped checkout discount and are never paid out as cash. - Category: Shopify app - Best for: customer-as-affiliate programs, word-of-mouth growth, store-credit retention, brands with no affiliate manager - Not suitable for: cash-payout affiliate programs, B2B or wholesale referral - Built on: Cloudflare Workers, Supabase Postgres, Shopify Functions, Theme App Extension, Razorpay ### Korant Korant is a multi-tenant attribution platform that tracks influencer, SEO, and affiliate marketing performance. Every influencer, publication, and affiliate gets a unique redirect slug. Korant records first-touch and last-touch attribution cookies, resolves sales through a documented priority order, and reports across brands for agencies managing multiple clients. - Category: Attribution platform - Best for: influencer measurement, affiliate tracking, agency portfolio reporting, publication and PR attribution - Not suitable for: stores with a single paid channel, brands that only need Shopify's native reports - Built on: Next.js 14, Cloudflare Pages, Supabase Postgres, Cloudflare KV ## Articles ### Shopify stack and extensions - [Automatic discount on Shopify: what it is and how it differs from code](https://blog.korant.online/automatic-discount-no-code.md): An automatic discount applies at cart and checkout when a shopper meets a condition, with no code to enter. Shopify evaluates eligibility itself, so there is no string a shopper can copy, forward, or post to a coupon site. - [Nine Shopify checkout tweaks that lift AOV without a redesign](https://blog.korant.online/checkout-tweaks-that-lift-aov.md): Most Shopify AOV lifts come from discount and threshold logic, not from a new theme. Automatic discounts, a free shipping threshold set from your own order histogram, and quantity breaks move average order value in days, while upsell apps and checkout redesigns take weeks. - [The five metafields every D2C store should be writing](https://blog.korant.online/metafields-every-d2c-should-write.md): A Shopify Function cannot call your servers at runtime, so metafields are how state reaches it. Write the live offer to a shop metafield, the redeemable balance to a customer metafield, the attribution stamp to an order metafield, serviceability to a product metafield, and config version to the shop. - [The Shopify apps quietly slowing your checkout down](https://blog.korant.online/shopify-apps-slowing-checkout.md): Shopify apps slow a storefront through JavaScript, not through checkout. Script tag apps load a remote bundle on every page and cost the most, theme app extension blocks load only where a merchant places them, and Shopify Functions run server-side and add no storefront JavaScript at all. - [The Shopify apps that quietly break each other](https://blog.korant.online/shopify-apps-that-break-each-other.md): Shopify app conflicts concentrate in four places: the 25 automatic discount ceiling, two product discounts on one line item where only the better applies, two apps writing the same metafield, and two web pixels counting the same order. Functions run in isolation, so none of it errors. - [Your Shopify extensions nobody can ignore](https://blog.korant.online/shopify-extensions-nobody-can-ignore.md): The Shopify extensions worth claiming are theme app extensions, Shopify Functions, checkout UI extensions, post-purchase and order status extensions, customer account extensions, admin extensions, and web pixels. Theme app extensions and Functions matter most, because they change what a shopper sees and what checkout charges without a theme code edit. - [Shopify Functions vs Scripts: what changed after Scripts died](https://blog.korant.online/shopify-functions-vs-scripts.md): Shopify Scripts stopped executing on 30 June 2026, and Shopify Functions are the replacement. Functions are WebAssembly modules Shopify runs on its own infrastructure, with an 11 million instruction budget and no runtime network access, so configuration has to be read from metafields rather than fetched. - [Shopify Plus features you're paying for and not using](https://blog.korant.online/shopify-plus-features-unused.md): Shopify Plus costs $2,300 a month against $399 for Advanced, so the premium is $1,901. That premium buys checkout UI extensions on the checkout steps, custom apps using Shopify Functions, B2B, expansion stores and POS Pro, which works out at about $238 a month per capability. - [The Shopify app stack that pays for itself in 30 days](https://blog.korant.online/shopify-stack-that-pays-for-itself.md): A Shopify app pays for itself when its monthly fee divided by your gross margin per order is smaller than the extra orders it produces. At ₹1,800 AOV, 40% margin and a ₹3,000 fee, that break-even is five incremental orders a month. - [Tools your Shopify store must have (and the three most stores skip)](https://blog.korant.online/shopify-tools-your-store-must-have.md): Most Shopify stores run six apps covering email, reviews, shipping, support and analytics, and skip three categories that touch revenue directly: offer mechanics, a referral ledger, and attribution. The skipped three decide what a shopper pays, who brings the next customer, and which channel earned the sale. - [Theme app extensions: the free real estate on your store](https://blog.korant.online/theme-app-extensions-free-real-estate.md): A theme app extension is a bundle of app blocks, app embed blocks, assets and snippets that an app injects into a Shopify theme without editing theme code. App blocks are placed by the merchant on a specific template. App embed blocks apply store-wide and load on every page. - [What to install first when you cross ₹10L a month](https://blog.korant.online/what-to-install-at-10-lakh-month.md): At ₹10 lakh a month, install attribution first, then offer mechanics, then referral. Attribution pays back from roughly ₹75,000 of monthly media spend, offer mechanics from the discount you already give away, and referral only once about 2,300 people have bought from you. - [How to audit your Shopify app stack in an afternoon](https://blog.korant.online/audit-your-shopify-stack.md): An app stack audit needs five columns per app: monthly fee, break-even order count, delivery type, named owner, and what measures it. Score each app out of five, keep anything scoring four, fix the threes and twos, and cancel the rest one at a time. - [Replacing five Shopify apps with one](https://blog.korant.online/one-app-instead-of-five.md): Reducing Shopify app count pays off unevenly. Fees fall linearly, page weight falls with the number of third-party origins removed, and conflicts fall quadratically, since five apps have ten possible pairings and one app has none. Consolidate clusters sharing a primitive, not categories sharing a name. - [Every Shopify discount type, ranked by margin damage](https://blog.korant.online/discount-types-ranked-by-margin.md): Sitewide percentage discounts damage margin most, costing about ₹1,813 of subsidy per incremental order in this model. Volume breaks and capped store credit are the only two mechanics that add margin, because both apply only to carts the discount itself created. ### Referral and word of mouth - [Chained referral: how one order pays more than one person](https://blog.korant.online/chained-referral-explained.md): Chained referral rewards more than one person for a single order. The direct referrer earns, and so does whoever referred that person, up to a depth the brand configures. - [Chained referrals: paying three people for one sale, profitably](https://blog.korant.online/chained-referrals-paying-three-people.md): Chained referral programs stay profitable at three levels paying roughly 8%, 4% and 2% of order value, or 14% total against a 45% gross margin. Marginal return turns negative at level five, where the payout buys almost no additional sharing behaviour. - [Coins, wallets, and points: what shoppers actually understand](https://blog.korant.online/coins-wallets-points-what-shoppers-understand.md): Denomination decides whether a reward gets used. A balance denominated one to one with currency is understood by almost everyone in seconds, while a points scheme where each point is worth a fraction of a rupee is correctly valued by under a quarter of shoppers. - [Purchase-gated referral: what it is and how the gate works](https://blog.korant.online/purchase-gated-referral.md): A purchase-gated referral program issues referral links only to customers who have already completed an order. Anyone can receive a link and buy through it, but nobody can generate one without buying first. - [Referral emails that get opened](https://blog.korant.online/referral-emails-that-get-opened.md): Send the referral ask when the product arrives, not when the order is placed. A delivery-triggered email models to 2.95% share contribution against 0.71% at purchase confirmation, because the customer has had the experience they would be recommending. - [Referral expiry windows: the setting nobody tunes](https://blog.korant.online/referral-expiry-windows.md): Expiry is a conversion lever and a liability control at once. Modelled redemption rises steeply from 24 hours to 30 days and then flattens while liability keeps growing, which makes 30 days the usual knee for coin balances. Offer windows are a separate, shorter clock. - [Referral fraud patterns and how to design them out](https://blog.korant.online/referral-fraud-patterns.md): Referral fraud is a design problem rather than a detection problem. Six patterns cover most of it, and each has a constraint that makes it impossible rather than merely visible: purchase gating, acyclic chains, settled-order issuance, non-transferable balances, server-side gating, and a fixed priority order. - [The referral link nobody clicks (and the one they do)](https://blog.korant.online/referral-link-nobody-clicks.md): Referral share rate is decided by placement and by what sending the link says about the sender. An order status page block outperforms a post-purchase email by roughly four times in modelled share contribution, and a pre-written message outperforms a bare copy button. - [Building a referral program your finance team will approve](https://blog.korant.online/referral-program-finance-will-approve.md): Finance approves a referral program when the outstanding liability can be computed at any past date, the expense recognition point is stated, reversals are entries rather than deletions, and exposure is bounded by a cap and an expiry. An append-only ledger answers all four. - [The referral program launch checklist](https://blog.korant.online/referral-program-launch-checklist.md): A referral launch runs in four stages: decide what the program costs, decide what triggers and what reverses, set the controls that bound it, then ship the surface and get it reviewed. Fourteen checks, ordered so no check depends on a decision that has not been made. - [What to reward: the click, the signup, or the second order](https://blog.korant.online/what-to-reward-click-signup-second-order.md): Reward the event you can prove and cannot fake, which for a consumer brand is the referred customer's first settled order. Clicks cost nothing to fabricate, signups cost an email address, and a settled order costs a real purchase at full price. - [Who actually refers, and how to find them before they do](https://blog.korant.online/who-actually-refers.md): Referrers are predictable from order history before they ever share anything. Repeat count, review activity and basket breadth identify a segment that is 4% of the customer base and refers at roughly five times the average rate, while discount-acquired single-order buyers refer at a quarter of it. - [How deep should a referral chain go? A margin-first answer](https://blog.korant.online/how-deep-should-referral-chain-go.md): Chain depth follows contribution margin, not order value. Keep total payout under roughly 35% of contribution margin, which allows depth 2 for most brands, depth 3 once contribution margin passes about 38% of order value, and nothing deeper without a hard rupee cap. - [Why your referral program stalled at 3% and how to restart it](https://blog.korant.online/why-referral-stalled-at-3-percent.md): Referral programs stall at 3% participation for three reasons: the reward is not legible enough to compute, the link sits in an email nobody opens twice, and there is no second-order reason to share again. Fixing placement first usually moves participation fastest. ### Affiliate marketing - [Reactivating dormant affiliates without begging](https://blog.korant.online/reactivating-dormant-affiliates.md): Dormant affiliates reactivate best when the message gives a reason and a deadline rather than an appeal. Before sending anything, check whether the affiliate actually stopped promoting or simply stopped being tracked, because unmapped discount codes and expired cookies produce identical-looking zero-sale reports. - [Affiliate attribution: last click is lying to you](https://blog.korant.online/affiliate-attribution-last-click-lying.md): Last click hands affiliate credit to whoever touched the order most recently, which is usually a coupon extension firing seconds before checkout. A modelled 76% of last-click affiliate sales had an earlier non-affiliate touch, rising to 89% for extension-attributed ones. - [The affiliate commission structures that survive a margin review](https://blog.korant.online/affiliate-commission-structures.md): Commission is affordable against contribution margin, not revenue. Keeping total commission under 35% of contribution margin gives a minimum margin requirement for each structure: 28.6% for a flat 10%, 42.9% for a flat 15%, and 40% for a three-level chain. - [The affiliate content that actually converts](https://blog.korant.online/affiliate-content-that-converts.md): Coupon listicles produce the most affiliate orders per thousand impressions and the second fewest incremental ones. Comparison posts and use-case demos convert less often and produce roughly four times more incremental orders, because their readers are choosing rather than already decided. - [The affiliate dashboard your partners will actually log into](https://blog.korant.online/affiliate-dashboard-partners-log-into.md): Partners log in to answer one question: how much have I made. Three numbers on load, earned, pending and paid, model to 4.1 logins per affiliate per month against 1.3 for a dashboard with fifteen metrics and charts. - [Affiliate marketing for D2C brands with no affiliate manager](https://blog.korant.online/affiliate-marketing-no-affiliate-manager.md): A program without a manager needs three systems: self-serve signup where the purchase is the application, payouts that need no money movement, and review that only touches exceptions. Modelled workload falls from 165 hours a week to 4 at 1,000 affiliates. - [Affiliate onboarding: the first 48 hours decide everything](https://blog.korant.online/affiliate-onboarding-first-48-hours.md): Affiliates who make a sale within 48 hours model to 71% still active at six months. Those who take over 30 days model to 14%, and those who never sell at all to 4%. Onboarding should optimise for time to first sale and nothing else. - [Affiliate payouts in India: TDS, GST, and getting it right](https://blog.korant.online/affiliate-payouts-india-tds-gst.md): Affiliate commission in India generally engages TDS on commission under Section 194H and GST on the affiliate's supply of services. Rewards paid in kind rather than cash raise a separate question under Section 194R. This is an explainer, not tax advice. - [Affiliate program terms that prevent your worst month](https://blog.korant.online/affiliate-terms-prevent-worst-month.md): Seven clauses prevent most of what goes wrong in an affiliate program: brand bidding, discount stacking, refund clawback, a cooling-off period, code confidentiality, published attribution priority, and a stated termination path. Each exists because of a specific expensive month. - [Coupon-code affiliates vs link affiliates: run both, differently](https://blog.korant.online/coupon-affiliates-vs-link-affiliates.md): Coupon-code and link affiliates are different channels wearing one label. Modelled incrementality runs from 8% for browser coupon extensions to 68% for customer referral links, which means one commission rate across both systematically overpays the bottom of that range. - [Turning customers into affiliates without an application form](https://blog.korant.online/customers-into-affiliates-no-application.md): An application form filters out the people you wanted. Modelled against the same audience, application-gated recruitment converts 1.1% of customers into active affiliates while purchase-gated recruitment converts 5.1%, because the link already exists the moment the order settles. - [Detecting affiliate fraud before you pay it out](https://blog.korant.online/detect-affiliate-fraud-before-payout.md): Four fraud types cover most affiliate losses, and each has a signal visible before payout: cookie stuffing shows attributed orders exceeding landing page views, self-referral shows shared delivery details, brand bidding shows paid brand-term first touches, and refund farming shows an elevated refund rate. - [Micro-affiliates: why 500 people at 2 sales beats 5 at 200](https://blog.korant.online/micro-affiliates-500-at-2-sales.md): Two programs producing the same 1,000 referred orders behave nothing alike. Five macro affiliates carry 20% dependency each and model at 38% incrementality; five hundred micro affiliates carry under 1% each and model at 66%, on a fifth of the management time. - [Quirky ideas in affiliate marketing that still work](https://blog.korant.online/quirky-ideas-in-affiliate-marketing.md): The affiliate ideas that still work share one property: they use an asset the brand already owns rather than buying reach. Nine of them, ordered by setup effort, from a thirty-minute packaging insert slug to a week-long pincode-exclusive partner offer. - [How to recruit your first 100 affiliates from your order list](https://blog.korant.online/recruit-first-100-affiliates.md): Your first hundred affiliates are already in your order list. Segment by repeat count and review activity rather than follower count, and the top two segments of a 5,000-customer base model to 153 activated affiliates against 12 from follower-based outreach. - [Tools to grow sales through affiliate marketing](https://blog.korant.online/tools-to-grow-sales-affiliate-marketing.md): Affiliate tooling covers four separate jobs: recruiting partners, tracking their sales, paying them, and controlling fraud. Most categories do one or two well, so a typical stack covers a modelled 2.4 of the four and the gaps get filled by somebody's spreadsheet. ### Hyperlocal and geo marketing - [Why your best customers are in six postcodes and you don't know which](https://blog.korant.online/best-customers-in-six-postcodes.md): Geographic customer concentration in ecommerce is steeper than most brands assume, with a modelled top decile of postcodes carrying 49% of revenue. Default store reports hide it because they group by billing region rather than by the shipping postcode that identifies where the customer lives. - [Delivery time is a marketing message you are not sending](https://blog.korant.online/delivery-time-as-a-marketing-message.md): A zone-specific delivery estimate is the cheapest conversion lever a D2C store has, because the transit data already exists in the courier serviceability file. On modelled costs the build breaks even at 6.3 incremental orders, against Rs 2,430 per incremental order for a sitewide discount. - [Delivery zones as a growth channel, not a logistics table](https://blog.korant.online/delivery-zones-as-growth-channel.md): Delivery zones are a segmentation axis, not just a shipping rate table. Loading shipping cost and return-to-origin rate into each zone produces a margin per order that varies by roughly a third between a metro core and a remote zone, which is enough to change where a brand spends and what it offers. - [Geo-targeted offers that don't leak to the wrong customer](https://blog.korant.online/geo-offers-that-dont-leak.md): A geo-targeted offer leaks whenever it gates on something the shopper controls. Of the eight signals a Shopify store can gate on, only the delivery address on the cart cannot be changed without also changing where the order physically ships, which makes it the only safe gate. - [Hyperlocal flash sales without a physical store](https://blog.korant.online/hyperlocal-flash-sales-no-store.md): A hyperlocal flash sale runs on the shipping postcode at checkout, not on a storefront. One delivery zone gets a live discount for a fixed window, the window rotates on a set cadence, and each zone is measured against its own prior weeks rather than against the rest of the country. - [Launching in one city at a time, on purpose](https://blog.korant.online/launching-in-one-city-at-a-time.md): A sequential city launch produces more customers than a simultaneous one from identical spend, because referral rate scales with how many of a customer's contacts can actually act on the offer. On modelled figures, concentrating a launch in one city yields 22% more customers than spreading it across six. - [Local scarcity: making 400 people feel like the only ones](https://blog.korant.online/local-scarcity-400-people.md): Local scarcity works because the audience is genuinely bounded. A strong postcode in a mid-size D2C brand contains roughly 400 of that brand's customers, so an offer restricted to one postcode is verifiably limited in a way that a countdown timer or a low stock badge is not. - [A marketing channel nobody has ever touched](https://blog.korant.online/marketing-channel-nobody-has-touched.md): Delivery serviceability data is the largest untapped marketing channel in D2C. Every Shopify store already records a postcode on every order and holds a serviceable-zone list from its courier, but almost none use either to vary an offer, because neither one appears in a default report. - [The pincode data you already have and have never exported](https://blog.korant.online/pincode-data-you-already-have.md): Building a pincode map needs three exports and one pivot. The Shopify order export supplies the shipping postcode and revenue, the courier serviceability file supplies transit time, and the courier invoice supplies delivery status. Joining them on the pincode key produces six measures no single system reports. - [Pincode marketing: the geo lever D2C brands ignore](https://blog.korant.online/pincode-marketing-geo-lever.md): A pincode is the smallest unit of buying intent a D2C brand already owns, because every order carries one and it encodes delivery speed, courier reliability and neighbourhood density at once. Pincode marketing varies the offer by that unit rather than by audience segment or by campaign. - [Regional pricing without a regional pricing team](https://blog.korant.online/regional-pricing-without-a-team.md): Most D2C stores already run regional pricing without deciding to. Shipping thresholds and cash on delivery fees produce a 7.2% spread in effective price between a metro and a remote pincode on modelled figures, and a zone-gated discount widens that spread to 41% in the direction of the metro. - [Serviceability data is a marketing asset, not an ops file](https://blog.korant.online/serviceability-data-is-marketing-asset.md): A courier serviceability file is a targeting list that operations happens to own. Its nine standard fields include delivery time, cash on delivery eligibility and reverse pickup coverage, and seven of the nine change a marketing decision if anyone outside operations ever opens the file. - [Testing new cities with a discount instead of a warehouse](https://blog.korant.online/test-cities-with-discount-not-warehouse.md): A discount window in a candidate city measures demand density before any capital is committed. On modelled costs a six month zone test across three cities runs at roughly 13% of a single dark store, and it is reversible, which a signed lease is not. - [Building a waitlist by pincode](https://blog.korant.online/waitlist-by-pincode.md): A pincode waitlist captures demand in areas you cannot yet serve, using the checkout attempts that currently get discarded. On modelled costs a zone needs about 24 orders to cover opening it, so an unlock threshold of 200 signups per pincode keeps the required conversion under 12%. - [The zip-code discount that beats a sitewide sale](https://blog.korant.online/zip-code-discount-beats-sitewide.md): A geo-targeted discount beats a sitewide sale by concentrating the same budget into fewer postcodes at a deeper rate, not by spending less. On modelled economics, a budget that funds 10% off everywhere funds 24% off in the top 20 postcodes, and that depth has to produce a 19.5% lift to win. - [Pincode targeting vs city targeting: why the delivery address wins](https://blog.korant.online/pincode-targeting-how-delivery-postcodes-gate-offers.md): Pincode targeting gates an offer on the delivery postcode entered at checkout, while city or IP targeting gates on an inferred browsing location. The delivery address is enforceable: a shopper who fakes it loses the parcel, so the gate holds where an IP guess does not. ### Urgency, scarcity and flash - [Why automatic discounts outperform code-entry every time](https://blog.korant.online/automatic-discounts-beat-codes.md): A discount code field is a price-comparison prompt placed at the most expensive point in the funnel. Research puts the share of shoppers who would abandon a cart to hunt for a voucher code at 27%, and on modelled figures the field costs more than simply applying the discount automatically. - [Countdown timers are dead. This replaced them.](https://blog.korant.online/countdown-timers-are-dead.md): Client-side countdown timers fail because shoppers test them and the test comes back false. At a modelled 15% chance of noticing per exposure, 80.3% of shoppers who see a resettable timer ten times have caught it. A server-held window survives the same test. - [Building a discount calendar your competitors cannot read](https://blog.korant.online/discount-calendar-that-doesnt-leak.md): A readable promotional calendar is read by three audiences you did not intend it for: competitors, coupon aggregators and your own customers. On modelled figures, moving from a published calendar to unforecastable timing lifts full-price share from 48.7% to 97.4% at an unchanged discount cadence. - [The discount stacking rules that quietly kill your margin](https://blog.korant.online/discount-stacking-rules.md): Shopify sorts discounts into three classes and applies them in a fixed order, so an order discount always computes on an already-reduced subtotal. Four individually reasonable discounts compound to a 46.6% giveaway on a modelled Rs 2,000 order whose break-even sits at 31.05%. - [The five-minute discount window: does it convert or annoy](https://blog.korant.online/five-minute-discount-window.md): A discount window converts only the shoppers who can decide inside it. A five minute window reaches 35% of interested shoppers in an impulse category with an eight minute median decision time, and 0.08% in a considered category with a three day median. Everyone else is annoyed. - [Flash mechanics for slow-moving inventory](https://blog.korant.online/flash-mechanics-slow-inventory.md): Slow-moving stock is rarely slow everywhere. On a modelled demand map, the three postcodes where a slow SKU already sells carry 31.6% of its demand while representing 0.9% of the brand's map, so a deep discount there reaches buyers at 34 times the exposure efficiency of a sitewide clearance. - [Flash sales without the brand damage](https://blog.korant.online/flash-sales-without-brand-damage.md): Discounting does not damage a brand. Frequency, breadth and predictability do, because together they teach shoppers to wait. On a modelled three week patience constant, fortnightly sitewide sales move 71.7% of demand off full price while a rotating window moves 3.6%. - [Running a flash window during a traffic spike](https://blog.korant.online/flash-window-during-traffic-spike.md): A flash window creates its own traffic spike, so the discount decision cannot happen on the request path. Pre-computing the live offer into a metafield costs one write per rotation against roughly 120,000 lookups for a per-request architecture during a ten minute spike. - [How often can you discount before shoppers wait for it](https://blog.korant.online/how-often-can-you-discount.md): Shoppers who have learned to wait show it in three measurable places: orders concentrating in sale windows, a trough in the fortnight before each event, and a repurchase interval drifting toward your sale period rather than their consumption rate. - [Rotating offers: one discount, many audiences, no code](https://blog.korant.online/rotating-offers-one-discount-many-audiences.md): A rotating discount runs in one audience at a time and moves on a fixed cadence, which turns the cycle length into a budget control. On modelled figures a 20 week rotation discounts 2.06% of orders and costs 5.2% of an always-on 10% offer while running at 25% depth. - [Scarcity that's true: how to run urgency you can defend](https://blog.korant.online/scarcity-you-can-defend.md): Real scarcity comes from three sources only: inventory that genuinely runs out, time that genuinely expires, and an audience that is genuinely bounded. Anything else is false urgency, which India's dark patterns guidelines list first among thirteen practices treated as unfair trade practices. - [Time-boxed offers and the two clocks you must never mix](https://blog.korant.online/two-clocks-you-must-never-mix.md): A time-boxed offer runs on two independent clocks. The offer clock decides whether a discount applies to an order, and the reward clock decides whether a referral earns. Tying the second to the first disables referral for 95% of referrals in a twenty zone rotation. - [Urgency that survives a second visit](https://blog.korant.online/urgency-that-survives-a-second-visit.md): Repeat visitors produce a modelled 56.2% of orders from 30% of traffic, and they are the only people who can detect a resettable urgency element. Any mechanic that restarts on a second visit is being tested by the exact segment that carries most of the revenue. ### Attribution and measurement - [Attribution window: what it is and how window length changes credit](https://blog.korant.online/attribution-window-how-window-length-changes-credit.md): An attribution window is the period after a click or impression during which a resulting sale is still credited to that touchpoint. A 7-day window credits purchases made within seven days; anything later is credited elsewhere. - [Attribution for a brand with nobody to run it](https://blog.korant.online/attribution-for-a-brand-with-no-analyst.md): A brand with no analyst needs four numbers reviewed weekly: total orders, new customer share, blended acquisition cost, and unattributed share. Fifteen minutes, one sheet, no dashboard. Everything more sophisticated goes unread and therefore changes nothing. - [Attribution for agencies running ten brands at once](https://blog.korant.online/attribution-for-agencies-ten-brands.md): Agency attribution needs per-client data isolation at the storage layer, not at the report layer. Run a tenant-isolation test before buying any tool: namespace collision, cookie scope, export boundaries, role separation, token scope, deletion, and portfolio aggregation. - [Cohort reporting your CFO will read](https://blog.korant.online/cohort-reporting-cfo-will-read.md): A cohort report a finance team will read fits on one page: acquisition month and channel down the side, months since acquisition across the top, cumulative contribution margin per customer in the cells, and the payback month marked. Everything else belongs in an appendix. - [How to measure a channel that has no clicks](https://blog.korant.online/measure-channel-with-no-clicks.md): Channels with no click surface are measured by geo-holdout rather than by tracking. Split matched delivery zones into exposed and held-back groups, compare total orders rather than attributed ones, and run long enough that the effect exceeds the natural variation in order counts. - [Server-side tracking for stores that can't afford a data team](https://blog.korant.online/server-side-tracking-no-data-team.md): Webhook-first attribution treats the Shopify order as the event rather than a browser pixel, which removes the 6.6% of orders a client-side pixel loses to blockers and broken sessions. It needs one webhook subscription, one storage table, and a deduplication key. - [What your CAC number is actually missing](https://blog.korant.online/what-your-cac-is-actually-missing.md): Most quoted CAC figures contain ad spend and nothing else. Adding agency and creative fees, first-order discount, and referral payout takes a modelled ₹1,091 to ₹1,364. Payment fees, shipping subsidy and return losses belong in contribution margin, not CAC, and are frequently in neither. - [Your attribution is wrong. Here's how wrong.](https://blog.korant.online/your-attribution-is-wrong.md): Ad platforms typically claim more conversions than a store has orders, because each counts a different event inside a different window with no shared deduplication. On a modelled 1,000-order month, four platforms claim 1,260 conversions while roughly 45% of orders are claimed by nobody at all. - [First-touch vs last-touch for D2C: pick one and defend it](https://blog.korant.online/first-touch-vs-last-touch-d2c.md): First-touch and last-touch attribution answer different questions and should be reported as two separate columns, never averaged. Scored on the same 100 orders, first-touch credits discovery channels like influencer and paid social, while last-touch credits closing channels like email and direct. - [Tracking influencer sales without a discount code](https://blog.korant.online/track-influencer-sales-without-code.md): Discount codes capture only the fraction of an influencer's audience that remembers to type them. Replacing the code with a unique redirect slug, a first-touch cookie and a post-purchase confirmation question takes modelled coverage of genuinely creator-driven sales from 34% to 78%. - [The attribution window that matches your buying cycle](https://blog.korant.online/attribution-window-matches-buying-cycle.md): Attribution windows should be derived from the 80th percentile of your own time from first session to order, not taken from a platform default. Modelled across three D2C categories that gives 4 days for consumables, 14 for apparel, and 47 for considered high-ticket purchases. - [Post-purchase surveys vs pixel data: use both, weight one](https://blog.korant.online/post-purchase-survey-vs-pixel.md): Pixels measure precisely what they can see and are blind to every channel without a click. Surveys see the dark channels and remember imprecisely. Weight pixel data for channels with a click surface, survey data for channels without, and state the rule on every report. - [Building a single source of truth for marketing spend](https://blog.korant.online/single-source-of-truth-marketing-spend.md): A single source of truth is one resolver that assigns each order to exactly one channel through a fixed priority order, with every dashboard built as a report on top of it. The priority order matters less than writing it down and never changing it mid-quarter. - [UTM hygiene: the thirty-minute fix worth lakhs](https://blog.korant.online/utm-hygiene-thirty-minute-fix.md): Unmanaged UTM tagging produces dozens of spellings for a handful of real channels, splitting every channel's performance across rows nobody adds back together. Lowercasing everything and generating tags from a fixed table instead of typing them fixes most of it in half an hour. - [What to do when Shopify, Meta, and GA4 disagree](https://blog.korant.online/when-shopify-meta-ga4-disagree.md): Shopify, Meta and GA4 disagree because they count different events, timestamp them differently, and apply different attribution windows. Reconcile by fixing Shopify orders as the denominator, restating everything on order date, separating view-through, then reporting the residual instead of averaging. ### Influencer and creator operations - [Ambassador programs: the middle ground between affiliate n influencer](https://blog.korant.online/ambassador-programs-middle-ground.md): An ambassador program sits between an affiliate scheme and an influencer roster: self-serve signup, a small standing reward, and no negotiation per post. On modelled figures that costs Rs 512 a post against Rs 14,400 for a negotiated micro-influencer. - [Barter, fee, or commission: a decision table](https://blog.korant.online/barter-fee-or-commission.md): Barter of a Rs 1,800 product at 35% gross margin costs Rs 1,225 including shipping, which is 68% of the value the creator assigns it. That arbitrage is entirely a function of margin, so the right structure depends on margin, creator tier and whether you need content rights. - [The creator payout report that ends the monthly argument](https://blog.korant.online/creator-payout-report.md): Payout disputes come from two invisible things: how a sale was credited and why a total changed. A creator sees 90 link clicks and is told 3.6 orders, a 4% conversion that reads as underreporting until the resolution path is shown per order. - [Creator contracts in plain language](https://blog.korant.online/creator-contracts-in-plain-language.md): Seven clauses cover almost every creator dispute worth preventing: deliverables, timing and minimum live period, usage rights, payment terms, attribution, exclusivity, and approval rounds. Each one exists because a specific argument happened to somebody, and the plain version prevents it. - [Finding creators whose audience overlaps yours](https://blog.korant.online/finding-creators-with-audience-overlap.md): Effective audience is followers multiplied by overlap, and overlap swings further than follower count does. A modelled 200,000-follower creator at 3% overlap reaches 6,000 relevant people while a 20,000-follower creator at 35% overlap reaches 7,000, at a fraction of the fee. - [Why your influencer campaign has no data and how to fix it](https://blog.korant.online/influencer-campaign-with-no-data.md): Creator campaigns lose data because tracking gets added after the brief rather than before it. Of a modelled 100 people who see a post, 40 tap the link and 25 search the brand later, so link tracking alone resolves 62% of those who acted. - [The influencer brief that gets you usable content](https://blog.korant.online/influencer-brief-that-works.md): A brief works by constraining rather than scripting. Three must-haves and one must-not give a creator four things to get right, against roughly forty in a written script, which is where most revision rounds come from. - [Managing 200 creators without a spreadsheet graveyard](https://blog.korant.online/managing-200-creators-without-spreadsheets.md): Creator management scales when every creator sits in one of five explicit states rather than in a free-text column. On modelled figures that shift takes a manager from 53 creators to 160, because computing attribution and payout falls from 0.50 hours per creator to 0.05. - [Nano-influencers in India: the unit economics](https://blog.korant.online/nano-influencers-india-unit-economics.md): Nano creators produce the lowest modelled CAC of any tier at Rs 1,688 against Rs 10,356 for macro, and the advantage does not erode with scale. What breaks is throughput, since 500 orders a month needs about 312 creators and eight people to manage them. - [Turning a one-off collab into a standing revenue line](https://blog.korant.online/one-off-collab-into-revenue-line.md): A one-off creator post delivers most of its orders in the first week and stops. On modelled figures a standing link reaches 14.98 orders over a year against 4.98 for the same post left to decay, and chained referral takes that to 17.6. - [Paying influencers on performance without insulting them](https://blog.korant.online/paying-influencers-on-performance.md): Commission-only offers get declined because of the number rather than the structure. A modelled 20,000-follower creator produces 3.6 attributed orders, so a 15% commission is worth Rs 972 against a Rs 8,000 flat fee. A base plus uncapped upside is the version that pays for itself. - [Publication and PR links: measuring what SEO teams can't](https://blog.korant.online/publication-and-pr-links.md): Publications introduce customers and rarely close them, so last-touch attribution credits a fraction of what coverage produces. On modelled figures a publication generating 19 orders among its clickers gets credited with 4, an undercount of 4.75 times. - [When a creator campaign flops, what to check first](https://blog.korant.online/when-a-creator-campaign-flops.md): A creator campaign has seven sequential failure points and a zero-order result looks identical at every one of them. Pulling reach, clicks and orders localises the break to a single link in the chain, and the creator is usually the last thing to blame rather than the first. ### D2C growth experiments - [Cart abandonment flows that don't beg](https://blog.korant.online/cart-abandonment-flows-that-dont-beg.md): An escalating discount flow recovers more carts and earns less. On modelled figures a reason-led flow recovers 8% of carts for Rs 35,280 of contribution, while a flow escalating to 20% off recovers 13% and nets Rs 24,120 after strategic abandonment. - [The growth loops that don't need paid media](https://blog.korant.online/growth-loops-without-paid-media.md): Three loops compound without paid media: referral, content and geo-density. Cycle time matters more than loop strength, because a modelled geo-density loop at a 1.10 coefficient and a 30 day cycle doubles in 218 days while a content loop at 1.25 and 120 days takes 373. - [Nine D2C experiments you can run this week](https://blog.korant.online/nine-d2c-experiments-this-week.md): Nine experiments ship in under a day each, and the useful discipline is checking the sample size first. At 36,000 monthly sessions, detecting a 20% relative change in checkout completion takes 2 days, while the same change in order conversion takes 25. - [The unlock nobody talks about: your second order](https://blog.korant.online/the-second-order-unlock.md): The second-order rate sets the ceiling on what a brand can pay to acquire anyone. Raising it from 25% to 45% lifts modelled lifetime contribution from Rs 980 to Rs 1,260, which is a 29% larger acquisition budget rather than a retention improvement. - [Viral D2C hacks to grow sales](https://blog.korant.online/viral-d2c-hacks-to-grow-sales.md): Mechanics that create their own distribution beat campaigns that buy it, because the reach arrives with every order rather than with every rupee. On modelled share rates, a referral link on the order confirmation reaches 96 people per 100 orders for half a day of work. - [WhatsApp as a growth channel for Indian D2C](https://blog.korant.online/whatsapp-as-growth-channel-india.md): WhatsApp forwards are the largest untracked channel in Indian D2C. A modelled share into a group of eight, forwarded onward at 10%, reaches 31.6 people across five generations, and none of them arrives carrying a referrer your analytics can read. - [What to do when your CAC crosses your AOV](https://blog.korant.online/when-cac-crosses-aov.md): Comparing CAC to AOV is a category error. AOV is revenue and CAC is paid out of margin, so at a 35% gross margin a CAC equal to AOV is 2.86 times the contribution of one order. The comparison that matters is CAC against cumulative contribution. - [Zero-budget launches that actually moved numbers](https://blog.korant.online/zero-budget-launches.md): Three mechanics launch a product with no media spend: a pre-built waitlist, a zone-gated early access window, and chained invites. On modelled figures they produce 202, 179 and 141 launch-day orders, and all three draw on the same customer base. ## Definitions - [Shopify Function](https://blog.korant.online/glossary/shopify-function): A Shopify Function is custom logic compiled to WebAssembly and shipped inside an app, which Shopify executes on its own servers at defined points in checkout such as applying a discount or hiding a delivery option. - [Pincode targeting](https://blog.korant.online/glossary/pincode-targeting): Pincode targeting varies an ecommerce offer, price, or availability by postal code, using delivery serviceability and order-density data the brand already holds rather than inferred audience segments.