Every campaign, every search term, every listing · 12 months to Oct 3, 2026 · 14-day attribution · audited Oct 4, 2026
Searches that describe our products return $3.80 per $1. Everything else returns under $1, and it takes 71% of the spend analysed. The auto campaigns are showing baklava to people searching for oreos, dates to people searching for strawberries, and honey to people searching for sliced cheese.
| Search-term group | Terms | Clicks | Spend | Sales | ROAS | What it is |
|---|
Groups cover the 569 terms with 3+ clicks, $1+ spend or a sale ($3,353). The other 2,315 terms had one or two clicks each and $851 in total.
47 campaigns with search activity, sorted by spend. Click a row for its best and worst search terms. Two more live campaigns (a second Turkish Delight exact and Delight Mix 650g) had no search clicks.
| Campaign | Status | Spend 12m | Sales 12m | ROAS | Orders | Wasted | 30d spend | 30d ROAS | Action → impact |
|---|
Spend and sales are the search-attributed figures from the Item Keyword report. "Wasted" is spend on search terms that never produced an order. 30-day figures are full campaign spend from the campaign list.
Everything else runs on auto-matching. The three that exist are the best performers here.
| Keyword | Campaign | Spend | Sales | ROAS | Action → impact |
|---|---|---|---|---|---|
ajwa date broad | Ajwa 400g | $65.49 | $353.98 | 5.40 | Add exact "ajwa dates" and "dates ajwa" → more volume at this return |
ajwa date broad | Ajwa 110g (paused) | $74.47 | $309.60 | 4.16 | Re-enable → it was the second-best keyword in the account |
baklava broad | Baklava 100g | $194.44 | $536.01 | 2.76 | Bid is $0.30 against a $0.66 suggestion → raise in 10% steps for more impressions |
baklava broad | Baklava 250g (paused) | $60.91 | $105.96 | 1.74 | Leave paused → the 100g wins this term at half the price |
turkey delight candy | Turkish Delight 250g | $10.10 | $23.74 | 2.35 | Misspelt. Replace with "turkish delight candy" → that term returned 6.2 across the account |
Added from Walmart's suggestions. Remove them → the ad group keeps only baklava terms.
| Keyword | Trend | Product | Evidence | Action → impact |
|---|---|---|---|---|
| turkish delight | 84 | Turkish Delight 250g mix | Never bid; we bid the misspelling | Exact → largest untapped term for the range |
| baklava pastry | 76 | Baklava 100g | 5 orders, ROAS 2.5 via broad | Exact → own bid on a proven term |
| indian sweets | 69 | Mini Mixed Cookies, Baklava 250g | 5 orders, $4.86 → $144 | Exact → cheapest converting term found |
| mop slippers / mop shoes / mop socks | 68 / 27 / 22 | Ribal Mop Slippers | All 7 orders came from these | Exact → relaunch the paused product at ROAS 2+ |
| turkish delight candy | 25 | Turkish Delight 250g mix | 4 orders, ROAS 6.2 | Exact |
| ajwa dates / dates ajwa / ajwa dates from medina | 22 | Ajwa 110g, 400g | 40 orders, ROAS 4.9–10.5 | Exact alongside the broad keyword → more share |
| sukkari dates | 21 | Sukkari 110g | 3 orders, ROAS 6.2 | Exact |
| date cookies | 16 | Leen Maamoul range | 17 units, $9.72 → $338 | Exact |
| barazek cookies / sesame cookies | – / 30 | Barazek 250g | 3 orders, ROAS 4.4–14.7 | Exact |
| arabic sweets / middle eastern sweets / lokum | 29 / 30 / 33 | Baklava, Turkish Delight | 1 order at 9.9; small volume | Phrase, low bid |
| dried dates / whole dates | 74 / 30 | Leen dates | Untested; narrower than "dates" (ROAS 1.0) | Phrase test, $0.40 cap |
| hot honey | 100 | Honeypino chili honey | Untested; auto matched it to cheese | Exact test after the listing fix |
| honey sticks / honey packets / honey straws | 96 / 94 / 69 | Honey Sticks | 69 clicks, 2 orders, ROAS 1.3 | Exact at a low bid once reviews exist |
| coffee alternative / caffeine free coffee / coffee substitute | 62 / 83 / 48 | BAL Datespresso, Groundate | Never advertised; converts 8% organically | New exact campaign → first paid traffic for BAL |
| portable bidet / travel bidet / peri bottle | 89 / 71 / 94 | Ribal travel bidet | Untested; 8 clicks ever | Exact test, small budget |
| petit fours / tea cookies | 67 / 89 | Petit Fours 350g | Untested | Test after the product-type fix |
| shortbread cookies / butter cookies | 100 / 100 | Ghraybeh | 143 clicks, 1 order | Hold → traffic is there, the listing does not convert |
Each produced two or more orders at ROAS 2+ through auto or broad matching. Bidding on them directly gives control of the bid.
Unrelated to any product, or a competitor/category term with $5+ spent and no order. The 30 most expensive are below; the full list is in the master table (filter: Negative).
| Term | Trend | Do we bid on it? |
|---|---|---|
| dates | 100 | Auto only. ROAS 1.0 on $252. Too broad; keep capped |
| medjool dates | 94 | No. We do not sell medjool; skip |
| baklava | 90 | Yes, broad, underbid |
| turkish delight | 84 | No. We bid on the misspelling "turkey delight" (score 0) |
| baklava pastry | 76 | No. Add exact (5 orders, ROAS 2.5 via broad) |
| pistachio cookies | 57 | No. 29 clicks, no sale; test on Barazek with a low bid |
| lokum | 33 | No. Add phrase |
| arabic sweets | 29 | No. Add phrase (ROAS 9.9 on small volume) |
| turkish delight candy | 25 | No. Add exact (ROAS 6.2) |
| maamoul cookies | 24 | No. 32 clicks, no sale → listing problem, see SEO section |
| ajwa dates | 22 | Broad only. Add exact (27 orders, ROAS 4.9) |
| sukkari dates | 21 | No. Add exact (ROAS 6.2) |
| date cookies | 16 | No. Add exact (17 units, ROAS 34.8) |
All 569 search terms with 3+ clicks, $1+ spend or a sale, across every campaign. Search, filter, click a header to sort.
| Search term | Group | Impr. | Clicks | Spend | Orders | Sales | ROAS | Top campaign | Action |
|---|
All 60 published listings, sorted by 30-day page views. Click a row for the proposed title, keywords and fixes.
Walmart matches ads and search by product type. Pistachio Turkish Delight is filed under Edible Nuts and drew 690 clicks for "pistachios". Pomegranate Turkish Delight is under Edible Seeds and drew 129 clicks for "pomegranate seeds". The 250g mix is under Halva. Petit fours are under Chocolate Covered Snacks.
"Maamoul", "Ghraybeh", "Barazek" and "Al-Madinah Al-Munawarah" carry the title, while the terms that convert are "date cookies", "shortbread cookies", "sesame cookies" and "ajwa dates from medina". Walmart's spell-checker also reads maamoul, safawi, khodhri and munawarah as typos. Keep the Arabic name in brackets and lead with the English one. Sizes are in grams only; add ounces.
This is the one Poor category in the 80% listing-quality score, and the likely reason 4,242 views of the 650g Turkish Delight produced one order. Enrol the top eight listings in Review Accelerator and turn on review syndication from our own site.
Description too short or too complex on 44 listings. Images below zoom resolution or loosely cropped on 31. Size Descriptor invalid on 18 (use "Regular"). Nutrition Facts image missing on 17. Milaf sodas lack Caffeine Designation, Container Material and Texture, and three titles use the UK spelling "Flavour".
| Listing | Quality | Views | Conv. | GMV 30d | Price | Top fix → impact |
|---|
| # | Change | Impact |
|---|---|---|
| 1 | Pause Milaf Cola Pack of 24, Milaf Cola - Auto, Safawi 400g, Khodhri 400g, Honey With Nuts | Saves ~$190/mo that sold nothing |
| 2 | Add the negative list to Catch All Zaitoune and Catch All Leen | Cuts the terms behind $971 of no-sale spend |
| 3 | Replace "turkey delight" with "turkish delight" and "turkish delight candy" (exact) | Moves from a term with no searches to one that returned 6.2 |
| 4 | Remove the 15 junk keywords from Baklava 100g; raise "baklava" from $0.30 by 10% | More reach on a 2.76 keyword |
| 5 | Re-enable Ajwa 110g broad | Returned 4.16 before it was paused |
| 6 | Add the proven keywords below as exact match | Bid control on terms that already sell |
| 7 | Fix the nine product types and retitle the top ten listings | Stops mismatched traffic at the source |
| Rule | Where we stand |
|---|---|
| Fix the listing before buying traffic: product type, title, reviews, buy box, in stock | Nine wrong product types; reviews score 17.9% |
| Auto campaigns are for discovery on a small budget; manual exact carries the spend | Reversed: auto carries almost everything |
| One product (or one tight family) per ad group | Catch All Zaitoune holds 14 items in one group |
| Harvest weekly: converting terms to exact, non-converting to negatives | Zero negatives in the account |
| Bid near Walmart's suggested bid, then move in small steps | "baklava" at $0.30 against $0.66 suggested |
| Watch placements: search in-grid usually converts best; cut Buy box and carousels if they lag | Turkish Delight Auto likely leaking to item pages |
| Judge on enough data: about 20 clicks per term before pausing or negating a relevant one | Applied in the master list |
| Keep prices at or below Amazon; Walmart suppresses the buy box otherwise | Five listings priced above Amazon |