Three services. One reconciled truth underneath all of them.
Trace, Retain and Allocate. Each works alone; together they compound, on one reconciled profit number.
Trace
Your best-seller might be losing money.
True profit and cash in real time, traceable to the source. With Trace:
- Every real cost lands on the unit: fees, commissions, returns and landed cost, read from your invoices, not a rate card.
- Every number opens up: the formula, the source rows, the invoice line. Your accountant can check us.
- Cash sits beside profit: settlement timing, supplier terms and stock cover, so you see where money is stuck.
- We find the leaks and start the fix: the reprice, the reorder, the disputed 3PL charge.
SKU contribution margin
CM1 (Contribution Margin) per SKU after landed COGS + channel variable costs
| Line | Key assumption | Amount |
|---|---|---|
| Net revenue (ex VAT) | £34.00 gross ÷ 1.2 | +£28.33 |
| COGS (landed) | SKU's own landed-cost cell | −£8.40 |
| Gross margin | £28.33 − £8.40 | +£19.93 |
| TikTok commission | 9.0% of GMV, flat since Sep 2024 | −£3.06 |
| Creator commission | Booked affiliate rate | −£3.10 |
| FBT pick & pack (gross) | Size tier M, derived from the bill | −£1.80 |
| FBT shipping (gross) | £3.22/order, per order not per unit | −£3.22 |
| TikTok shipping funding | Withdrawable funding, credited back | +£1.29 |
| Paid ads | Attributed, allocated to unit | −£4.83 |
| Contribution margin 1 | £19.93 GM − £14.72 costs | +£5.21 |
Every line shows its assumption and its arithmetic. Click any figure to see the source rows it came from.
Synthetic data.
Retain
Repeat purchases. Nothing else.
We send each customer a tailored message at the right time, so they buy from you again and again. With Retain:
- We build it, and we run it. Every flow is made for your brand. We build it, run it, and report it, so your team gets its evenings back.
- Every email fits one person. We write to the individual: the product they bought, how fast they get through it, and what other buyers say about it online.
- Your margin comes first. We rarely discount. When one would win a customer back, we bring it to you first.
- Proof, not a dashboard. We hold a small group back and compare. You see the lift the flow truly caused, not a number a tool takes credit for.
Run-out table
example brand · 4 of 38 products
| Product | Pack | Daily use | Runs out after |
|---|---|---|---|
| Barrier Serum | 30 ml | 2 pumps | 42 days |
| Collagen Powder | 300 g | one 10 g scoop | 30 days |
| Scalp Shampoo | 250 ml | 8 ml, alternate days | 62 days |
| Magnesium Capsules | 90 | 3 a day | 30 days |
The figures above belong to an invented example brand.
Allocate
Stop paying to sell at a loss.
Ad spend steered on profit, not platform ROAS. With Allocate:
- Every pound clears a profit floor: after the cost of the click a product must make money, or its spend stops.
- The whole basket counts: a product that looks unprofitable alone often pulls a profitable basket, and gets the credit.
- You approve, we run: we build the campaigns and move the budget on your sign-off.
- Unproved moves wait: we hold them and name the test that would unlock them.
Reallocation, bounded ±30%
Tier 2 · across channels, robustness-gated
Survives all four haircut extrema, incl. halo-widened bound
Depends on the incrementality assumption; one four-week holdout resolves this
tier-1 moves + tier-2 moves that survive the whole haircut grid
2 cross-channel moves the data cannot yet support
Synthetic data.
Join the waitlist.
A few brands at a time. A free 30-day audit on your own data, and the findings are yours either way.