Marketplaces make the sale.Your brand makes the choice.
Sarthi is the old word for the one who drives the chariot — the one who holds the reins so the warrior is free to fight. It is the job we named ourselves after, and the only job we have ever done.
Ten years in
Brands built from zero to crore scale
Ratings held at 4★ or better on every account we run
Agra
Why we are called Sarthi
On the field of Kurukshetra, the greatest archer alive did not drive his own chariot. His sarthi held the reins — choosing the ground, steadying the horses through the chaos, saying the quiet word at the right moment. He never lifted a weapon. And nobody doubts who carried that war.
That is the job we took our name from. The product, the ambition, the name on the pack — those are yours. We hold the reins: the platforms, the listings, the seasons, the daily unglamorous care of your shelf.
You fight. We drive.
What being chosen looks like
Getting found is table stakes. Our work is the quiet thread that makes yours the one they pick.
Case files · names and numbers published with each brand's permission
These are running accounts, not testimonials. The figures are as reported on each brand's own seller account. Ranges are ranges, not claims rounded upward.
Sterling silver jewellery · built from scratch
Listed on Amazon unsure it would work at all. Built from nothing, then taken to the other marketplaces once the shelf held. Now selling over a crore a year online.
Sterling silver · built from scratch
The client did not want to sell online, and at fifty thousand a month it was hard to argue. Then one Rakhi season did four to five lakh, and the plan changed.
Groceries, colours, rangoli, pooja, home
An established offline house, 250 to 300 SKUs, B2B and B2C. The big seasonal number is B2B led; our patch is the B2C shelf, kept healthy all year so the brand is present when its enormous season arrives.
Kids apparel · where it started
The first brand, ten years ago. We made the early mistakes here — process, manpower, Amazon itself — and in the end we parted ways, because we could not meet their expectations then. It stays on this page because everything we now know how to prevent, we first paid for here.
Silver Chest and Satkoti both make sterling silver. They know about each other, and their accounts are run by separate hands with separate keyword work. If you compete with a brand we run, we tell you before you sign, not after.
Two reins, one harness
ListingThe catalogue is the shop floor. Titles, images, bullets, variants, compliance.
SearchRanking for what the category actually types, not what the brand calls itself.
AdsBids built around what converts, and switching off what does not.
Buy boxPrice, stock, fulfilment and rating, watched daily.
Account healthThe quiet one. It decides whether any of the above is allowed to run at all.
Positioning, story, socialThe newer rein. The team for it sits in-house. It is drawn open until we have a preference number to show you, because we would rather show that than hide it.
Why the ratings hold
Six people and a leader. Small enough that your account is never a ticket in a queue; senior enough that the person answering has filed the claims, run the listings and sat in the client’s chair before ever managing yours.
One runs Tota, one runs the silver brands. Your account has a name on it, not a ticket queue.
Graphics and design for listings and campaigns, done at the same table as the people who run the shelf.
For brands that need it we run the physical side too: dispatch, returns, and the stock claims nobody enjoys filing.
Multi-SKU brands get a second pair of hands for content and the social calendar.
Rachi plans the growth and runs the advertising herself. Seven to eight years of working directly with the Amazon team sits in this seat.
Listing videos, dispatch-proof videos and social cuts, made in-house next to the people who run the shelf.
The ceiling, said out loud
Four to five brands. That is the honest capacity of this table today, and we say the number out loud because pretending otherwise is how ratings slip. We grow our own account managers — claims first, then orders, then listings, then the client’s chair — which is slow, and is why the quality holds. We take accounts slowly and we vet who we work with; we have been burned once, and it made us careful.
A note from the founder
Ten years ago we put our first brand on Amazon and made every mistake there was to make. Nobody taught us; the marketplace did. Since then we have taken brands from their first listing to more than a crore a year — and the thing I am proudest of is not any number. It is that every account we have ever run still holds its rating. That only happens when someone cares for a brand the way its owner would.
Engine one, running for ten years
Titles, images, A plus content, variant structure, and the category rules that decide whether a listing stays live.
Keywords taken from what the category types, backend terms, and the slow job of earning rank instead of renting it.
Campaign structure, bids, negatives, and the discipline of switching off what does not pay.
Price, stock, fulfilment and rating watched daily, because losing it quietly costs more than losing it loudly.
Policy warnings, suppressions and appeals. Boring until it is the only thing that matters.
Platform dashboards disagree with each other. When they do, the report says which number we used and why.
Amazon, Flipkart, Blinkit, FirstCry, Myntra, JioMart. Each with its own rules and its own way of breaking.
Rakhi, Diwali, Holi, the wedding months. Half this business is being ready in the quiet months for the loud one.
Engine two, newer
Positioning, story and social. The reason a customer types your brand name instead of your category, and the reason a cheaper listing one slot above you does not take them.
What we are not claiming
The content desk for this sits in-house and runs it daily. What we do not yet have is a preference number that moved and a client's permission to show it. When we do, it goes here with the dates attached. Until then this rein stays drawn open.
What a bad Tuesday looks like
These are conditions we manage on live accounts. No ranking positions or percentages are simulated here.
Before you contact anyone
Put in your own last thirty days. Every chip is one percent of your gross sales. Chips peel off at each cost. This is arithmetic, not a case study.
Product cost is not a leak. It is your product. The three costs above it are the ones marketplace work can actually move, and that is the number worth staring at.
ACOS is ad spend against ad attributed sales. TACOS is ad spend against all sales. Your overheads are not in any of these numbers.
Worth saying out loud
Most of what damages a young brand on a marketplace is something an agency agreed to do because the client asked for it. Ten years taught us which requests to refuse.
Fake reviews and rating manipulationIt works until the platform notices, and then it takes the account with it. Our ratings are slower and real, which is why they survive audits.
Inflated MRP so the discount looks biggerCustomers price check across six apps. The theatre is visible from all of them.
Spending into a broken listingIf the images, title and price are not fixed first, advertising only buys traffic for a page that cannot close.
Reporting a number we cannot reconcileWhen two dashboards disagree, the report says which one we used and why.
Telling you that you need brand work when you do notSometimes the honest answer is that the shelf is fine and the problem is the product or the price.
The deal, plainly
Sarthi works on commission. Our revenue is a share of what your shelf actually sells — if the month is bad, it is bad for both of us. Ten years in, we have not found a faster way to keep an agency honest.
Manufacturing, inventory, dispatch planning. Nobody knows your product like you do, and we never pretend to.
Listings, search, advertising, buy box, account health, claims and returns. The daily unglamorous care of the shelf.
Which product to push, at what price, with how much stock. Most agencies sell this as consulting. Here it is simply part of holding the reins.
Clients describe it in different ways. Rachi’s phrase for it is “ease of business” — you build the product, and stop carrying the marketplace in your head.
A launch, honestly: we need the product’s basics, its costing, and the product itself in hand before we begin. Fifteen to twenty days from signing to selling in an open category, and we plan a month. Around a third of the work is the photoshoot. The hardest stretch is the first ten to fifteen orders — after that, a good product starts carrying itself.
How it starts
We open your listings the way a customer does, on every platform you sell on, and write down what is losing you orders. You keep that document whether or not you hire us.
Fix the listing and the account health, then rebuild the advertising around what converts. Engine one only. Brand work does not start on a leaking shelf.
Once the shelf holds without being propped up by spend, we start on the reason someone chooses you by name.
How to reach us
Pick whichever is costing you the most this week. Two fields, then it reaches Rachi with the problem already written at the top.
Nothing is stored on this page. The button opens your own WhatsApp or mail app with the message already written.