Glossary/GMV Max

What is GMV Max?

GMV Max is an automated marketplace campaign type in which the platform takes over keyword selection, audience targeting, placement, and bid management, and optimizes toward gross merchandise value under the budget and ROAS target the seller sets. Shopee runs it under this name, and TikTok Shop offers an equivalently named automated campaign; in both cases the seller trades granular control for the platform’s conversion model. The objective is in the name: GMV Max maximises gross merchandise value, which is revenue before any seller cost, so the only lever that connects it to profit is the Target ROAS the seller supplies.

01/Formula

Example

A GMV Max campaign at Target ROAS 5, budget ฿20,000/day, on a catalogue
averaging 18% contribution before ads.

Gross return at target = ฿100,000/day of attributed GMV
True return = 5 × 0.18 = ฿0.90 per ฿1 of spend
Daily contribution after ads = 20,000 × (0.90 − 1) = −฿2,000

The campaign hits its target exactly, and loses ฿60,000 a month.

02/In detail

What do you still control in GMV Max?

Four things, and they carry the entire profit outcome. The budget sets how much the optimizer may spend. The Target ROAS sets the efficiency bar it must clear, and is the only input in the whole campaign that has any relationship to your costs. The product selection decides which contribution rates are in scope — the single most under-used control, because a mixed-margin product set forces one target onto products that need very different ones. And the creative and listing quality determine what the optimizer has to work with. Everything else, including which keywords it buys and what it pays for them, is delegated.

Is GMV Max better than manual keyword ads?

It is better at the job it is given, which is finding volume under a gross-revenue constraint. Automated bidders re-price every auction against a live conversion forecast, and no manual bid schedule matches that. What they cannot do is see your cost structure: the optimizer does not know that one SKU in the campaign carries 40% contribution and another 12%, so it will happily shift spend toward whichever converts cheapest in revenue terms, which is often the thin one. The practical answer for most sellers is neither-or-both: automated campaigns for scale on product groups with homogeneous margins, manual keyword ads where a specific term is worth defending or excluding.

How should you group products inside GMV Max?

By contribution rate, not by category, brand, or price. Since a single Target ROAS applies to everything inside the campaign, every product in it should share roughly the same break-even ROAS, or the target is wrong for most of them. A workable pattern is three to four campaigns split on contribution bands — for example above 35%, 25–35%, 15–25%, and below 15% — each with its own target derived from that band. This is unglamorous and it is usually the largest single improvement available to a shop running one big automated campaign across the whole catalogue.

03/Why it matters

The trap, in one paragraph.

Automation moves the seller’s job from operating the campaign to specifying its objective, and most of the value now sits in that specification. A GMV Max campaign will hit whatever target it is given with more consistency than a human bidder, which is exactly why a target set below break-even is more dangerous under automation than it was under manual bidding: the system will pursue the loss efficiently and at full budget.

Common mistake

Judging a GMV Max campaign by whether it hit its Target ROAS. Hitting the target only proves the optimizer worked. Whether the campaign made money depends on whether the target was above the break-even ROAS of the products inside it — a question the platform never asks, because it cannot see your COGS.

04/In DataGlass

How GMV Max is used in DataGlass.

DataGlass reconstructs contribution per SKU from Shopee, Lazada, and TikTok Shop order data, so the products inside an automated campaign can be grouped by real contribution rate rather than by category, and the configured Target ROAS can be compared against the break-even ROAS of the actual product mix the campaign has been selling.

05/Sources

  1. [1]
    Shopee Thailand — Seller Education Hub, campaign metrics and automated campaigns

    Shopee seller documentation on automated campaign types, reported ROAS as attributed revenue over ad spend, and the learning period automated campaigns run before performance stabilises.

  2. [2]
    TikTok for Business — Help Centre

    TikTok Ads documentation of value-based and GMV-objective automated campaigns and their learning phase after budget or target changes.

Stop guessing. Start deploying.

Join the sellers using DataGlass to turn shop data into the next profit-maximizing action.