In short
- Give every creator one short, speakable, first-order-only code with a cap and an expiry, and pair it with a link that applies the code automatically.
- A discount is a cost. Work the maximum discount back from margin, and count either revenue net of discount or the discount as a cost line, never both.
- Leakage shows up as redemptions with no matching link click, odd timing, off-audience cities or existing customers. Report verified orders and total code orders as a range.
- Codes miss buyers who never type them and credit leaked redemptions to creators who did not reach those buyers, so cross-check against a post-purchase survey.
A unique discount code for each creator tracks influencer sales by recording which code a customer typed at checkout. It catches buyers who watched a Reel, searched your brand and never clicked a link, which is why it is the most common tracking device in India. It is also the leakiest: codes spread to coupon sites and WhatsApp deal groups, and buyers who do not type a code go uncounted. Treat code orders as one reading, not the answer.
This guide covers code design, the maths of the discount itself, how to catch leakage, what codes miss and how to pay creators on them. It goes deeper than the code rules in our performance influencer marketing service page, and it feeds the tracking steps in our guide to measuring influencer marketing ROI. Clean code data is rare: only 13.3% of surveyed Indian brands tie influencer work to formal revenue targets, according to a 2026 vendor survey covered by Campaign India. For the wider picture, see the complete guide to influencer marketing in India.
What a code can and cannot tell you
A code is wrong in two directions at once, so it helps to name both.
| Direction | Cause | Effect on your numbers |
|---|---|---|
| Undercount | Viewers who buy without typing the code; buyers on marketplaces; people who buy later after searching | Creator impact looks smaller than it is |
| Overcount | Codes posted on coupon sites, shared in deal groups or reused after expiry; existing customers using a new code | Creators get credit for people they never reached |
The two errors do not cancel out in any predictable way. Neither direction has a published India size that we could find, so measure your own with the checks below.
Design the code
Most code problems are design problems. Decide these before the first post.
| Decision | Rule | Reason |
|---|---|---|
| Scope | One code per creator per campaign; one per deliverable if you compare formats | Shared codes block creator-level cost per order |
| Spelling | Short, speakable and unambiguous: a stem the creator picks plus the discount, such as TADKA15 for a cooking creator | Viewers hear codes in video and type them on a phone |
| Characters | Avoid letters that look like digits (O and 0, I and 1) | Typos create failed redemptions and support tickets |
| Eligibility | First order only, one use per customer | Stops the code becoming a standing discount |
| Cap | A total redemption limit set from the creator's expected reach | Limits the damage if the code leaks |
| Expiry | A hard end date that matches the campaign and usage-rights window | Old posts keep promoting dead offers otherwise |
| Stacking | Not combinable with sale or welcome codes | Prevents deeper discounts than you priced |
| Minimum order | Set it if shipping would otherwise make small orders loss-making | Protects margin |
| Linked URL | A tracked link that applies the code automatically | Clickers do not lose the offer by forgetting to type it |
Keep every code in a ledger: creator ID, code, discount, start and end dates, cap, linked URL, owner and status. The ledger is what lets you reconcile redemptions with creators later, and it is the first thing to hand over when a creator relationship ends and the code should be switched off.
Discount depth: a cost, not a gift
A discount comes straight out of gross profit, and the first order is where you can least afford it. This example uses placeholder numbers: an order value of ₹1,200 and a 55% gross margin, so product cost is ₹540.
| Code discount | Price paid | Gross profit | Change against no discount |
|---|---|---|---|
| None | ₹1,200 | ₹660 | n/a |
| 10% | ₹1,080 | ₹540 | -18% |
| 15% | ₹1,020 | ₹480 | -27% |
| 20% | ₹960 | ₹420 | -36% |
Each extra 5 points of discount costs another ₹60 on every order, paid on top of the creator fee, the agency fee and shipping. We found no published India benchmark for what discount works best, so work yours back from the numbers: gross profit per order, minus the profit you want to keep, minus the acquisition cost you can afford.
Whatever you choose, count it once. Either report revenue net of the discount, or report gross revenue and add the discount as a cost line. Doing both subtracts it twice, and it makes campaigns look worse than they were. Our cost per acquisition post shows the full break-even method.
How codes leak
Leakage means redemptions by people the creator did not reach. Each route leaves a different footprint.
| Route | What you see in the data | Response |
|---|---|---|
| Coupon sites and browser extensions that test codes at checkout | Redemptions with no matching link click; spikes when someone abandons a cart | Link-applied codes, one use per customer, caps |
| WhatsApp and Telegram deal groups | Sudden bursts at odd hours; many different cities | Cap, pause and rotate the code |
| Reposts after expiry | Redemptions after the creator's last post; failed code attempts | Hard expiry, tell creators to remove pinned comments |
| Existing customers | A high share of redemptions from emails already in your store | First-order-only rule; report them as revenue, not acquisitions |
| Staff or support sharing | Codes issued outside the campaign | Keep codes out of support scripts |
Run a weekly check while a campaign is live. Search your brand name plus "coupon code" on Google and see which codes appear. Compare redemptions by city with the creator's audience geography. Look at timing: a post usually produces a burst of redemptions soon after it goes live, so redemptions that run at a steady rate long after the last post deserve a closer look.
A leak check you can run in an afternoon
This is a worked example with made-up numbers for one creator's code, TADKA15, with a creator cost of ₹30,000. You need the order export with customer email, the discount code and the session source from your UTM data. Our guide to UTM links for influencer campaigns explains how to get the source onto the order.
| Step | Count | Reading |
|---|---|---|
| Total redemptions of the code | 160 | Every order that used it |
| Order session came from the creator's UTM link | 70 | Verified: link and code agree |
| Of those, new customers | 56 | Verified acquisitions |
| Code only, no creator link session | 90 | Route unknown |
| Of those, existing customers | 54 | About a third of all redemptions: revenue, not acquisitions |
| Of those, new customers | 36 | Plausible: saw the Reel, searched, typed the code |
Report a range, not a single figure. The low end credits only the 56 verified new customers, giving a cost per new customer of about ₹536. The high end adds the 36 plausible code-only new customers, 92 in all, for about ₹326. The 54 existing customers sit outside the acquisition count, and a share that high deserves a conversation with whoever manages the code. Never quote only the flattering end.
What codes miss: the capture ratio
To estimate how much creator-driven demand carries no code, compare code orders with a post-purchase survey. Suppose there were 2,000 orders in the campaign window. If 31% of respondents say a creator was how they first heard of you, that is about 620 creator-influenced orders. If all creators' codes together account for 420 orders, the capture ratio is 68%: roughly a third of creator-influenced orders never touch a code.
Survey answers are self-reported and some buyers misremember, so use this as a rough guide, not a correction factor to multiply into revenue. It does tell you whether codes are close to the truth or far from it. The reconciliation of all the methods is in our post on influencer marketing attribution.
Marketplaces, stores and apps
- Marketplaces. Creator-specific codes are often not possible on marketplaces. Check the seller tools, and where codes are unavailable, read sales lift on the listing against a pre-campaign baseline. You will not get creator-level numbers.
- Stores and restaurants. Give the cashier a code to enter at billing and keep a tally. A code printed on a table card does not tell you which creator sent the diner, so use one code per creator.
- Apps. Offer code entry in the onboarding flow, and use a mobile measurement partner link as the primary source. Codes are the backup.
Paying creators on code sales
If the creator's pay depends on code orders, the contract must define the order. Pay on net orders after returns and refused cash-on-delivery parcels, release the bonus after the return window, and cap the total. State whether existing customers count. A bonus on gross redemptions pays the creator to push the code onto deal sites. A fixed fee plus a bonus on verified net orders is the safe structure, as described in our influencer contract guide. Payments to creators carry TDS and GST questions, covered in TDS and GST on influencer payments.
Disclosure still applies
A code in a caption does not replace a disclosure label. The ASCI guidelines permit labels including Advertisement, Ad, Sponsored, Collaboration, Partnership and Affiliate, and they must be upfront and prominent, not only in the bio or buried in hashtags. Our ASCI influencer guidelines explainer has the placement rules. This is general information, not legal or tax advice.
A pre-launch checklist
- Write the code ledger and fill every column.
- Place a test order with each code and confirm the discount, the cap and the first-order rule.
- Test that each creator's link applies the code automatically.
- Confirm the order export shows the code and the UTM source.
- Set up the weekly "coupon code" search.
- Add a post-purchase "How did you hear about us?" question.
- Record a baseline for orders and branded search before the first post.
- Agree in the contract how codes are paid and when they are switched off.
Mistakes to avoid
- One code for all creators. It tells you the campaign worked and nothing about who did the work.
- Public, permanent, generous codes. They leak first and expire last.
- Counting discounted revenue gross and the discount again as cost. One or the other.
- Crediting existing customers as acquisitions. Their orders are revenue, not new customers.
- Judging a creator on code orders alone. Read codes beside links, survey answers and branded search, as the KPI guide sets out.
If you want codes, links and a leak check built into a creator plan, send us a brief. The cost guide shows what the creator fees themselves look like.
Sources
- Coverage of a 2026 India creator-economy report: brand survey on revenue targets and campaign-by-campaign judging, Campaign India (21 May 2026)
- ASCI Code and guidelines for influencer advertising in digital media (permitted disclosure labels, placement), ASCI (Undated, accessed 31 Oct 2026)
Frequently asked questions
How do I track influencer sales with a discount code?
Create one unique code per creator, set it to apply once per customer on a first order, and read redemptions from your store's discount report. Add a tracked link that applies the code automatically, so you can separate link-plus-code buyers from code-only buyers. Treat code orders as a floor for the creator's influence, not the total.
How many discount codes should each influencer have?
At least one per creator per campaign. If you want to compare formats, issue one per deliverable, for example one for the Reel and one for the Stories. A code shared by several creators makes creator-level cost per order impossible to calculate, so avoid shared codes except for a deliberate community code.
What discount should an influencer code offer?
We found no published India benchmark for the right discount, so work it back from your numbers. Take gross profit per first order, subtract the profit you want to keep and the acquisition cost you can afford, and what is left is the most you can give. A 15% code takes about 27% of first-order gross profit at a 55% margin.
How do I stop influencer discount codes leaking to coupon sites?
Restrict each code to one use per customer and first orders only, cap total redemptions, set a hard expiry, and apply the code through the creator's link instead of printing it publicly. Search for your brand plus 'coupon code' weekly. If redemptions rise with no link clicks, pause and replace the code.
Can I track discount codes on marketplaces or in offline stores?
Marketplace tools often do not let you issue creator-specific codes, so check the seller settings and otherwise read sales lift on the listing against a baseline. In stores and restaurants, give the cashier a code to enter at billing and log each use. Our performance influencer marketing service covers both.
Should I pay influencers on discount code sales?
A fixed fee plus a bonus on verified net orders is the usual safe structure. Pay after the return window, count only new-customer or link-verified orders, and cap the bonus. Paying on gross redemptions tempts creators to post codes on coupon sites. See influencer contracts in India.
Does a discount code in a caption count as disclosure?
No. The code is part of the promotion, and the post still needs a clear, upfront label such as Collaboration, Partnership or Affiliate under ASCI's influencer guidelines. Read our ASCI guidelines explainer for placement rules. This is general information, not legal or tax advice.