Harpy Guide

How to Rank Top-3 on Amazon India

10 min read · Harpy Media

The practical version of the Top-3 System: six steps in order — fundamentals, keyword front-loading, review engineering, keyword isolation, inventory-synced bids, and the 14-day defence.

Part 1

What a Top-3 position actually is

Everyone asks how to rank on Amazon. The better question is where — because on Amazon, position is economics. The first screen of search results collects the overwhelming majority of clicks; anything below it fights for scraps. A Top-3 position on your money keyword means your product is on that first screen, in front of buyers with intent, without paying for every single visit.

That last part is the point. Sponsored placements rent you the position — the meter runs on every click. Organic Top-3 owns it: the same buyers, zero marginal cost. That difference, compounded over months, is the entire economics of a brand on Amazon.

🔑 Key insight. Rank without fundamentals is a rented position. Ads can place you anywhere; only conversion, fulfilment, and stock can keep you there. Every step below exists to make the position defensible instead of leased.

This is the practical version of the Top-3 System (our Frameworks № 02) — six steps, in order, the way we run them for client accounts. Skipping steps doesn't speed this up; it just decides in advance which step you'll repeat.

Part 2

Step 1 — win the fundamentals

Before a single bid: stock depth, fulfilment speed, and Buy Box eligibility. These aren't hygiene — they're ranking inputs. Amazon reads them the way a bank reads your credit history.

FundamentalWhy the algorithm cares
Stock depthEvery lookback window — days, weeks, months — is accumulating sales. A stockout writes zeros into all of them at once, and recovery costs more than the inventory ever did
Fulfilment speedDelivery promise is a conversion input; conversion is the master ranking signal. Slow promise, fewer sales, weaker rank
Buy Box eligibilityNo Buy Box, no ad placements, degraded conversion — three losses from one miss
⚠️ Watch out. Never overpay into a stock-out. Bidding aggressively on rank you can't fulfil is the single most expensive way to lose twice: you pay premium CPCs, then the algorithm resets you when inventory blanks. Bid to what you can ship, always.

Part 3

Step 2 — front-load the money keyword

The money keyword is the term your best buyers actually type — the root phrase the rest of your category's searches hang from. Relevancy is scored in a strict hierarchy, and the top of it is the exact-match keyword in the first words of your title.

Title = [Money keyword, exact form] + [benefit/variant words a human would still say out loud]Front-loaded, exact, and still readable — never a keyword dump

Everything else supports: backend search terms carry the misspellings, synonyms, and long-tail variations your title shouldn't waste characters on. Bullets answer buyer questions in natural language. The common failure isn't ignorance — it's decoration: a clever title that never says what the product is. The algorithm can't rank what it can't read.

💡 Tip. Test your title like a shopper: in a grid of twenty competitors, does a stranger know what you sell, and for whom, in two seconds? If the answer needs a second line, the money keyword isn't front-loaded enough.

Part 4

Step 3 — engineer your first 15 reviews

Early conversion decides everything after it — and nothing moves early conversion like visible proof. Fifteen real reviews change the math: a listing that converts at 2% converts at two or three times that once shoppers see people who bought before them and stayed happy.

Engineered doesn't mean manufactured. It means systematized: the ask arrives at peak satisfaction — the moment of use, not the moment of delivery — through Amazon's own tools, Vine for the first units, a compliant insert that invites honest feedback, and a follow-up sequence that doesn't beg. Reviews that answer the top two objections in a buyer's own words are worth more than a hundred generic stars.

🔑 Key insight. Your first 15 reviews are a conversion asset, not a vanity metric. They set the ceiling on every campaign you run afterward — bids that would be profitable at 4.5 stars are donations at 3.8.

Part 5

Step 4 — isolate the money keyword

When the listing can convert, the money keyword gets its own campaign — one keyword, one budget, one unambiguous set of numbers. Bidding toward Top-3 placements on that term, and watching conversion hold under the higher traffic, is the test that decides everything.

1. Isolate.One campaign for the money keyword. Negate it everywhere else so its traffic and data stay pure.
2. Bid to position.Calibrate to Top-3 economics: what a click is worth at your conversion rate, times your target — never a rupee more.
3. Watch conversion hold.Higher placement brings more traffic; if conversion holds, the position deserves the aggression. If it slides, the listing — not the bid — is the work.
4. Then add aggression.Only after the hold. Adding spend to slipping conversion is paying the algorithm to learn the wrong thing about you.

Part 6

Steps 5–6 — sync to inventory, then defend

Step 5: bids follow inventory. Deep stock supports aggressive bids; thin stock means throttle down before the world notices. The algorithm resets rank on a stockout with brutal indifference — and the rebuild always costs more than the discipline would have.

Step 6: defend for 14 days. Rank compounds if you hold it — the flywheel (position → sales → stronger position) needs time to spin without interruption. Judge nothing before the window closes: day-4 position wobbles are noise; day-14 position is a fact. In our client work, Top-3 pushes are judged on 90-day windows for exactly this reason — it's long enough for the compounding to show, short enough to steer.

💡 Tip. The defence window is also your diagnostic window. If position oscillates between 3 and 7, that's the algorithm testing you — conversion holding steady through the wobble is what settles the test in your favour.

Part 7

The cadence that holds it

Getting to Top-3 is a project; staying there is a cadence:

RhythmThe check
DailyPosition on the money keyword · stock cover · Buy Box status
WeeklyConversion on the isolated campaign · search terms to negate · bid drift vs. position
Every 14 daysThe defence verdict — hold, push, or diagnose the listing
QuarterlyThe 90-day picture: rank trend, organic share of sales, next keyword to isolate

Run properly, this system has taken a client from launch to Top 5 on Amazon India in 60 days — not because of one clever move, but because every step was done in order and held long enough to compound. That's the whole difference between the brands that rank and the brands that rent. If you want the full Top-3 System run on your account, the first diagnosis is free.

Part 8

The math underneath the six steps

You can run the six steps on faith and they'll still work. But knowing the mechanics underneath makes every judgement call easier — because Amazon's ranking decision, stripped to its bones, is:

Ranking score = Performance × RelevancyPerformance = conversion, velocity, engagement. Relevancy = how exactly your listing matches the search

Step 2 (the front-loaded keyword) is the relevancy multiplier — the cheapest score to move, because it's pure editing. Steps 3–4 (reviews, isolated bidding, conversion that holds) are the performance side — more expensive, but compounding. A weak listing with aggressive bids scores high × low and loses to a modest performer with perfect relevancy — which is why the steps run in this order: relevancy first, then performance on top of it.

And the memory: the algorithm keeps multiple lookback windows — days, weeks, months — with recent history weighted heaviest. That's why Step 6 exists: fourteen days of holding a position writes a consistent record into the windows the algorithm actually reads. One good day is a data point; fourteen is a pattern.

🔑 Key insight. The six steps are one sentence in disguise: make the listing match the search exactly (relevancy), make the people who arrive buy (performance), and keep both true long enough for the algorithm to believe it (the windows).

Part 9

Amazon India specifics that change the playbook

The six steps are universal; the weighting inside them is not. On Amazon India, four market realities bend the playbook:

RealityWhat it changes
Mobile-first shoppingThe first fold is a phone screen: the main image and the first line of the title do nearly all the work. If the money keyword doesn't survive truncation on mobile, it isn't front-loaded — it's buried
Festive-quarter spikesThe big festive-sale window concentrates a frightening share of annual demand. Stock depth (Step 1) and bid-sync discipline (Step 5) matter more in the six weeks before it than in the six months after
Price-aware buyersCoupons and offers are read closely; the price-plus-proof cluster on the results screen decides the click. Premium pricing works — with premium proof visible at thumbnail size
Transliterated searchPlenty of category terms arrive in romanized Hindi or Hinglish mixes. These live in backend terms (Step 2's supporting cast) — often untracked by tools, and quietly won by the listings that bother

None of this replaces the system — it just tells you where to spend extra care when your market is India. The fundamentals still come first, the keyword still goes first, and the defence window still closes before you judge. Run it with the local weighting, hold it for ninety days, and the position stops being a rental. That's the whole system — and if you want it run on your ASIN, the first diagnosis is free.

Want this run on your account instead? The first diagnosis is free.

Book Free Consultation

New guides, straight to your inbox.

Practical D2C playbooks as we publish them. No fluff, no spam — unsubscribe anytime.