Proof

The numbers count work done, not sales.

RECORDED · STORE CHECK

A free store check

Three plain coffee bags with rolled tops stand in a row on a dark grey backdrop: a near-black bag, an off-white bag and a pale grey-blue bag, lit softly from the left, with long shadows to the right.
Our picture, not the store’s.

A check of a coffee store’s public pages. Its products met and read each other’s pages. It took 20 minutes.

  • 12of its 148 products met
  • 36talks between them
  • 356things said
  • 56options they read

What they found, each checked again on the live store

  1. Main findingIts refund rules disagree: 60 days, 30 days and a $40 cap.
  2. Main findingOn 9 product pages, the price at the top is for a smaller pack than the one the page opens on.
  3. Smaller noteA 5 lb bag says Sold out while it is in stock.

See more findings

RECORDED · CLIENTS 1 AND 2

Prices checked

MindLoop read each store’s catalogue from its store admin.

  • 258prices confirmed for Client 1
  • 106more Client 1 items priced on request, as the store chose
  • 61prices confirmed for Client 2

RECORDED · CLIENT 1

Would 10% off help?

MindLoop had AI write 8 shopper reactions to each price. Neither price is shown as better. Forecast, not recorded sales.

Client 1, 25 September 2026. Current price: 43.2 percent, range 34.7 to 51.8. 10 percent lower: 51.5 percent, range 46.2 to 56.9. The ranges overlap. Forecast, not recorded sales.

RECORDED · CLIENT 1

A change in Telegram, then undone

A proposal in Telegram changed MindLoop’s goals. The original 3 goals were put back.

RECORDED · CLIENT 1

Customer conversations read

Each finding quotes the messages it came from.

  • 580messages
  • 115conversations
  • 139sources quoted

RECORDED · CLIENTS 1 AND 2

Interviews with real people

  • 13interviews for Client 1
  • 9findings for Client 1
  • 5interviews for Client 2
  • 4findings for Client 2

Questions

Two stores that work with MindLoop, names left out.

No. We ran the free store check on it. Names left out.