Tabeer Homes sells handcrafted bone inlay and mother of pearl furniture from Dubai into the Gulf. Between February 2025 and April 2026 we built the organic demand engine, rebuilt the paid account, and then found the thing nobody was looking at.
How to read this document
The first half is the record: what the brand looked like in February 2025, what was built, and what the numbers did. The second half is the argument.
Section 09 contains the finding that matters most commercially. Section 11 contains the reason the results are not larger than they are. Appendix B lists eleven errors and unresolved contradictions found across the source reports, two of them material enough that the affected figures should not be quoted until corrected.
01 · The Brand
Pieces are made by hand, priced between roughly AED 2,000 and AED 5,000, and bought perhaps once every two or three years. Almost every constraint in this case study follows from those facts.
The catalogue is small, around fifty active products at the April 2026 audit, and roughly half is made to order on a six to ten week lead time. The brand competes on craft, not price: curation, exclusive design, bespoke sizing, ethically sourced materials and a concierge-style buying experience.
That positioning is also the source of the friction. A buyer spending AED 3,000 on furniture they cannot touch, from a brand they do not know, waiting up to ten weeks, needs a great deal of reassurance. How well the website supplies it turns out to be the single largest determinant of the results.
Seven competitors were named at onboarding: Home Kode, Marina Home, 2XL and CB2 locally; Anthropologie Home, Graham & Green and Fenton & Fenton internationally. Only one mattered for search.
Home Kode ranked for sixteen times as many keywords as Tabeer, on roughly a quarter of the referring domains.
The buyer. Source: client onboarding workbook, February 2025
02 · The Engagement
The proposal was explicit that organic search was the primary engine and paid media the accelerant, a sequencing decision the results would later vindicate more strongly than anyone expected.
| Line | Monthly (AED) | Monthly (USD) |
|---|---|---|
| SEO | 3,750 | $1,020 |
| PPC management | 1,250 | $340 |
| Local SEO | 500 | $136 |
| Total retainer (excludes ad spend) | 5,500 | $1,496 |
The original February 2025 proposal priced SEO, backlinks and content at AED 2,800 per month and paid media management at AED 2,000, a combined AED 4,800 discounted to AED 4,000. Recommended opening ad budget was AED 5,000 per month, with management fees stepping down from 25% to 10% of spend as budgets grew.
Sales, not calls, not leads, not traffic. Worth holding onto, because much of the reporting that followed drifted towards traffic and lead counts instead.
The proposal set out six client-side dependencies: platform access, timely content approvals, active communication, backend access, stakeholder alignment and budget approvals. Five of the six became recurring friction points.
03 · The Starting Position
February 2025. More referring domains than any competitor in the category, and almost no traffic to show for it. The site had links; the pages they pointed at were not built to rank.
The clearest illustration sat on one collection page. Coffee tables drew 16 organic visits a month from 19 keywords, with zero backlinks. Home Kode’s equivalent drew 3,300 visits from 63 keywords, carried optimised body copy beneath the grid, and cross-linked to related collections.
Coffee tables collection page: monthly organic visits
December’s budget ran 3.4 times November’s, for 41% of the return, a measure of how little of it was being steered.
04 · The Strategy
Organic search as the compounding asset, paid search as the volume lever that could be tuned monthly, and the website itself as the multiplier on both. The first two were executed. The third is where the story turns.
Take the entire material vocabulary and own it before anyone notices it is available.
Programme one: organic search
Meta titles and descriptions across target pages, heading hierarchy, alt text remediation, canonical and robots checks, sitemap validation, 404 and redirect cleanup, schema review.
Keyword-mapped body copy for collection pages that had been bare grids, the highest-leverage on-page fix, because collection pages are what commercial searches want.
Content mapped to three clusters: luxury furniture in the UAE, design inspiration, and care and maintenance. Each cluster feeding internal links down to collection and product pages.
Foundational links, social bookmarking, blog and directory submissions, Quora placements, plus a competitor backlink gap analysis and a separate Kuwait-specific link sheet.
City-level targeting across Dubai, Abu Dhabi, Riyadh and Jeddah, with Arabic-language terms treated as a distinct and largely uncontested opportunity.
Programme two: paid search
The Google Ads account was not tuned; it was rebuilt. Six pillars: a full account restructure segmented by category and margin with branded traffic isolated; a product feed overhaul with rewritten titles, GTINs, material attributes and margin-tier custom labels; intent-based keyword targeting with luxury qualifiers and aggressive negative lists; audience layering with three RLSA tiers and customer match; a full-funnel retargeting engine built from nothing; and a move from manual CPC to target ROAS bidding with per-tier rather than global targets.
Programme three: the website
Conversion work was scoped from the outset (a UI and UX review sat in the February 2025 task list) but was not executed with the same force until April 2026, fourteen months in.
05 · Execution
Drawn from the client workbook task log, the monthly reports and the April 2026 audit set. A record of work delivered, including the parts that stalled.
Onboarding and questionnaire; analytics, Search Console and Tag Manager configured. Initial report on authority, rankings, traffic, backlinks, indexation and tag coverage. Keyword research, competitor gap analysis, buyer personas, and on-page and technical audits.
Platform access secured. Meta titles and descriptions, alt text remediation, H1 and H2 fixes, duplicate heading cleanup, canonical and robots validation, sitemap checks, 404 and 301 sweeps, schema review, speed analysis. Collection page content published. Foundational links, bookmarking, Quora and blog submissions. Backlink gap analysis. First sample blog drafted for approval.
Content cadence, continued link acquisition, ongoing on-page work against the target keyword set. Rankings begin to move in the bone inlay and mother of pearl clusters.
UAE organic traffic 339 to 453 month on month, up 33.6%. 52 keywords ranking across 44 pages. Search Console clicks 508 to 585.
Eight-month performance reviewed against the prior eight months. 28 keywords holding position one. The review reports a 6.5× return on tracked marketing cost.
Return on ad spend swings from 5.86× in November to 0.71× in December and back to 4.02× in January. An ad account payment failure pauses campaigns in December. Conversion tracking problems surface in January.
Paid conversions rise from 47 to 130 quarter on quarter while cost per conversion falls from AED 652 to AED 346. Every core website engagement metric improves.
A full SEO audit, a conversion audit and a rebuild report are produced in the same week. A Shopify theme cleanup ships: tag manager deferred out of the critical path, font loading switched to swap, six dead pages unpublished, three orphaned templates removed, all validated through a fifteen-check quality assurance pass before publishing.
06 · Results · Organic Search
Across the review window organic search grew clicks by half and impressions by seventy percent, and took twenty-eight keywords to position one in a vocabulary no competitor was defending.
| Metric | Before | After | Change |
|---|---|---|---|
| Search Console clicks | 2,120 | 3,170 | +49.5% |
| Search Console impressions | 159,000 | 271,000 | +70.4% |
| Click-through rate | 1.33% | 1.17% | −0.16 pts |
| Keywords at position 1 | n/a | 28 | +28 |
| Further keywords in top 10 | n/a | 7 | +7 |
| Referring domains | 347 | 360 | +13 |
| Total backlinks | 3,200 | 3,600 | +400 |
Review window against the preceding equivalent period
Impressions grew faster than clicks, so click-through fell from 1.33% to 1.17%. Two things cause that pattern. The first is healthy: as a site starts ranking on page two for broader head terms, it accumulates impressions that will never convert to clicks until the rankings improve further. The second is structural: AI overviews increasingly answer informational queries without a click. The team identified this in the October review and shifted content briefs toward formats that earn citation inside those overviews.
This is the weakest part of the organic programme and it should be stated plainly. Between the February baseline and the August audit, total backlinks rose from 3,200 to 3,600, but referring domains moved only from 347 to 360, a net gain of 13 domains against a commitment of 150 backlinks.
Four hundred links from thirteen sources is a directory-and-profile footprint, not an authority programme. Semrush authority score sat at 9 in August. No editorial coverage in Gulf lifestyle or interiors publications was secured during the period, despite being identified as the highest-value available link opportunity in both the original proposal and the April 2026 audit. The rankings were won on relevance and low competition, not on authority, which is fine while the niche stays uncontested, and fragile the moment it does not.
07 · Results · Paid Search
Paid search produced the programme’s most quotable numbers and its least trustworthy ones. Both deserve setting out.
Monthly return on ad spend, November 2025 to January 2026
Within the October 2025 business review window the paid figure was tighter and better attributed: AED 13,400 of spend against AED 66,549.03 of Shopify-attributed sales, a 4.97× return across 29 orders, roughly AED 462 of media cost per order at an average order value of AED 2,295.
| Metric | Q4 2025 | Q1 2026 | Change |
|---|---|---|---|
| Conversions | 47 | 130 | +176.6% |
| Cost per conversion | AED 652 | AED 346 | −46.9% |
| Ad spend | AED 30,644 | AED 45,225 | +47.6% |
| Impressions | 2.4M | 3.7M | +55.7% |
| Conversion rate | 0.17% | 0.35% | +106% |
| Sessions | 112,820 | 133,608 | +18.4% |
| Engagement rate | 22.73% | 27.76% | +22.1% |
| Bounce rate | 77.27% | 72.24% | −6.6% |
| Conversions per dirham | 0.001534 | 0.002875 | +1.87× |
It is the ratio of the conversion growth rate to the spend growth rate. Conversions per dirham went from 0.001534 to 0.002875, an improvement of 1.87×, not 3.8×. The corrected number is still the best quarter in the account’s history. It is simply half of what was claimed, and the claim is easy to check.
Cost per conversion by campaign, Q1 2026, lower is better
An Arabic Performance Max campaign launched in February spent AED 673 and produced nothing at all, because it ran Arabic targeting against English landing pages. The April 2026 restructure cut daily budget from AED 505 to AED 380, a saving of AED 3,750 a month, by moving money out of Kuwait and Arabic Performance Max and into remarketing and branded search.
The 3.63× return is derived from a conversion count and an assumed order value. If a meaningful share of those conversions are form fills, WhatsApp taps or add-to-carts rather than purchases, the true figure is lower. Until purchase tracking is verified end to end, the Shopify-attributed 4.97× is the number to stand behind, because it counts orders that actually happened.
08 · Results · The Blended Picture
The October 2025 business review put the headline at 6.5×. That number is defensible with two qualifications attached, and it is better to attach them here than to have a prospective client find them later.
| Component | Amount (AED) |
|---|---|
| Agency fees (approx. eight months at the AED 2,800 SEO line) | 22,161 |
| Google Ads spend (media only) | 13,400 |
| Total tracked cost | 35,561 |
| Organic search revenue (Shopify attribution) | 165,695.92 |
| Paid search revenue (Shopify attribution) | 66,549.03 |
| Total tracked return | 232,244.95 |
| Reported return, as published | 6.53× |
First, the denominator is incomplete. The AED 22,161 of agency fees corresponds closely to eight months of the SEO retainer alone; paid media management was billed separately. Adding it back at AED 1,250 to AED 2,000 a month gives an honest range of 4.5× to 5.1×.
Second, this is a revenue-to-cost ratio, not a return on investment. Cost of goods, fulfilment, payment processing and the separately managed Meta budget sit outside it. A 6.5× revenue multiple is not a 6.5× profit multiple and should never be presented as one.
Between February and October 2025, AED 35,561 of tracked marketing cost was associated with AED 232,245 of Shopify-attributed revenue from organic and paid search, a 4.5× to 6.5× revenue multiple depending on how paid-media management fees are treated. Stated that way it survives scrutiny, and it is still an excellent result.
Store-wide movement: Shopify marketing report, all channels including the separately managed Meta budget
09 · The Finding That Matters Most
Buried in a screenshot on slide five of the business review is the most commercially consequential fact in this entire engagement. It was never called out in the deck. It should have been the deck.
Revenue per session, by channel, in AED
The same five channels, by share of sessions and share of revenue
| Channel | Type | Sessions | Revenue (AED) | Orders | Conv. rate | AED / session |
|---|---|---|---|---|---|---|
| Google Search | organic | 6,243 | 165,696 | 52 | 0.83% | 26.54 |
| paid | 11,907 | 66,549 | 29 | 0.24% | 5.59 | |
| Direct | direct | 26,333 | 70,448 | 26 | 0.10% | 2.68 |
| paid | 50,220 | 90,075 | 21 | 0.04% | 1.79 | |
| paid | 29,798 | 37,947 | 17 | 0.06% | 1.27 |
Someone searching “bone inlay coffee table Dubai” has already decided what they want and is looking for somewhere to buy it. Someone scrolling Instagram has decided nothing. Paid social was reaching an audience at the wrong moment and being asked to close a AED 3,000 considered purchase on the strength of a photograph. Organic also carried the second-highest average order value in the table at AED 3,186.
Last-click attribution flatters search and punishes social. Some of those organic sessions are people who first met the brand through a paid social ad and later searched for it by name, demand that social created and search harvested.
That caveat softens the ratio. It does not overturn it. Facebook carried the highest average order value in the table at AED 4,289 and still could not convert; and the branded-search conversion rate of 3.07% in Q1 2026 shows the harvesting layer was working precisely as intended. The conclusion is to re-measure paid social as an awareness channel feeding search, not to switch it off.
10 · The Conversion Ceiling
In April 2026 the measured conversion rate was nine hundredths of one percent, against a category benchmark of one to three. Every dirham of search and media work through 2025 was landing on a page that could not close.
Where the traffic is actually lost
A leftover app was telling every visitor to a AED 3,679 coffee table that they were signing up for recurring charges. Estimated impact of removing it: 0.4 to 0.6 percentage points, on a base of 0.09%. It is a five-minute uninstall.
To a buyer weighing AED 3,000 with an unfamiliar brand, a stale promotion says nobody is home.
48% of the catalogue (24 of roughly 50 products) ships in six to ten weeks. Customers discovered the wait at checkout. Checkout-to-purchase ran at 22.4% against a 60%+ benchmark; roughly 267 people a month added an item and left.
Beyond those: no reviews or star ratings on a site selling AED 3,000 handcrafted goods; no warranty mentioned anywhere; buy-now-pay-later logos present but instalment amounts never shown; no express checkout, exit-intent capture, cross-sell, scarcity signals or session recording.
Two days after the rebuild report, a follow-up audit checked every recommendation against the live site.
Done, of 30 checks
Partial: started, not finished
Not done: including the critical one
The subscription disclaimer, a five-minute uninstall flagged as the highest-impact fix on the site, was still live. Of the fifteen fixes in the original report, four were complete, five partial, six untouched.
The traffic problem was solved in 2025. The conversion problem is still open.
11 · What Slowed This Down
A case study that lists only wins is a brochure. These are the things that cost this programme real performance, including the ones that were the agency’s to own.
Client-side
Agency-side: ours to own
Ramadan, running from late February to late March 2026, suppressed Gulf conversion throughout the quarter and produced a run of zero-conversion days in the ad account. And the brand’s own stated seasonality, a reliable dip through July and August, sits directly inside the review window, which if anything makes the organic growth recorded across it slightly more impressive than it appears.
12 · What Transfers
Stripped of the specifics, five conclusions apply to any considered-purchase e-commerce brand in a visually driven category.
Skip the generic category fight and take the bone inlay and mother of pearl vocabulary instead: lower volume, near-zero competition, and a searcher who has already chosen the material and is therefore most of the way to buying. Twenty-eight position-one rankings came from refusing the obvious fight.
The highest-leverage on-page work was not the homepage or the blog. It was writing real body copy for collection pages that had been bare product grids. Commercial searches want a category page, and a category page with no text on it cannot rank.
Paid social looked like the dominant channel on every traffic dashboard and was the weakest revenue channel in the store. Revenue per session made that visible in one line. Any brand buying reach in a visual category should run this calculation before its next budget review.
A site converting at 0.09% wastes most of what marketing sends it. Fourteen months of ranking and media work sat on top of a product page that told buyers they were signing up for a subscription. Conversion work is not the phase after acquisition; it sets the ceiling on it.
Google Ads conversions and Shopify orders differed by an order of magnitude for months. Every efficiency claim built on the larger number inherits that uncertainty. Verify the purchase event end to end before anything is reported against it.
13 · The Road Ahead
The acquisition engine is built. The next step change does not come from more traffic; it comes from converting the traffic that already arrives, and from moving budget toward the channel that has already proven it closes.
Immediate: the five-minute fixes still outstanding
This quarter: close the funnel
This year: compound what is working
The one-line summary of fourteen months
The demand was created and captured. What remains is a checkout that lets people through it.
Appendices A to C
Every figure in this document is traced to a named source. Where the source documents disagree, the discrepancy is listed and resolved here rather than quietly averaged away.
| Document | Date | What it supplies |
|---|---|---|
| Tabeer Homes Business Review (.pptx) | Oct 2025 | Headline return, revenue split, Search Console growth, position-one keyword export, Shopify channel table |
| SEO and PPC Combined Proposal (.pdf) | Feb 2025 | Baseline traffic, authority, keywords, referring domains; competitive set; scope; commercial terms |
| Client Workbook (.xlsx) | 2025 to 26 | Onboarding responses, task log, monthly ad spend and return |
| Monthly Report, August 2025 (.pptx) | Sep 2025 | Month-on-month traffic, keyword and page counts, backlink audit |
| Q1 2026 Performance Report (.pptx) | Apr 2026 | Quarterly paid performance, campaign cost per conversion, GA4 engagement, channel mix |
| CRO & Website Rebuild Report (.pdf) | 6 Apr 2026 | Conversion rate, checkout completion, ranked defects |
| CRO Audit: Recommendations vs Reality (.pdf) | 8 Apr 2026 | Implementation status of every prior recommendation |
| Complete SEO Audit Report (.pdf) | 6 Apr 2026 | Dimension scorecard, technical and content findings |
| Google Ads report, Windsor AI pull | Apr 2026 | Ninety-day performance, campaign decisions, proposed budget |
| Paid media case study (.html) | Mar 2026 | Six-month paid rebuild figures and restructure detail |
| Billing and communications records | 2025 to 26 | Retainer structure, payment history, relationship context |
Attribution throughout is Shopify’s marketing report unless stated otherwise. Where Google Ads and Shopify disagree, both are shown and the disagreement is treated as a finding rather than resolved by preference.
Eleven issues were found across the source reports while assembling this case study. Two are material. All are listed so that the affected figures are corrected before they are quoted again.
Q1 2026 report, slides 4 and 10. The figure is the ratio of the conversion growth rate (176.8%) to the spend growth rate. The true improvement in conversions per dirham is 1.87×.
The correct figure at a 5% close rate is approximately AED 20,482. The slide overstates by roughly three orders of magnitude and appears to be inherited from another client’s template. It must not be shown again.
Google Ads reports 130 conversions in Q1 2026 alone; Shopify attributes 29 orders to paid Google across the whole eight-month review window. The Q1 deck calls them leads, the paid case study calls them orders. Resolve before either number is quoted.
Business review, slide 3. AED 22,161 corresponds to approximately eight months of the SEO retainer alone. Including paid media management gives 4.5× to 5.1×.
Search Console had been live since February 2025 and the site was recording roughly 585 clicks a month by August. The audit was crawl-based and states it ran without Semrush or Ahrefs access; its estimates should not override measured data.
The workbook records collection page content and a sample blog draft as complete in March 2025. Either publication never followed drafting, or the crawl missed it. Verify before repeating either claim.
Q1 2026 report gives a Q1 average of 72.24% on slide 7 and 68.55% on slide 9.
The same report shows 130 leads on slide 10, 131 on slide 4, and GA4 monthly figures summing to 118 on slide 6. The Windsor pull gives 127.77 over a different ninety-day window.
Q1 2026 report, slide 3 states conversions dropped 57% from January to March; the GA4 figures on slide 6 show 47 to 32, a fall of 32%. The sources differ and neither is labelled.
Baseline domain rating of 3 is Ahrefs; the August figure of 9 is Semrush authority score. The scales are not comparable and no growth claim should be made across them.
Including the slides showing Google Search Console and Shopify data. Cosmetic, but it undermines confidence in the sourcing of the rest.
| Metric | Value |
|---|---|
| Monthly organic traffic | 183 |
| Ranking keywords | 378 |
| Referring domains | 347 |
| Total backlinks | 3,200 |
| Domain rating (Ahrefs) | 3 |
| Google Ads return | Below breakeven |
| Metric | Prev | Aug | Change |
|---|---|---|---|
| Organic traffic, UAE | 339 | 453 | +33.6% |
| Organic traffic, worldwide | 442 | 587 | +32.8% |
| Search Console clicks | 508 | 585 | +15.2% |
| Search Console impressions | 45,500 | 52,100 | +14.5% |
| Keywords ranking | n/a | 52 | n/a |
| Pages ranking | n/a | 44 | n/a |
| Metric | Measured | Benchmark / target |
|---|---|---|
| Site conversion rate | 0.09% | 1.50% target · 1 to 3% category |
| Checkout completion | 22.4% | 60%+ |
| Average order value | AED 3,151 | n/a |
| Pre-order share of catalogue | 48% | 6 to 10 weeks lead time |
| Recommendations implemented | 6 of 30 | 7 partial · 17 not started |
| Overall SEO audit score | 38/100 | n/a |