No. 03 · E-commerce
Audit / 5 business days
AI Stack Audit for a $2M D2C brand — what to skip and what to ship
Four AI vendors. $1,800 a month combined. Three demos that all looked great. The owner wanted an outside read before signing anything. We told her to skip three.
Client: Direct-to-consumer apparel brand (anonymized)
Two-million-a-year D2C apparel brand on Shopify. Owner runs it with one ops lead. Both of them spend their week being pitched. Every vendor has a deck, every demo goes well, every contract feels like an obvious yes in the moment and a bad idea by Tuesday.
The brief
"I have four contracts on my desk. Each vendor sounds plausible. I cannot tell which of them, if any, would actually save me time. I'd rather pay you a flat fee for a sober opinion than sign $1,800 a month of pretty dashboards."
What we found
Vendor A ($600/mo) — "AI inventory forecasting." She had 14 SKUs. Existing Shopify reports already handled the forecast. Skip.
Vendor B ($450/mo) — "AI customer service." Her ticket volume was ~30/week, all handled by her ops lead in 90 minutes total. Skip.
Vendor C ($300/mo) — "AI-generated product descriptions." She'd rewrite every one anyway because brand voice. Skip.
Vendor D ($450/mo) — "AI vendor reconciliation." Real problem; vendor's pricing reasonable. We said yes here, with a 60-day exit clause.
One workflow nobody had pitched her: weekly Klaviyo email drafts. She spent ~6 hrs/week writing them. Claude could draft them in 5 minutes; she'd edit. Worth ~$28k/year of her time at her own hourly rate.
Monthly AI vendor spend
Combined vendor pitches she was actively considering, before and after the audit.
“I came in expecting to sign two contracts. I left having canceled three demos and built the thing myself for free.”
What we recommended
- 01
60-minute discovery call
Mapped every workflow she or her ops lead spent more than 30 minutes a week on. Not what they wished they were doing — what they actually did. The list ran to 14 items.
- 02
Score each vendor against the real list
For each of the four pitches, we asked: which of those 14 workflows does this vendor measurably change, and by how many hours per week? Three vendors scored zero. One scored real hours.
- 03
Flag the workflow nobody had pitched
The Klaviyo email draft workflow was the largest single time sink and no vendor on her desk addressed it. We sketched a $0/month Claude prompt that would handle it.
- 04
Deliver three documents
Workflow map (PDF, 1 page). Vendor scorecard (1 page). Recommendations doc with skip / try-yourself / hire-out per workflow (2 pages). Total reading time, maybe 12 minutes.
- 05
Don't oversell
Audit is $49. We told her: if you want us to build the Klaviyo workflow, that's a separate Build engagement. She built it herself in a weekend. We were fine with that — that's what an honest audit looks like.
Numbers
- audit delivered
- 5 days
- vendor pitches told to skip
- 3 of 4
- ongoing spend avoided
- $1,350/mo
- in-house workflow worth building
- 1
Outcome
Owner skipped three vendor contracts (~$1,350/mo combined). Signed Vendor D with a 60-day exit clause. Built the Klaviyo workflow herself in a weekend. Came back six months later for a Plan tier on her next bottleneck — wholesale order intake.
What this cost
Audit tier · $49 fixedAudit tier, $49 fixed. The audit paid for itself within the first weekend she didn't sign Vendor A.
Stack
- 60-min discovery call
- Workflow map (PDF)
- Vendor scorecard
- Recommendations doc
Want a similar audit?
A $49 audit beats $1,800 a month in the wrong vendors.
A senior engineer responds personally within one business day. If there isn’t a fit, we’ll tell you and point you somewhere better.