GMV Max for fashion: weekly tracking, ROI 7.69 and 266.7 million VND in revenue in 7 days

Context & challenge

This is a fashion store on TikTok Shop that runs GMV Max steadily. The brand name is withheld by agreement.

Fashion has a particular trait: styles change fast. A video can sell very well for a few days and then “fatigue”. When a style runs out of sizes, orders from that video drop too. So looking at reports monthly is usually too slow.

The store’s challenge:

• Track GMV Max performance weekly to spot fatigued videos and out-of-stock styles early.
• Keep cost per order stable while still raising budget slightly.
• Grow revenue week over week, steadily.

Diagnosis from the data

GMV Max dashboard for 02/10 – 09/10/2026, compared with the previous 7 days:

• Spend: 34,674,136 VND (+9.76%).
• SKU orders: 1,958 (+14.50%).
• Cost per order: 17,709 VND (−4.14%).
• Revenue: 266,671,920 VND (+15.26%).
• ROI: 7.69 (+5.05%).

The previous period worked out backwards (estimates, calculated from the dashboard data):

The good news: this was a “healthy growth” week. Spend went up, but orders and revenue grew faster than spend. As a result, cost per order fell and ROI rose at the same time. This is the ideal state when running ads.

What to watch: the chart on the dashboard shows spend and orders going up and down by day, with two peaks around 04/10 and 07/10. 09/10 on the chart is clearly lower. This may be because the last day’s data was not yet complete when the screenshot was taken, so it should be rechecked at the end of the day before drawing conclusions.

Growth strategy

This is the framework LDH Media applies to fashion stores running GMV Max on a weekly rhythm.

Phase 1 – Foundation (prepare stock and content)

• Check inventory by size and colour for the styles being advertised.
• Prepare try-on and styling videos for each hero style.
• Set a break-even ROI for each product group, since fashion margins vary by line.

Phase 2 – Activation (read the numbers weekly)

• Each week, compare spend, orders, cost per order and ROI with the previous week.
• Separate the videos that generate orders from those that spend money with few orders.
• Replace fatigued videos with new videos of the same style, or new styles in the same look.
• Check styles about to run out of sizes and lower their priority before they sell out.

Phase 3 – Scale (raise gently, raise steadily)

• Only raise budget when the previous week’s cost per order was stable or falling.
• Raise moderately — this week, spend rose only 9.76%.
• Align ads with the new-style release calendar and the platform’s sale events.

Results

For comparison within fashion, here is the 30-day GMV Max dashboard of F-Basic Việt Nam, another fashion case LDH Media has published:

The two stores are in the same category, but their order values are very different: ≈ 136,200 VND for this store and ≈ 677,800 VND for F-Basic (calculated from the dashboard data). A higher order value lifts ROI even with a higher cost per order.

Analysis: where the growth came from

The GMV Max dashboard has no visitor count, so revenue is broken down as:

Revenue = Spend × (Orders per unit of spend) × Order value

Calculated from the dashboard data:

• Spend rose 9.76%.
• Order efficiency rose about 4.3%, as cost per order fell 4.14%.
• Order value rose about 0.7%.

Multiplied together: 1.0976 × 1.043 × 1.007 ≈ 1.153, matching the revenue increase of +15.26%.

So about two-thirds of the increase came from extra spend. The rest came from ads generating cheaper orders and slightly higher-value orders.

This week’s marginal ROI was very good (estimated from the dashboard):

• Extra spend ≈ 3.1 million VND.
• Extra revenue ≈ 35.3 million VND.
• Marginal ROI ≈ 11.5, higher than the average ROI of 7.69.

When marginal ROI is higher than average ROI, it signals that budget can be raised further — as long as there is stock and there are good videos.

The weaker points:

• One week is too short a window for long-term conclusions. Weekly results can be skewed by one trending style or one sale day. The 30-day window should also be reviewed.
• Order value is low compared with many other fashion stores. It can be improved with top + bottom bundles or buy-2 offers.
• Daily fluctuation is still large. The two peaks in the week show revenue depends on a few good days. More videos are needed to even it out.

Lessons for store owners

• Fashion stores should read the numbers weekly. Videos and styles change fast; waiting until month-end is too late.
• A good week is one where orders grow faster than spend. Then cost per order falls and ROI rises at the same time.
• Calculate marginal ROI before raising budget. If marginal ROI is still high, there is room to grow.
• Check stock by size and colour before scaling. Running out of sizes mid-way wastes what the algorithm has learned.
• Don’t compare ROI between fashion stores with different price points. Compare against your own break-even ROI.

Actual results depend on product, price, inventory, budget and platform policies. This is not a commitment to results for every client.