peanut
NewEarly access for Shopify merchants

Your store has a profit problem.
Peanut finds it.

The AI profit analyst that watches every order, product and cost line, investigates what changed, and tells you what to do about it, with the evidence attached.

Try the live demo Join early access

No sign-up. The demo is an invented store with illustrative data.

app.peanut.ai/dashboard
Overview
7 days30 days90 days

You made an estimated £43,660 contribution profit, down 8% even though sales grew. 9 issues need attention, worth about £11,916/month.

Net sales
£135,151
▲ 4.8%
Contribution profit
£43,660
▼ 8.3%
Contribution margin
32.3%
▼ 4.6 pts
Profit per order
£21.71
▼ 15.0%
Contribution profit by dayParcelFast rate rise · 17 Oct
10
profit detectors running daily
26
tools behind every answer
0
invented numbers: code does the maths
Read-only
access to your Shopify store

Dashboards wait for you to notice. Peanut notices first, works out why, and brings you the evidence before the month is over.

How it works

From store data to a clear next step.

Shopify · read-only
✓ Orders ✓ Refunds ✓ Products
1

Connect your store

Read-only access to orders, refunds, products and costs. No customer names or addresses are stored.

10 detectors · daily
9 open cases in the demo store
2

Peanut monitors profit

Every product, carrier, channel and cost line, every day, checked for statistics and for money.

Case 3 · ParcelFast
Next step: review invoices since 17 Oct
3

You get the case and next step

An alert when something changes, a case file with evidence, and next steps you approve.

Features

Everything an analyst would do, grounded in your numbers.

Why did my margin drop last week?
✓ get_profit_summary✓ list_incidents✓ get_incident
Contribution margin fell from 36.9% to 32.3%. The biggest driver is Meta: spend up 39.6% while attributed sales fell 4.8%, about £6,117/month. ParcelFast labels add another £1,694/month.
Attribution is last-click, so the Meta figure is medium confidence at best.

Can I just ask why my profit dropped?

Yes. Ask Peanut in plain English and it answers from your store’s data, shows which figures it used, and tells you what it couldn’t see. The AI explains; deterministic code does every calculation.

  • ✓26 tools that compute every figure
  • ✓Keeps follow-up context
  • ✓Says “I don’t know” when the data can’t tell
Ask the demo store
Profit watchdogDetected 6 Nov
Meta spend up 39.6%, sales down 4.8%
£423 → £591 median daily spend
−£6,117/mo
Aurora Hoodie refunds at 14.7%
was 5.6% · “sizing runs small”
−£2,527/mo
ParcelFast cost per label up 22.1%
£6.72 → £8.21 · 532 orders
−£1,694/mo
Trail Bottle unit cost up 25%
£9.60 → £12.00 · price unchanged
−£1,229/mo
New recurring cost
Review booster app · £349/month
−£349/mo

What does Peanut watch while I’m busy?

Every product, carrier, channel and cost line, every day. Ten detectors compare recent weeks with the weeks before and only speak up when a change is both statistically real and worth money.

  • ✓Change-point dating: when it started
  • ✓Impact in £ per month
  • ✓Lower thresholds for watchlisted items
See the open cases
Case 3 · high confidence
Shipping cost per order · ParcelFast
−£1,694/mo
Measured
£6.72 → £8.21 per label across 532 orders; units per order 1.49 → 1.45.
Can’t determine
New rate, surcharge or different destinations.

How do I know a finding is real?

Each significant problem becomes a numbered case file with the baseline, the measured change, the orders behind it, and a clear line between what was measured and what the data can’t determine.

  • ✓Dated evidence chart
  • ✓Supporting transactions
  • ✓Status from new to resolved
Open a case file
What if…Hypothetical
Raise Trail Bottle price
£29 → £31
Contribution profit, next 30 days
+£489
Still ahead unless units fall by more than 14.9%
Assumes typical price sensitivity (−1.2); everything else held constant. Not a forecast.

What happens if I change my price?

Model price changes, ad budgets, shipping rates, refunds and unit costs before you touch your store. Every result is labelled hypothetical, with its assumptions and break-even point.

  • ✓Price, ads, shipping, refunds, costs
  • ✓Break-even shown for every scenario
  • ✓Refuses to run on unknown costs
Try the simulator
p
Daily briefing
Fri 6 Nov · compared with last Fri
Net sales were £4,217 from 72 orders, down 11%. Estimated contribution profit £1,227, a 29.1% margin, down 19%.
  • Meta spend up 39.6%, sales down 4.8%£6,117/mo
  • Aurora Hoodie refunds at 14.7%£2,527/mo
  • Estimate: Ember Camp Mug has no costData

Will it tell me first?

A daily or weekly briefing that leads with what changed, plus one alert when a problem starts and another only if it gets worse. You choose categories, minimum amounts and quiet hours.

  • ✓In-app and email delivery
  • ✓Digest or instant
  • ✓Quiet hours and snooze
Read today’s briefing

And the rest of the toolkit.

Product intelligence

Profit per product and per unit after every cost, with catalogue photos. Unknown costs stay unknown, never zero.

Open in demo →

Full P&L

Gross sales to profit after fixed costs, with “How is this calculated?” on every line.

Open in demo →

Expense capture

Photograph an invoice; Peanut drafts the supplier, date, amount and category for you to confirm.

Open in demo →

Action centre

Save next steps with a priority and deadline, then see how the metric moved afterwards.

Open in demo →

Watchlists

Pin the products, channels or costs you care about and watch them with tighter thresholds.

Open in demo →

Business memory

Your target margin, key products and preferences. See, edit or delete anything it remembers.

Open in demo →

Profit inbox

Risks and evidence-backed opportunities in one place, ranked by money at stake.

Open in demo →

Data health

Checks for missing costs, gaps and unconnected sources before they skew your numbers.

Open in demo →

Recovery journal

What you acted on, and what happened to profit next, in one running record.

Open in demo →

Example figures are illustrative and taken from the demo store.

An investigation, step by step

What actually happens when profit slips.

A real case from the demo store, from the first signal to the decision you make.

Explore the cases
  1. 1

    Something moves

    Shipping is growing faster than sales: £7,400 in the last 14 days against £6,622 the fortnight before, while net sales fell.

  2. 2

    Peanut measures it

    Shipping is split by carrier. ParcelFast alone costs about £1,694 a month more than before, with a t-statistic of 29.4.

  3. 3

    It finds where

    ParcelFast labels went from £6.72 to £8.21 per order from around 17 Oct, across 532 orders. Units per order didn’t rise (1.49 → 1.45).

  4. 4

    It says what’s unknown

    Whether it’s a new rate, a surcharge or different destinations can’t be told from order data: zone and weight aren’t available.

  5. 5

    You decide

    Suggested next step: review ParcelFast invoices since 17 Oct, or get a competing quote. Save it as an action and Peanut tracks what happens next.

Why you can trust it

Built so the numbers are never invented.

The AI never does the arithmetic.

Every figure comes from tested, deterministic calculations. The language model only explains them and is forbidden from inventing numbers.

Unknown is not zero.

A product with no cost on record or an ad account that isn’t connected is flagged on every page, and affected figures carry an Estimate badge.

Measured, not caused.

Findings separate what was measured from what the data can’t determine. Ad findings are capped at medium confidence.

Read-only, always.

Peanut never changes prices, ads or shipping settings, and never messages anyone on your behalf without approval.

Per-store data isolation Full audit log Delete everything, any time Never used to train AI models
p
peanut
AI profit analyst
How's my store doing?
Figures from the demo store. Apple messaging is subject to approval.
Messaging-first

Your analyst, one message away.

Ask how yesterday went, request a chart, or open an investigation from a conversation. Peanut replies with figures, product images and charts rendered from your data, linking to the full evidence.

  • In-app chatAvailable in the demo
  • Email alerts & briefingsBuilt; enabled per deployment
  • WhatsApp BusinessIn development; needs Meta approval
  • Apple Messages for BusinessPlanned; needs Apple approval

We only call a channel available once it works end to end.

Pricing

Simple plans. No surprises.

Preliminary early-access pricing. Billing isn’t switched on yet; we’ll confirm prices before anything is charged.

Starter

£29/month

For a single store getting started with profit tracking.

  • ✓ Profit dashboard with full P&L
  • ✓ Automatic monitoring and case files
  • ✓ In-app alerts and email
  • ✓ AI agent (limited monthly questions)
  • ○ Shopify connectionSoon
Join the waitlist

Growth

Most popular
£59/month

For stores that want investigations, simulations and personal monitoring.

  • ✓ Everything in Starter
  • ✓ Simulations with break-even
  • ✓ Watchlists and business memory
  • ✓ Visual reports and briefings
  • ✓ Expense capture from photos
  • ○ WhatsApp deliverySoon
Join the waitlist

Pro

£99/month

For brands with more stores, people and integrations.

  • ✓ Everything in Growth
  • ○ More AI usageSoon
  • ○ Multiple storesSoon
  • ○ Team membersSoon
  • ○ Ad platform integrationsSoon
  • ○ Apple Messages for BusinessSoon
Join the waitlist

✓ = working in the demo today. Plans and limits may change before launch.

FAQ

What merchants ask first.

Something else? Get in touch.

What is Peanut?

An AI profit analyst for Shopify merchants. It calculates your profit from orders, refunds, product costs, shipping, fees and ad spend, watches for meaningful changes, investigates them, and explains what it found, in a dashboard or a conversation.

How does it work?

Peanut imports your store data, calculates profit with deterministic code, and compares recent weeks with earlier ones. When a change is statistically meaningful and worth money, it opens a case file with the evidence and tells you. The AI explains the results; it never does the arithmetic.

How do I connect Shopify?

Through a Shopify app with read-only permissions for orders, products and costs. The Shopify connection is in development; today you can explore everything with a demo store. Join the waitlist to be invited when it opens.

How does it calculate profit?

Contribution profit = net sales − product costs − shipping − fulfilment − payment fees − advertising. Fixed costs you enter give an estimate of profit after fixed costs. That isn’t accounting net profit, and Peanut always says so. Missing costs are flagged, never treated as zero.

Does it access customer information?

Peanut needs orders and refunds but not who your customers are. It doesn’t store customer names, emails or addresses, and we’ll request the minimum Shopify permissions needed.

Can it change my store automatically?

No. Peanut can suggest actions, draft an email for you to review, or model a price change, but it never changes prices, ads, shipping settings or anything else in your store, and never sends messages on your behalf without your approval.

Does it work with Apple Messages?

Not yet. Messaging businesses through Apple’s Messages app requires Apple Messages for Business, which needs Apple’s approval and an authorised messaging provider. We’re preparing for it, and we won’t describe it as available until it genuinely works. In-app chat works today; email works when enabled; WhatsApp is in development.

Can it send me charts and reports?

Yes. Peanut renders charts, product comparisons, profit bridges and weekly summary cards from your data, and can send them in a conversation or by email. Images of your financial data are only reachable through short-lived private links.

Does it support multiple stores?

Not yet. Each account currently covers one store. Multiple stores are planned for the Pro plan.

How is my business data protected?

Each store’s data is kept separately, every change is written to an audit log you can review, and you can delete everything Peanut stores at any time. We don’t sell your data or use it to train AI models.

Stop wondering where
your profit went.

Let Peanut investigate your business so you can get on with running it.

We’ll email you to confirm, and only about Peanut. Unsubscribe any time.

Or explore the live demo →