GMV Max budget: how much to spend and when to scale

Start GMV Max in the $50 to $150 a day band per hero product, leave it alone for 3 to 4 days while it stabilizes, judge it on 5 to 6 days of data, then raise budget 20 to 30 percent at a time for as long as the return holds above your floor. That is the whole system in one sentence, and it is the one we run across 60+ client shops. The rest of this guide is the reasoning: the readiness gate that comes before your first ad dollar, how to size a starting budget for your margins, how to set the return floor, and the signals that say scale, hold, or cut. If you are still deciding what GMV Max even is, start with our explainer, then come back.

What GMV Max does with your budget

GMV Max is TikTok's automated campaign type for Shop sellers. You give it a product, a budget or a target return, and it assembles the rest: which videos to amplify, which audiences to hit, and how to split spend across the feed, search, and the Shop tab. That automation is why budget strategy matters more than targeting strategy here. You cannot out tweak the machine. You can only control what you feed it, how much runway you give it, and when you add fuel. For how it compares to running Spark Ads yourself, see GMV Max vs Spark Ads.

The gate before your first dollar

Ads are fuel, not a starter. GMV Max amplifies the content attached to your product, so the state of that content decides your return before you spend anything.

• 20 to 30 plus creator videos live. This is the readiness rule we hold every client shop to. The system needs a library of proven angles to pull from, and you need organic signal telling you which videos deserve amplification. Building that library is the job of our affiliate and creator program.
• 65 percent margin or better after fees. Platform fees of roughly 6 to 8 percent, payment processing, and creator commission of 15 to 20 percent all come out before ad spend does. Thin margins turn a working campaign into a treadmill.
• At least one video already selling organically. If nothing converts without paid push, the ad account is about to teach you an expensive lesson that a sample batch would have taught you cheaper.

How to size the starting budget

The starting budget has one job: buy readable data without buying it at scale. Too small and the campaign never exits learning, so every number you see is noise. Too big and you find out about a weak angle after a very expensive week.

• $50 to $150 a day per hero product is the band we start client shops in. Impulse priced products in the $15 to $35 range sit at the lower end. Higher priced products need the higher end so the campaign books enough orders per day to read.
• One hero product first, not the catalog. Spreading $100 across 5 products gives you 5 unreadable campaigns. Concentrate, learn, then expand to the next SKU with what you learned.
• Fund a full week before you judge. Whatever daily number you pick, be prepared to hold it for 7 days. A budget you would panic cut on day 3 is too big.

Set the return floor from your margin, not from hope

Your return floor is the ROAS below which an order loses money, and it comes straight from the cost stack. Work it once and write it down: product cost, platform fees, processing, creator commission if the attributed video carries one, and shipping. On a 65 percent margin product that math typically lands the breakeven floor around 1.5x to 2x, which is why we aim campaigns comfortably above it. Across the US brands we run, the ceiling is real: one food brand did $58,173 in gross sales in a single 30 day window at a 4.01x return on ad spend, with a peak 19 day stretch at 4.72x across 801 orders. A wellness brand generated $13,129 in revenue on $2,600 of spend in one week, a 5.05x return across 314 orders. Those numbers came from proven creator content plus disciplined scaling, not from a clever launch day setting.

The stabilization rules

• Days 1 to 4: hands off. GMV Max recalibrates every time you touch budget or target, so early edits reset the learning you are paying for. Let it stabilize for 3 to 4 days.
• Days 5 to 6: judge. Now the numbers mean something. Compare the campaign's return to your floor across the full window, not the best single day.
• One change at a time. If you adjust budget and target ROI together, you will never know which change did what. Move one lever, wait, read, repeat.

When to scale, and by how much

Scale when the campaign holds above your floor across 5 to 6 consecutive days, not because one day spiked. Raise budget 20 to 30 percent at a time, then let the campaign restabilize before the next step. Doubling budget overnight is the classic way to break a working campaign: the system chases volume it has not learned to buy efficiently yet, the return craters, and the seller concludes ads do not work. Step scaling feels slow in week one. By week 6 it is the reason the campaign still works at several times the spend.

When to cut, and what to fix first

• Below the floor for 3 plus days after stabilizing: drop budget back a step or pause. Do not negotiate with a losing campaign.
• Return decaying while spend is flat: creative fatigue. The fix is new creator videos, not a budget change. Feed the campaign fresh angles and it usually recovers. What to brief is covered in what content sells on TikTok Shop.
• Great return that will not spend: your target is set too aggressively. Loosen the target ROI slightly so the system has room to buy volume.

The pattern behind all three: when GMV Max underperforms, it is almost always a content or offer problem wearing an ads costume. Fix the input before you touch the dial.

A note on Q4 budgets

From late October, auction costs climb toward Black Friday. The playbook does not change, the calendar does: campaigns need to be live and stable by mid October so November is pure step scaling into demand you already validated. The full season plan is in our Q4 strategy playbook and the weekend execution in the Black Friday guide.

Quick answers

How much should I spend on GMV Max per day?

Start in the $50 to $150 a day band per hero product. That is enough for the system to exit learning and produce readable data within the first week without buying expensive lessons at scale. Once a campaign holds its return target across 5 to 6 days of data, raise budget 20 to 30 percent at a time. Proven campaigns on client shops often run several hundred dollars a day, but they earned that level step by step.

How long before GMV Max is profitable?

Give a new campaign 3 to 4 days to stabilize without touching it, then judge on 5 to 6 days of data. Campaigns built on creator videos that were already selling organically usually read profitable inside that first week. If a campaign is still below your return floor after a full week, the problem is almost always the content or the offer, not the budget.

Can I run GMV Max without creator videos?

You can, but you should not. GMV Max pulls from the content attached to your product, so with no creator videos it amplifies a weak library. The readiness rule we use across client shops is 20 to 30 plus creator videos live before the first ad dollar. Ads amplify proof. They do not create it.

Run it in house, or have it run for you

Budget discipline is simple to describe and hard to hold, especially the part where you do nothing for 4 days while money spends. If you would rather have the campaigns run by a team that manages this daily across 60+ brands, that is our ads and GMV Max service, with the creator engine behind it and the client results in the open.