Basket size up 39 per cent on the same platform.
A Jeddah pharmacy with an online store and a growing wholesale side. Over two years the average order rose from SAR 282 to SAR 392 on the same Salla platform. July 2026 shows how: sales, search, a five-month content experiment and a feed fault that had hidden 95 per cent of the catalogue.
Period: July 2026, one month inside an ongoing retainer. Every figure is taken from the client report and labelled with its source. No figure has been rounded upward.
The client
Ghaydaa Medical is a Jeddah pharmacy with an online store on Salla, shipping across Saudi Arabia and the GCC, and a growing trade in professional brands to salons and clinics. Cush is its marketing department: ecommerce, paid media, search, content, social and the technical plumbing between them.
Where they started
An average order that moved with promotions, and a catalogue of 8,244 products that Google Shopping had quietly stopped showing. Content effort spread thin across products without evidence of what it returned.
What was wrong
The store had a pricing feed quoting riyals while the platform still had the US dollar and twenty other currencies switched on. Google saw mismatched prices and disapproved 7,848 products. Nobody had noticed, because nothing looked broken.
What we did
The Merchant Center currency fault identified and fixed on 24 July, with the foreign currencies switched off and the GCC set retained.
A five-month experiment across forty products: twenty with no copy received it, twenty with existing copy had it rewritten. Views were tracked monthly from March to July.
Six product campaigns, duplicate product pages merged so the brand stopped competing with itself in search, a pricing audit of 71 professional SKUs against noon.com and se7rek.com, and a landing-page structure for the Kaypro range.
What changed
Gross sales SAR 110,778, up 2.8 per cent on June and the second-highest month of the year. Orders fell 15 per cent while average order value rose 20.6 per cent to SAR 486.5, so fewer, larger baskets. Sessions rose 34.2 per cent to 86,930. YouTube delivered 72,116 views. The content experiment gave an operational answer: products given copy for the first time rose 38.4 per cent in views; products whose copy was rewritten fell 14 per cent. August's writing time goes where the return is.
The evidence
Every number on this page carries one of three labels, and the period it covers is stated above the figures.
Measure the boring things. That is where the money was hiding.
The sector was pharmacy retail. The moves were a feed fault found by someone paid to look, and a content experiment that told the team where to stop. Every ecommerce brand has both waiting.
Read the next case, or find out which pattern your marketing fits.
DTS Arabia: From a cold email to a live session with 181 registrations in twelve days, and a median watch time of 44 minutes.