Upstream for Shopify

More of your best customers.

We turn the data you already have into profitable actions — integrated into your existing systems, so you can optimize marketing spend and get more of your best.

Action Queue

Don't just read the diagnosis. Apply the fix.

Every data story pairs with a one-click move: pause a product that's losing money — live in Shopify the second you click — or queue a budget shift to your ad accounts. Nothing executes without your approval, and every move can be undone.

Take Action

The actions you take move your business from bleeding cash to compounding it.

Extra profit in 90 days

+$0

1Shift Budget

Shift $3,500/mo from Flash Sale Campaign → Signature Bundle

96% of Flash Sale buyers never come back — Signature Bundle keeps 46% of them. Move the money.

From: The Lost Causes · Cold Prospecting
+$14,000
2Shift Budget

Shift $2,000/mo from Retargeting → Brand Search

You're paying more for Retargeting and getting less repeat business than Brand Search. Switch it.

From: The Expensive Regulars · Retargeting
+$6,200

2 of your queued moves shown · more waiting

Cumulative Net Margin by Cohort

Live · After Product Cost

Each line represents one group of buyers. If their line is above $0 = profit, below $0 = losing money. Every number is counted after the product cost, shipping, returns, and ad spend are paid.

1,840

Server-Verified Orders

Traced to the exact ad they clicked

38.4%

Blended Post-COGS Margin

Profit after product, shipping & fees

42 Days

Second-Order Payback Latency

Days until their 2nd order pays back the ad

Signature Bundle(Buyers come back · pays back the ad in 42 days)20% Flash Sale(One-time buyers · never earns back the ad cost)
Day 0Day 42Month 3Month 6Month 12−$4k$0+$4k+$8k+$12k+$16k$0.00 BreakevenOrder 2 Payback

Signature Bundle

Spend: $4,500 · 100 Customers

46% buy again within 42 days

+$14,000 total profit over 12 months

3.11× — profit is 3.11× the ad cost

20% Flash Sale

Spend: $3,500 · 140 Customers

4% buy again (96% never return)

−$1,274 total loss over 12 months

0.64× — profit is less than the ad cost

Action Required · A group of buyers is losing money

+$14,000 extra profit expected in the next 90 days

Shift $3,500/mo from Flash Sale Campaigns (96% never buy again) → Signature Bundle

The Gateway SKU Bridge

Which first order predicts a repeat buyer

The buyer who pays full price and comes back for more.

  • • First product they buy: Trailhead Backpack
  • • How soon they buy again: 42 days
  • • Profit from each buyer in a year: $153
  • • How often they send it back: < 2.1%

Ready to push as a lookalike seed to your ad platforms — one click, whenever you want it.

Attribution Engine

First-Party Click-to-Order Ledger

A record of which ad click led to each order — not guessed in the browser.

  • • Ad type: Paid Social (lookalike audience)
  • • Ad click tracked directly, not modeled
  • • Cost to get each first buyer: $45.00
  • • Profit is 3.11× the ad cost over 12 months

Flags the ads mostly bringing one-time buyers who never come back — so you can cut them.

— The Segment Matrix

See which customers make you money — and which ones don't come back.

Every group of customers lands in one of four boxes. Each box tells you the one move to make: re-invest in it, fix it, or cut it.

This is the same buyer groups as the curve above, plotted a different way: LTGP:CAC against buy-again rate, so you can see which groups fund the business and which drain it. As a group's numbers change, it can move from one box to another — a group scoring well today isn't locked in that box forever.

Cohort Profitability × Retention Matrix

Side-to-side = profit per ad dollar (higher is better). Up-and-down = how often buyers come back. Bubble size = how many buyers are in the group.

0123456Profit per ad dollar · LTGP:CAC →00.511.52How often they buy again · Repurchase Multiplier →Compounding ZoneOne-Hit WonderThe TrapThe Drain
Bubble size = buyers in the group
100
1,000
5,000
Less profit per ad dollar · buyers come back

The Trap

They love the product and buy repeatedly, but acquisition costs are too high. Fix your creative or cut the channel.

More profit per ad dollar · buyers come back

The Compounding Zone

Scale spend aggressively. These customers print cash and stick around.

Less profit per ad dollar · buyers don't come back

The Drain

Cut immediately. These customers lose money on day one and never come back.

More profit per ad dollar · buyers don't come back

The One-Hit Wonder

Profitable on day one, but dead afterward. Good for cash flow — dangerous to scale without retention hooks.

What's inside a segment

Every bubble above is made of real buyers. Click a bubble in the chart, or pick a group below, to see where they live, who they are, and what they buy.

Example store · illustrative, not averaged across merchants

The Word-of-Mouth Winners · Referral · 950 buyers

Where they are

United States83.8%
Canada9.8%
United Kingdom6.4%

Age & gender

From a connected Meta/Google Ads account, once each platform clears our review

25-34 · female410 buys
35-44 · female330 buys
25-34 · male140 buys
45-54 · female120 buys

What they buy

Real purchase data — what this group actually bought and when. Not a lifestyle/values profile: we don't run surveys, buy panel data, or license third-party psychographic datasets.

Sweaters42%
Outerwear28%
Accessories18%
Footwear12%

Business buyers

Channel-presence only — this example store has B2B/wholesale channel orders, and some of this segment's buyers also appear there, but no company name, size, or industry is captured.

B2B orders store-wide

14

In this segment

6

Your best customers are already in your data. Typically scattered across ad dashboards, order histories, and cluttered spreadsheets.

Upstream connects your true cost of goods, logistics, and even return costs to every ad click. You see which campaigns build compounding repeat buyers—and which ones subsidize one-time deal hunters who never come back.

See how it works

Limited early access · Real data, not modeled · No code required

Want the full math? See how it works.

A Fox & Otter Data product