Imagery project — product grouping & volume map
1,681 active SKUs · by group × product type × photo approach · basis for the quote · 2026-06-13
Reconciliation: the four groups sum to G1 860 + G2 83 + G3 255 + G4 483 = 1,681, which matches the live active product count exactly. (Full catalog is 1,705; the 24 orange/purple SKUs we archived are excluded from imagery scope. The 33 specials + ~31 net-new with no base photo are inside the 1,681 but can't be AI-generated until Jessar supplies a packshot — see Q3.)
Before pricing, here's what we're actually working with — every active SKU mapped to its real product type (confirmed by sampling titles, not by the brand label) and to the photo approach that type needs. This is the quote basis: the project isn't "1,680 images," it's a mix of effort tiers, and a large share collapses into near-duplicate clusters we shoot once and swap.
Five things the data showed (some surprising)
- "JS Maison" is tiny and isn't what we pictured. Only 83 SKUs, and 76 of them are anti-slip mats (the "Beauty Cover" line) — colorways of one product. The linens / bedding / furniture you described aren't here.
- The home goods are mislabeled under JS Gourmet. Bath/laundry/bins/mirrors (124), side-table furniture (16), baskets, door-hooks, blankets, even extension cords are vendor-tagged "JS Gourmet" (Group 1). Brand label ≠ product type — so we price by type, not group.
- ~Half the catalog is simple packshot work. 584 SKUs (bulbs, electrical, mats, cords) need one clean color-background product shot — not lifestyle. Fast and cheap per unit.
- Big near-duplicate clusters = the efficiency lever. 271 bulbs are spec variations, 76 mats are colorways, extension cords are lengths, drinkware/bins come in colour families. We shoot one template and swap — unique shoots are far fewer than SKUs. This is what makes full-catalog viable.
- Only ~64 true appliances + ~276 light fixtures + ~20 furniture carry the heavy, high-value treatment. That's where the craft (and the price) concentrates.
Photo-approach tiers — the real cost driver (whole catalog)
Product + lifestyle · 2 img
691
Simple packshot · 1 img
584
Lighting fixture · clean + hero lit
277
Full treatment · 3 img (appliances)
64
In-room / environment · furniture
20
Rough image volume at these tiers ≈ 2,600 base images → ~3,560 with a re-roll buffer → ~7,100 credits at 2K — comfortably inside one 9k month (~$310 compute). So the quote to Jessar is workflow + QA + margin, never the pixels.
Group 1 · Kitchen umbrella (JS Gourmet) — 860 SKUs
The bulk of the catalog
sub-lines: Veraroma 32 · Inspiration Italia 25 · Laguiole 8 · iBake 8
| Product type | SKUs | Photo approach | Tier |
|---|
| Drinkware (glasses, mugs) | 142 | product + lifestyle | 2 img |
| Bath / laundry / bins / mirrors ↤ home, mislabeled | 124 | product, colorway dupes | 2 img |
| Serveware (boards, platters) | 112 | product + styled flat-lay | 2 img |
| Food storage / organizers | 110 | product, some in-context | 2 img |
| Kitchen gadgets & tools | 100 | product, some in-use | 2 img |
| Small appliances | 64 | full: product + lifestyle + features | 3 img |
| Prep · glassware · bakeware · other | 66 | product + lifestyle | 2 img |
| Furniture — side tables / ottoman / divider ↤ home | 19 | in-room / environment | env |
| Extension cords · hooks · tape · step-stool ↤ misc | 48 | simple packshot | 1 img |
| Baskets · blankets · untyped kitchen | 75 | styled / review | mixed |
Group 2 · JS Maison — 83 SKUs
| Product type | SKUs | Photo approach | Tier |
|---|
| Anti-slip mats ("Beauty Cover" line) | 76 | flat product, colorway dupes — ~1 template | 1 img |
| Storage / bath / shelving / misc | 7 | product | mixed |
JS Maison as a standalone "home & linens" group barely exists in the data — it's effectively one mat line in colorways. The actual home goods live in Group 1. Decision needed: re-tag the home items into JS Maison, or just price the whole job by product type and ignore the brand grouping for imagery.
Group 3 · Jessar Lighting — 255 SKUs
| Product type | SKUs | Photo approach | Tier |
|---|
| Pendant fixtures | 111 | clean product; hero ones lit in-situ | fixture |
| Desk / table lamps | 43 | clean product | fixture |
| Wall sconces · track · vanity · floor · ceiling | 101 | clean product; premium lit | fixture |
Lighting is its own discipline: a clean product shot reads catalog-grade, but the premium / touch lamps earn a lit in-room hero (that's the "premium lighting" tier). Reflective/glass shades + emitted light are the hardest fidelity case — budget extra re-rolls here.
Group 4 · Xtricity electrical — 483 SKUs
| Product type | SKUs | Photo approach | Tier |
|---|
| Bulbs | 271 | simple packshot — spec dupes, few templates | 1 img |
| Recessed/ceiling LED · security/flood · switches · taps · power bars · boxes · panels · night lights · misc | 188 | simple packshot — "vs the garbage it has now" | 1 img |
| Desk lamps · heaters · fans (consumer items) | 24 | product | 2 img |
Electrical is almost entirely simple packshots on a brand-color background — high SKU count, low effort per unit, heavy near-duplication (271 bulbs ≈ a few dozen visual templates + label/spec swaps). This is the cheapest tier per SKU despite the volume.
Open questions before pricing
- The "Pop" sub-brand isn't in the data. Product-lines present are Veraroma, Inspiration Italia, Laguiole, iBake, Beauty Cover (= the mats), and Limpus (6). No "Pop" — is it discontinued, renamed, or one of these? (Did you mean Limpus?)
- Brand re-tag, yes or no? Home goods sit under JS Gourmet. For imagery we can price by type and ignore it — but if Jessar wants clean brand grouping on-site too, that's a separate data-cleanup line.
- Base-photo gaps remain. 33 Cook & Fresca specials + ~31 net-new request SKUs have no packshot — reference-locked AI can't touch them until Jessar supplies a base photo. They're carved out of the count or priced as "client supplies source."
- Lighting depth: clean product only, or lit in-room heroes for the premium fixtures? Drives the fixture-tier price.
Next step: you validate these groupings + approaches, answer the four questions, and I turn this into the priced quote — per-tier rates × volume, with the near-duplicate efficiency baked in (unique-shoot count, not SKU count). Source data: work/imagery_volume_map.csv.