A UAE Hydration Brand: Amazon and Noon Needed Three Different Growth Decisions

One brand. Three marketplace environments in the UAE: Amazon UAE, Noon Core, Noon Minutes.

It would have been easy to put all three into one advertising report and ask: "Did ROAS improve?" But that would have missed what was actually happening.

Amazon needed to grow with less dependence on paid traffic. Noon Core needed better campaign structure and tighter control of waste. Noon Minutes had a different issue — order volume was holding up, but the sales mix was shifting towards lower-value trial packs.

Three channels. Three different jobs. Three different decisions.

The key takeaway

A multi-marketplace account should not be managed against one blended advertising metric. Between Mar–May and Jun–Aug 2026, each channel moved differently — and each needed a different intervention.

• AmazonRevenue +9.1% while advertising spend fell 15.8%.
• Noon CoreRevenue +8.5% while advertising efficiency more than doubled.
• Noon MinutesRevenue −12.9% despite advertising conversion improving. The problem was purchase mix.

At account level, the quarter looked almost flat after advertising

Revenue after advertising is not profit. It excludes product cost, marketplace fees and other costs. The account did not have one simple growth story — Amazon and Noon Core improved while Noon Minutes weakened. The channel view showed where.

The first decision was to stop treating every channel the same

The review separated the business into three operating roles, each judged by different measures.

• Amazon UAEGrow with less paid dependence. Measure: total revenue + TACoS.
• Noon CoreTurn advertising efficiency into growth. Measure: revenue + ROAS + cost per ad order.
• Noon MinutesProtect volume, improve purchase value. Measure: packs sold + revenue per pack.

That is more useful than ranking all three channels by one blended ROAS — because each channel was solving a different commercial problem.

Amazon UAE — more sales with less advertising

Amazon delivered the clearest positive business movement during the quarter.

Fewer visits. More units sold.

Amazon sessions fell by approximately 21% — from 11,780 to 9,347. Yet units sold increased by approximately 23%. Store conversion moved from 15.4% to 23.8%. The quarter was not driven by buying more traffic. The available traffic converted better. At the same time, estimated organic revenue increased from approximately AED 100.7K to AED 120.5K, with estimated organic share moving from 52% to 57%.

"Estimated organic revenue" is calculated as total revenue minus ad-attributed sales. It is a directional operating measure — not proof that advertising caused incremental organic growth.

Brand advertising became cheaper, but brand dependence was not solved

From July to August, brand-search advertising spend fell from AED 2,364 to AED 1,262 — approximately 47%. Brand clicks fell by only 6%. Brand CPC improved from AED 6.44 to AED 3.67. The account was paying materially less for much of the same branded traffic. But the conclusion should not be that brand dependence was fixed — brand-led demand still represented a meaningful part of advertising sales.

What changed on Amazon? The account reduced its overall advertising burden, lowered branded CPC later in the period, and saw stronger store conversion particularly around the hero product. Listing and media changes also preceded the stronger August conversion result — but the data does not isolate their individual effect.

The useful operating result is simpler: Amazon produced more total revenue while requiring less advertising spend. The next job is to extend that without allowing paid or branded dependence to creep back up.

Noon Core — better structure changed the advertising decision

Noon Core had a different starting problem. The account was heavily dependent on one broad automatic campaign. The quarter moved towards a more structured portfolio with separate decisions around brand defence, generic keywords, competitor activity, category targeting and product campaigns. The objective was not simply to create more campaigns. It was to stop treating every type of demand as though it deserved the same investment.

Revenue grew. Reported advertising spend reduced substantially. ROAS more than doubled. Cost per advertising order more than halved.

The more interesting change happened inside the campaigns

The team did not simply lower every bid. Weak areas and stronger areas were treated differently.

So the decision was not spend less everywhere. It was spend less where the evidence is weak and keep room to invest where the evidence is stronger.

ROAS did improve — from 1.53x to 3.54x. But the operational story underneath it is more useful. The account moved from a broad campaign structure towards more separate investment choices. That gives the operator something to repeat: not the exact number, but the decision process.

Noon Minutes — better conversion did not mean better business

Noon Minutes told a very different story. This is exactly the kind of channel that can be misread if you look only at advertising metrics.

Advertising conversion improved. Pack volume was broadly flat. Yet revenue fell by approximately AED 46.1K. Why? The sales mix changed.

Trial packs took a much larger share of volume

Trial-pack share of volume moved from approximately 47% to 62%. Customers were still buying — but more of the volume was coming from lower-value packs. More orders were not automatically more valuable orders.

If the team looked only at 23.4% advertising conversion, the channel could appear healthier. But the business question was: what are customers actually buying after they convert? That led to a different next-quarter priority — not get more clicks, not simply improve conversion, but: rebuild larger-pack contribution to the sales mix.

Availability and visibility also deserved attention

The review found two additional signals worth investigating. First, the recorded store footprint for the 30-pack narrowed materially after early August — enough to justify reviewing replenishment across dark stores. Central warehouse inventory does not automatically mean a customer can buy the product from the location serving them. Second, a category exclusion had restricted visibility in the Beverages category. That exclusion was removed on 16 September, after the quarter being analysed. The correct next step was to restore visibility, improve local availability where needed, and measure what happens next. The data does not establish that this change caused a recovery.

One account. Three different next moves.

By the end of the review, each channel had a different priority.

All three channels were ultimately working towards more total sales with a sustainable advertising burden. But they did not need the same intervention to get there. That is the point.

What this case does not claim

• Total marketplace revenue increased — it did not. Account-level revenue declined 3.2%.
• Profit improved — the source does not provide full cost economics beyond advertising.
• Amazon's conversion improvement was caused entirely by listing or media changes — correlation is noted, not causation.
• Inventory caused the Noon Minutes decline — the data identifies an availability issue worth investigating, not a proven full cause.
• Restoring category visibility has already produced recovery — there was insufficient time to establish this.
• Isolated incrementality was proven for any channel — these are observed movements, not controlled experiments.

Frequently asked questions