Your catalog has products that are out of stock, missing images, or priced wrong.
They're still listed. Customers see them, click, and leave.
We find them, and hand you a table you can check row by row.
One catalog we audited: 249 products. 107 were out of stock — 43%. 23 had no image at all.
The store didn't know.
249
products in one catalog we audited
107
were out of stock and still listed
43%
of the catalog 23 had no image at all
What actually comes back
Three artefacts, every job. Not a summary of the work — the evidence of it.
01
Change log
Every edit recorded as old value → new value, row by row. If a number changed, you can see what it was.
02
Dedupe report
What merged, what survived, and why. Deduplication runs after normalisation, so Acme Corp. and acme corp are caught.
03
Status column
Every row marked verified, uncertain or rejected. Ambiguous values are flagged with a reason, never filled in with a plausible guess.
// a row that could not be proven from the source
{ "company": "Delta Co",
"date_raw": "",
"date_norm": null,
"email": "delta@example",
"status": "uncertain",
"note": "no date in source; not inferred" }
Pricing
Fixed price per job. No hourly surprises. Every tier includes the change log, dedupe report and status column.
Rush delivery (48 h): +50%. Beyond the top tier: quoted per 1,000 rows or records.
01
Catalog audit (for online stores)
We go through your catalog and find what's costing you sales: out-of-stock products still listed, missing images, wrong prices, dead links. You get the table, the change log, and the dedupe report.
Single catalog, ≤ 500 products
$120
1–2 days
Single catalog, ≤ 5,000
$350
3 days · raw extraction log included
Multi-site + clean + dedupe
$700–1,000
5–7 days
Monthly monitoring
$99/mo
2 re-runs per month, diff against the last run
02
Competitive monitoring (what your rivals just launched)
Every morning we read the public product feed of the stores you name and tell you what is new, what moved in price, and what disappeared. No account access needed — only the store URL.
Dirty file in, auditable table out. Original columns untouched; normalised values go beside them.
≤ 1,000 rows
$40
1 day · change log + dedupe report
≤ 5,000 rows
$120
2 days · includes MX validation
≤ 20,000 rows
$350
3–4 days · full audit trail
Ongoing, monthly
$90/mo
2 batches per month, ≤5,000 rows each
04
Public records & site extraction
Government and public-record sources. One row per entity, every field traced back to its source URL.
Single site, ≤ 500 records
$80
1–2 days
Single site, ≤ 5,000
$250
3 days · raw extraction log included
Multi-site + clean + dedupe
$500–800
5–7 days
Monthly monitoring
$120/mo
2 re-runs per month, diff against the last run
05
AI output audit
Your pipeline produced numbers. Before you act on them, someone checks whether they are trustworthy.
≤ 1,000 rows
$40
Every row marked verified / uncertain / rejected
≤ 5,000 rows
$120
+ discrepancy report: what doesn't reconcile
≤ 20,000 rows
$350
+ root-cause notes on where it broke
Standing monthly audit
$90/mo
2 batches per month
06
Product video (for stores)
TikTok Shop and Reels push video over still images. Send product shots and get a 30-second 1080p clip back the same day. Change a price or a word and the re-render takes 15 seconds, not an afternoon — you get the source, not just the MP4.
15–30 s code film
$80
1–4 shots · 1080p30 · source included
60–90 s code film
$200
4–8 shots · source included
Batch transcode / concat
$40
≤ 20 files, 1 day
Subtitle burn-in
$50
≤ 30 min footage, timing aligned
Transcript + summary
$60
≤ 60 min audio, structured output
Monthly video pack
$280/mo
4 films per month, ≤30 s each
What I don't take
Live-action editingNo camera work, no on-set footage.
After Effects compositingCode-driven only, not AE-grade VFX.
Colour grading / sound designNeeds a treated room and a specialist.
Anything behind a loginPublic sources only. No bypassing walls.