Discount Optimizer
One clear action per product, per city, every Monday — raise, invest, hold or test. Safety-capped moves, verified by a scorecard you keep.
See the full offer
For consumer brands on
Cut discount waste. Grow real sales.
A slice of every discount budget goes to shoppers who'd buy at full price — today, that slice is invisible. We make the waste visible and hand you one clear action per product, per city, every Monday — done for you, on your own data.
Verify before you commit — a one-week data check up front. If the data isn't enough, we say so and stop.
Worked example
One representative brand, shown in full — so you can judge the method, not a testimonial.
Discount spend
0%
was subsidy — spent on shoppers who'd have bought anyway
Surfaced
₹0L
of recoverable waste a month, concentrated in a handful of SKUs
Figures throughout this page are illustrative — a demonstration of how the system reasons on representative data, not the result of a client engagement.
What we do
One causal engine. Three ways in.
The Discount Optimizer is the flagship. The same engine — your own data, causal analysis, rupee-denominated answers — also powers two focused engagements, built for the questions consumer brands on quick-commerce most often need answered.
Platform-wise ROAS reallocation
Where does the next ₹10,000 of ad spend earn most — Blinkit, Zepto or Instamart? Marginal-ROAS curves per platform and product show where budget still compounds, and where it's saturated.
Ask about ROASRoot-cause analysis (RCA)
Sales moved — but why? We trace what's working and what isn't in your own data — price, discounts, availability, visibility — and hand you the fix, not another chart.
Ask about RCADiscount money leaks in three ways.
Dashboards count units sold — none of them separate the sale you caused from the one you'd have made anyway. So the wasted discount hides in plain sight, and it hides in three distinct ways.
Paying shoppers who'd buy anyway
Your regular buyer doesn't need 20% off. That discount is margin gone for zero extra volume.
Paying for stockpiling
Deep promos pull next month's sales into this month. The spike looks great; the dip cancels it.
Paying one pack to rob another
The 1L promo steals from your own 500ml. Units move sideways — but it books as a “win”.
Net effect: your largest promotional line is also your least measured one.
The Discount Optimizer.
One clear action per product, per city — every Monday. Delivered as a ready-to-execute workbook: exact prices, this week and the weeks ahead.
Illustrative workbook — your version carries your own products and exact prices, with the reason behind each move.
Price up in small steps where shoppers were buying anyway. Margin comes home.
Redirect the savings to where deeper discounts genuinely create new volume.
Where the discount is earning its keep, nothing changes.
Where evidence is thin, a small price test builds it — no guessing.
A done-for-you weekly rhythm — about one hour of your team's time.
Every Monday
Action workbook
Exact price per product per city — and why.
Every Monday
30-min review call
You approve before anything changes.
Always current
Leadership dashboard
Spend, savings and sales impact on one screen.
Updates weekly
Reality scorecard
Every prediction graded against what happened.
Worst case: one week of small mispricing, in one city.
Small steps only
About 3 percentage points a week — never a shock.
Only proven prices
Every target is a price this product has already sold at, in this city.
It refuses to guess
Thin data gets a test plan, not an invented number.
You stay in charge
Nothing changes without your approval at the Monday review.
From you: a daily sales export (6+ months) and your brand names. Setup takes about a day.
Book a 30-min call
A 30-minute call scopes the one-week data check — no commitment.
Proof by method, not by logo
Tested on data it has never seen. Its error rate is published.
We're a new lab, so we won't borrow a logo or a quote you can't verify. Instead, here is the method run end to end on a representative brand — measured, graded, and honest about what it can't yet see.
0%
of discount spend
was recoverable
was recoverable
Waste concentrated in a handful of SKUs
Measured, not estimated — every rupee tied to a product and a city.
Of sales response explained on unseen data
Labelled “moderate” by the system itself — never rounded up.
Flagged −13.8% downside; held-out weeks came in at −8.8%
Built to err toward caution when evidence is thin — it would rather over-warn than surprise you.
Is an answer it's allowed to give
Weak evidence gets flagged for testing — not guessed at.
| Product | City | Price now | Move to | Action | Recovered / mo |
|---|---|---|---|---|---|
| Cold-pressed oil 1L | Bengaluru | ₹285 | ₹294 | Raise | ₹42,800 |
| Peanut butter 340g | Mumbai | ₹249 | ₹229 | Invest | – |
| Wild honey 500g | Delhi | ₹449 | keep | Hold | – |
| Millet muesli 400g | Hyderabad | ₹315 | ₹325 | Raise | ₹31,500 |
| A2 ghee 500ml | Pune | thin data | A/B test | Test | – |
| Illustrative recovery · 11-city sample | ₹1,76,000 | ||||
These numbers are one worked example — the logic is what carries. If even 1 in 10 discount rupees is subsidy, then on a ₹1 crore monthly discount budget that's ₹10 lakh, every month, waiting to be won back.
“Not enough data” is an answer it's allowed to give.
Weak evidence gets a test plan — never an invented number. Honesty is a feature, not a disclaimer.
How it works
Technical under the hood. Simple on the surface.
Four steps, one weekly loop. No proposals to negotiate, no scope creep — your data in, Monday's plan out.
-
01
Learn your history
6–12 months of daily sales show how volume responded to every price move you've already made.
-
02
Grade the evidence
Strong evidence acts. Weak evidence is locked out until a test builds it — no guessing allowed.
-
03
Recommend the moves
The best price per product, per city — written into Monday's plan as an exact, ready-to-execute number.
-
04
Score itself
Every prediction checked against reality, on a scorecard you keep. Accuracy is public between us.
Full methodology available for your analysts on request.
Why this wins
Beats the agency playbook — and building in-house.
| Typical agency | Build in-house | This service | |
|---|---|---|---|
| Counting wins | Every promo unit counted | Depends who builds it | Fake lift netted out |
| Accountability | Rarely tracked | Unaudited | Weekly scorecard |
| Time to value | Ongoing retainer | 3–6 months | Verdict in week 1 |
| If data is weak | Runs anyway | Found out late | We say so — and stop |
Who we are
Your data speaks. We translate.
StatIQ Lab is a boutique analytics practice for consumer brands whose discount spend on quick-commerce has grown large enough to be worth measuring: statistics and causal methods, applied to your own sales data, returned as rupee-denominated decisions. No dashboards to babysit, no black box — every recommendation arrives with its evidence attached.
The aim: a world where no brand leaves growth to guesswork.
How we start
Verify first. Commit only while the results hold up.
No long contract, no leap of faith. One decision today: agree the data export.
-
Week 0
Share the export
Daily sales per product per city. Setup: about a day.
-
Week 1
Get your verdict
A clear read on what can be optimized now. Go / no-go is yours — if the data isn't enough, we say so and stop.
-
Weeks 2–12
Weekly optimization
Monday workbook, 30-min review, and a scorecard filling with results from day one.
-
Month 3
Decide on evidence
Continue, renegotiate on documented savings, or walk away. Everything stays with you.
Commercials: discussed on our first call — structured so the service funds itself out of recovered waste. Prefer a lighter start? The same call can scope a one-off promo-leak audit if you're not ready for the weekly service.
Before you ask
The questions every brand asks first.
Is our sales data safe with you?
Yes. We sign an NDA before any data moves, work on read-only exports you control, and never share, resell or benchmark your data against other brands. Everything we build from it — models, scorecards, workbooks — stays yours.
What do you need from us to start?
One daily sales export per product per city (6+ months of history) and your brand names on each platform. No access to your ad accounts or seller panels is required. Setup takes about a day.
What if our data isn't good enough?
Then we tell you in week one and stop — that's the point of the up-front data check. You'll know exactly what's missing and how to start capturing it. You pay for answers, not for effort.
Will this break our promo calendar or platform relationships?
No. Moves are capped at roughly 3 percentage points a week, only to prices your product has already sold at in that city, and nothing changes without your approval at the Monday review.
Why not just build this in-house?
You can — it typically takes a data hire 3–6 months to reach a first trustworthy model, unaudited. This service gives you a verdict in week one and grades its own predictions every week — a low-risk way to de-risk an in-house build before you commit to it.
How is the engagement priced?
Commercials are agreed on the first call and structured so the fee funds itself out of documented recovered waste — if the scorecard doesn't show the savings, you have every reason (and the evidence) to walk away.
Let's find out what
your data supports.
A 30-minute call is enough to scope the data check and agree commercials. If the data isn't enough, we say so — and you've lost nothing. The waste, though, doesn't wait: every month it stays unmeasured, it's spent again.
Response within one business day