Weft turns the cameras and conversations already in your stores into a record of real demand: what shoppers looked at, asked for and couldn't find, by attribute, store and week.
The till and the door counter see how many came in and how few bought. Everything that happened in between, and every visit that ended without a sale, is lost by closing time.
That's the whole story your systems hold about this hour: 212 people came in and 9 of them bought. A 4% conversion, and no reason why.
None of this reaches your systems today. Weft records all of it.
Every week Weft rolls your cameras and consented conversations into one view for merchandising, planning and store operations, with a proposal you can act on.
| Part | Share |
|---|---|
| On the floor | 25% |
| In the range, out of stock | 41% |
| Not in the range | 34% |
| Week | Requests for linen (requests) |
|---|---|
| W1 | 14 |
| W2 | 15 |
| W3 | 17 |
| W4 | 16 |
| W5 | 19 |
| W6 | 22 |
| W7 | 24 |
| W8 | 27 |
| W9 | 29 |
| W10 | 31 |
| W11 | 35 |
| W12 | 38 |
| W13 | 41 |
proposes, line plan
Customers keep asking for a length you don't cut, and most of them leave without buying.
Proposed. Waiting for a decision.
| Size | Units sent (%) | Requests (%) |
|---|---|---|
| 30L | 40 | 22 |
| 32L | 35 | 38 |
| 34L | 20 | 28 |
| 36L | 5 | 12 |
| Item | Size | Asked | On the floor | Stores short |
|---|---|---|---|---|
| Straight fit jean, mid wash | 32L | 64 | 0 (none on the floor) | 9 of 14 |
| Linen midi dress, sage | 10 | 41 | 0 (none on the floor) | 6 of 14 |
| Merino crew, navy | L | 33 | 1 | 5 of 14 |
| Relaxed chino, sand | 34 | 28 | 0 (none on the floor) | 4 of 14 |
| Rain jacket, olive | M | 22 | 2 | 3 of 14 |
proposes, replenishment
Store 07 keeps being asked for a size it doesn't have on the floor, while Store 03 holds more than it sells.
Proposed. Waiting for a decision.
| Part | Share |
|---|---|
| Browsing | 58% |
| With a rep | 27% |
| Waiting for help | 15% |
| Hour | visits | Reps on the floor |
|---|---|---|
| 10 | 22 | 2 |
| 11 | 29 | 2 |
| 12 | 41 | 2 |
| 1 | 63 | 2 |
| 2 | 84 | 2 |
| 3 | 92 | 2 |
| 4 | 86 | 2 |
| 5 | 58 | 2 |
| 6 | 47 | 2 |
| 7 | 35 | 2 |
| 8 | 24 | 2 |
| Week | Picked up (per week) | Sold (per week) |
|---|---|---|
| W1 | 64 | 10 |
| W2 | 76 | 9 |
| W3 | 88 | 8 |
| W4 | 94 | 6 |
| W5 | 90 | 5 |
| W6 | – | 21 (projected) |
| W7 | – | 26 (projected) |
| W8 | – | 28 (projected) |
| W9 | – | 27 (projected) |
| W10 | – | 24 (projected) |
| W11 | – | 20 (projected) |
| W12 | – | 16 (projected) |
proposes, staffing
Traffic doubles in the afternoon while the floor keeps the same two reps.
Proposed. Waiting for a decision.
The cameras you already have become a live, anonymous map. Weft knows who is browsing, who is being helped and who has been waiting too long.
The Weft Floor app shows every rep what needs doing right now: a shopper left waiting, a queue at the till, an empty rack, a table to fold. One tap and it's taken.
Reps tap to record a styling conversation. Speech becomes text, the audio is deleted, and each request becomes a row of structured demand.
Every week rolls up into store and conversation intelligence: layouts, staffing, markdowns and next season's line plan.
Weft Floor puts the whole store on every rep's phone: a shopper left waiting, a queue at the till with no one behind it, an empty rack, clothes left unfolded. A free rep takes it with one tap.
Camera 2, main floorNumbered visits. No faces, no names.Needs attention0
Recording with consentAudio is deleted once transcribed
Do you have these in a 32 long?
Only 30 and 34 on the floor.
Anything in a lighter wash?
Not in this fit, sorry.
Visit 3, store 07, 2:17 pm
Run that across 14 stores for a quarter and two decisions appear that no sales report can show you: which sizes to rebalance this week, and which attributes to add next season.
From the cameras you already have
Every visit, every hour, with nothing for staff to do. Six things the cameras see that your tills never will.
212visits in one Saturday hour. 41% stopped at the window first.
Used by Store operations, marketing
71%of open hours with no rep near the back-right corner.
Used by Store operations
64 → 6shoppers reached the denim wall; 6 bought denim. Attention, set against what sells.
Used by Visual merchandising
2xthe traffic on Saturday 1 to 4 pm, with the same two reps. 11 of 31 shoppers left the fitting-room queue.
Used by Store operations, HR
12 minthe fitting rooms were busy with no rep nearby, flagged in the moment.
Alert sent to Maya, 2:41 pm
Used by Loss prevention
412 / 38times the linen midi was picked up versus sold. Mark down 20% in week 6.
Used by Merchandising, pricing
From consented conversations
A conversation holds far more than what to buy next season. Each one feeds the team that can act on it.
6,480requests in a quarter, sorted by what you could do about them.
“Do you have this in a 32 long?”
Used by Merchandising, planning
38%of walk-outs came down to a size that wasn't on the floor.
“It's cheaper on your website.”
Used by Pricing, e-commerce
2.1xmore sales when a rep offers an alternative.
Used by Training, store managers
2,140answers this week to one question, asked naturally in conversation.
“I saw the lighter wash in the window.”
Used by Marketing
3xmore often, linen shoppers mention travel.
“Is this good for a week away?”
“Will it crease in a suitcase?”
Used by Marketing, e-commerce
19%of walk-outs were about fit, traced back to the product.
“The 10 is tight across the shoulders.”
“The hem came down after one wash.”
Used by Product, design, quality
Runs on your existing cameras.
Best for store operations and loss prevention.
Runs on staff phones.
Best for merchandising, sales and marketing.
Vision and Voice, linked.
Best for retailers who want the whole picture.
Each option is priced per store, per month. Start with one and add the other any time.
Book a demoLive demo, 30 minutes
Computer vision and consented conversations bring out the hidden metrics in your stores: who waited, what they asked for, and what walked out the door. See them for your own floor.
Weft is designed to work with the IP cameras most stores already have.
No. Weft tracks anonymous visits and never identifies a person.
Recording only starts when a rep taps record, after a spoken disclosure, with signage in store. We set this up state by state with your counsel.
They get live alerts when a shopper is waiting, the till needs someone or a rack needs restocking, plus personal coaching from their conversations.
The first demand report arrives two weeks after install.