Facebook Ads
×
Papercrane
Facebook Ads says 48 sales. Shopify says 31.
Neither number is lying; each counts by its own rules. Papercrane joins Facebook Ads with Shopify, GA4, and Google Ads, shows you where the gap comes from, and builds live dashboards on the numbers you can bank.
Free to connect and ask. Paid plans start when you publish a dashboard. See pricing
Where “48 purchases” comes from
Ads Manager's purchase count is a claim assembled from five different kinds of counting, each documented, each defensible, none of them the same thing your store counts. Papercrane doesn't add a fifth opinion; it puts Meta's claim, GA4's count, and Shopify's orders on one screen and traces the disagreement day by day.
48 purchases · Meta Ads Manager
19 Clicked the ad, bought, order is in Shopify too
7 day click window, observed
9 Saw the ad, never clicked, bought within a day
1 day view window
8 Estimated to cover what tracking can't see
modeled conversions
7 Bought this week, counted on last week's click date
reported at click time
5 Also claimed by Google Ads or your email platform
overlapping windows
Shopify counted 31 orders the same week. Both totals are internally consistent. Neither is the answer to “what did my ads make.”
Example figures; the mechanisms are Meta's documented attribution rules. Connect your account and the agent runs this dissection on your real weeks.
You already run this reconciliation by hand
The 7am tab shuffle
Ads Manager in one tab, Shopify in another, yesterday's spend typed into the MER spreadsheet. Every morning, before coffee.
The number that moved overnight
Last week's ROAS changed since you last looked, because conversions land on click dates and the modeled fill keeps settling.
The question you can't answer cleanly
“So are the ads working?” Your co founder asks it with a bank balance open, and you answer with three numbers that disagree.
Dashboards built on the numbers that agree
Spend is a fact. Orders are a fact. Papercrane's AI agent joins them into live dashboards that skip the argument: blended MER across Meta, Google, and TikTok; spend pacing against budget; creative and audience performance from Meta's own reporting, side by side across every ad account you run.
Blended MER, daily
All revenue over all ad spend, every platform on one line, so the morning spreadsheet has nothing left to do.
Creative fatigue watch
Frequency climbing while click through falls, flagged per creative while there is still time to rotate.
Every ad account, one screen
Connect each account and address them by name, whether that's two brands or twelve clients.
Ask the attribution question anyway
Ask “which ad really made the sale” and the agent won't invent another model; it gathers what each platform claims, sums the claims against your actual orders, flags the orders more than one platform takes credit for, and reads your attribution window settings back to you. You end up knowing why your tools disagree, which is more than any “true ROAS” ever told you.
The questions the tabs can't answer
What's my blended MER this month, all revenue over all ad spend?
What did Meta claim versus what Shopify recorded, day by day?
Which creatives are fatiguing: frequency up, click through down, last 14 days?
How did prospecting versus retargeting spend shift this quarter?
Which audiences' CPMs rose fastest since the holidays?
Show both ad accounts' pacing against this month's budgets, one screen.
Every answer comes back as a chart on a live hosted dashboard your team can open tomorrow.
Three ways to live with the disagreement
| The job | Ads Manager + exports | An attribution pixel ($149 to $1,290/mo) | Facebook Ads + Papercrane |
|---|---|---|---|
| Know spend against real revenue | Two tabs and a spreadsheet, every morning | Its own third number, tracked its own way | Blended MER from Meta spend and Shopify orders, live joins Shopify |
| Explain why the numbers differ | You, improvising, on a client call | Rarely; its model is the product | The dissection above, on your real weeks, down to the orders claimed twice |
| Creative reporting across accounts | Export per account, paste, repeat | Varies by plan | Ask once, every account answers by name |
| Share with the team or a client | Screenshots into a deck | More seats, more cost | A live dashboard link they can open anytime |
| When a number looks wrong | Rebuild the export | File a support ticket | Errors route back to the agent, which repairs the dashboard |
1. Connect Facebook Ads with read access you approve on Meta's consent screen
2. Ask a question
3. The agent plans the queries, joins the sources, and builds a hosted dashboard
4. It stays live, and errors route back to the agent
Read access to the ad reporting you already have. No tracking added to your site, nothing installed in your store.
When you don't need us
If your day is bids, budgets, and creative tests inside one ad account, stay in Ads Manager: for steering Meta's own auction, Meta's numbers are the right ones, and no outside tool improves them. If you want journey level attribution with its own tracking of your site visitors, Triple Whale and Northbeam run their own measurement and are built for exactly that; Papercrane doesn't track your visitors and won't pretend to. We are worth paying for when the question spans tools, real revenue over real spend, claims compared honestly, and when the answer needs to live somewhere your team or your clients can watch it.
Questions advertisers ask us
Most Facebook Ads connections are joined with Shopify for the real orders, GA4 for the session count, and Google Ads for the other half of the budget.
Connect your ad account, ask what Meta claimed versus what your store recorded, and watch the answer come back as a dashboard.
Free to connect and ask. Paid plans start when you publish a dashboard. See pricing
Papercrane
Integrations: