Arabic Parfums

Client Overview

Arabic Parfums sells Middle Eastern fragrance (oud, attars, and designer-alternative perfumes) to customers across Europe. It is a catalogue business in a category where the buying decision is driven by scent family and brand familiarity rather than specification, and where a large share of revenue comes from customers who reorder.

ADFIX runs the brand’s paid media across search and social, and its retention programme: email, SMS and WhatsApp flows and campaigns. Search came first.

Challenge

Arabic Parfums came to ADFIX running at small scale and on autopilot. Search ran on a low budget that nobody was actively steering, and the tracking behind it recorded orders without recording what those orders were worth. On paper every campaign looked identical: a sale for a €15 sample counted the same as a sale for a €200 bottle.

In a catalogue with that much price spread, autopilot bidding does real damage. It optimises toward whatever converts most often, which in fragrance is almost always the cheapest item. With no way to prove what spend returned, the budget stayed small.

Retention was in the same state. The brand had an email list, but no automated flow was producing revenue and campaigns went out only occasionally. Repeat customers are a large share of this business, and nothing was working to bring them back.

Nothing could be scaled until the measurement was true and the basics were switched on.

Strategy

Rebuild the measurement

• Implement true purchase-value tracking so that revenue, not order count, became the optimisation target.
• Re-align analytics and ad platform conversion data so store performance and channel performance could be read against each other.
• Establish a clean baseline before increasing spend, so that later growth could be attributed rather than assumed.

Then scale search

• Move bidding onto value-based targets now that value was visible.
• Restructure the product feed around scent family and price tier, so higher-value lines could be bid on separately from entry-price ones.
• Increase budget in steps, holding the return threshold at each one before moving up.

Then add social

• Launch Meta as a second acquisition channel once search efficiency was stable and proven.
• Lead with product-led creative suited to a discovery-driven category, where shoppers browse by scent profile rather than search a brand by name.

Then build retention

• Rebuild the email programme in Klaviyo: welcome series, abandoned checkout, browse abandonment, post-purchase and back-in-stock flows.
• Warm up sending reputation gradually, widening campaign reach step by step instead of mailing the whole list at once.
• Add WhatsApp and SMS alongside email, starting with abandoned checkout reminders, order confirmations and post-delivery review requests.
• Write campaigns around how people shop fragrance: scent comparisons, similar-scent recommendations and single-fragrance features.

Execution

Search

Once value tracking was live, the account could be scaled on evidence. Average monthly search budget grew roughly sixfold over the following eighteen months, in steps rather than in one move, with the return threshold re-tested at each increase.

The return did not degrade under that scale. Comparing April to September against the same six months a year earlier, spend rose 97% while revenue from search rose 120%. ROAS improved 11% while the channel doubled in size. Average order value from paid search rose 40% over the same comparison, which is the direct consequence of bidding toward value instead of volume.

Social

Meta launched in January 2026 and has returned 9.6x since. As expected for a new account, the opening quarter ran well below that average while the algorithm learned the catalogue and creative was tested. Efficiency improved substantially through the year as both settled.

Retention

ADFIX took over email, SMS and WhatsApp in mid-July 2026. In the quarter before the handover, automated flows produced no attributed revenue, and no campaign had gone out since November 2025.

The new flows went live in July. Campaigns started with a warm-up sequence sent to a small engaged segment. By late September a single campaign reached 22 times as many people as the first warm-up send, and campaign revenue per recipient was 51% higher than on the brand’s last campaigns before the handover. The welcome series alone produces 60% of flow revenue, which shows how many first-time buyers paid media was sending to a store with no follow-up.

Store

The combined effect is visible at store level, not only inside the ad platforms. From April to September, sessions grew 113%, orders doubled and revenue rose 140% against the same months last year, with average order value up 20%.

Results

Store performance

April–September 2026 vs April–September 2025

Paid search

April–September 2026 vs April–September 2025

Google Ads returned 11.2x across the last twelve months. Meta has returned 9.6x since launch in January 2026.

Retention: email, SMS & WhatsApp

Since ADFIX took over in mid-July 2026

Insights

• Broken conversion values don’t slow an account down. They point it in the wrong direction. Automated bidding optimised faithfully toward the cheapest possible sale.
• The single highest-return action of the engagement was not a campaign change. It was making revenue visible.
• Average order value did more work than click cost. Bidding toward value moved AOV up 40% on search, which is most of the efficiency gain.
• Nearly doubling spend while improving ROAS 11% indicates the account was under-invested rather than saturated. There is likely headroom left.
• Paid media was bringing in first-time buyers with nothing to bring them back. Switching on the welcome and abandoned checkout flows turned traffic the brand was already paying for into revenue.
• Traffic grew faster than orders, so conversion rate is now the constraint, not reach. That is the next thing to work on.

Key Takeaways

You cannot optimise toward a number the platform cannot see.

• Measurement rebuilt first, scale second. In that order, deliberately.
• Search budget grown roughly sixfold with return improving, not eroding.
• A second channel added once the first was proven, returning 9.6x since launch.
• Retention rebuilt from zero: Klaviyo-attributed revenue equal to half of store revenue within one quarter.
• Store revenue up 140% and orders doubled, April to September year on year.

Conclusion

Arabic Parfums did not have a demand problem. It had a visibility problem, and the visibility problem was capping the demand. Once the platforms could see what an order was actually worth, budget could follow evidence instead of guesswork, and the account absorbed several times its previous spend without giving up efficiency.

Retention followed the same pattern. The customers were already arriving. Once the flows were switched on, they had a reason to come back.

The order matters. Scaling an account that reports the wrong numbers just buys more of the wrong thing, faster.