NorthStar

Know you're going to miss the number while there's still time to fix it.

NorthStar sets a target on every metric that matters, watches the full funnel against your actual orders instead of platform claims, and puts the gap, and what caused it, in your inbox at 8am.

$15B in revenue monitored. $10K+ in recoverable spend in week one, or you don't pay.

$15B in revenue monitored160+ connectorsSOC-2 certified
ThirdLoveThe Beard StruggleTractor Supply

Are you going to miss the number, and find out too late to fix it?

Most teams learn they've missed on the fifth working day of the following month, from a spreadsheet someone assembled the night before. By then the period is closed and the only thing left to do is explain it.

What's going wrong

The report is built by hand

Someone on your team pulls the same numbers from five to nine tools every morning and pastes them into a sheet. The version that reaches the board is the version that got assembled last night.

What's going wrong

Nobody is looking at the same number

Every ad platform reports its own performance against its own attribution window. Meta's number, Google's number and the Shopify admin's number disagree, and every budget decision sits in the gap between them.

What's going wrong

Nothing is watching between logins

Dashboards are pull. They wait for you to open them and ask the right question. A checkout that broke on Tuesday afternoon stays broken until someone thinks to check that particular chart.

The gap isn't measurement. It's that measurement arrives after the period it was measuring. NorthStar moves the finding to the morning it happened.

One report. Every morning. Already analyzed.

NorthStar connects the full funnel, reconciles it against your actual orders, holds it against targets you committed to, and tells you what moved and why, before you open anything.

160+ connectors

Shopify, Meta, Google Ads, TikTok, GA4, Klaviyo, Adobe Analytics, site, checkout and OMS, in one view.

Reconciled against orders

Every platform-claimed conversion checked against what actually sold.

A target on any metric

Daily to quarterly, with pacing against the forecast landing point.

One semantic layer

The same metric means the same thing in the report, the dashboard, Slack and your warehouse.

Delivered, not displayed

Email or Slack, at the time you set.

This is the report. Read it.

Not a screenshot of a dashboard. The actual thing that lands at 8am, rendered here in full. Every figure below is a sample, anonymized from a live account.

Daily report · Tuesday 18 August 2026 · compared to same weekday, last year (NRF 4-5-4)

Summary

Net revenue closed the day at $412,880, 5.8% below the comparable weekday last year, against a month-to-date target that now requires $448k/day to reach. The shortfall is not an acquisition problem: sessions and new-customer volume are both ahead. It sits in checkout, where completion fell 3.1 points, concentrated almost entirely in mobile iOS traffic following the express-pay configuration change on 16 August.

Written after the analysis, not before. If nothing moved, it says so, and a quiet day reported as quiet is what makes a loud one credible.

KPI
Net revenue
$412,880
baseline $438,120
-5.8%z -2.4
Blended MER
2.41
baseline 2.68
-10.1%z -3.1
Checkout completion
68.2%
baseline 71.3%
-3.1ptz -2.7
New-customer CAC
$61.40
baseline $52.90
+16.1%z +2.9

Every number carries its baseline, the absolute delta, the percentage delta and a z-score. The z-score is why a 4% drop on a noisy metric does not get your attention and a 4% drop on a stable one does.

Callouts
  • HighCheckout completion is $18.4k of the $25.2k gap. Cart-to-checkout is flat, so the drop-off is inside the checkout step itself. Express-pay availability changed on 16 August; the decline starts the same day.
  • MediumGoogle Shopping new-customer CAC up 22%. Shopping absorbed 31% of new spend. Brand search CAC is unchanged, so this is Shopping-specific rather than a blended auction shift.
  • LowKlaviyo flow revenue within normal range. Lifecycle revenue is 1.2% under baseline, inside variance.

Ranked high, medium, low. Three callouts, not thirty. Ranking is the product's job, not yours.

Drilldown
Mobile · iOS62.4% vs 71.8%
Mobile · Android69.1% vs 70.2%
Desktop77.3% vs 76.9%
Tablet and other71.0% vs 71.4%

iOS carries the entire decline and is 54% of checkout sessions, so a 9.4-point drop there produces the 3.1-point blended figure on its own.

The metric broken down by the dimension that actually explains it, with the long tail collapsed into one bucket and a sentence saying what the split means.

Context
  • Marketing: back-to-school creative refresh went live 17 Aug across Meta and TikTok. Spend +11%, sessions +8%. Acquisition is not the constraint.
  • Funnel: sessions +8.1%, add-to-cart +6.4%, cart-to-checkout flat, checkout completion -3.1pt. The break is one step wide.
  • Product mix: bundle SKUs 34% of orders vs 31% baseline. Average order value +2.1%, partially offsetting the volume loss.
  • Annotations: "Express pay reconfigured on mobile" logged by your team, 16 Aug 14:20.

What else was happening: marketing activity, funnel state, product mix and your own annotations, so a drop arrives with a cause attached rather than as a chart.

Generated
08:00
Time to produce
41s
Reconciled against
14 sources
Compute cost
$0.38

What the report cost us to produce, on every run. We don't bill you per analysis, so there is no reason to hide it.

Compared to what?

A number is only a signal against the right baseline. You choose it once, and every report and every target uses it.

Default

Same weekday, last year

On the NRF 4-5-4 retail calendar, so Black Friday compares to Black Friday and not to the 24th of November.

Operational

Previous seven days

For reads where last year is irrelevant: a launch, a new channel, a promo that has no precedent.

Finance

Same date, last year

For the people who need the calendar to line up with the ledger rather than with the trading week.

A target on every metric, not just on revenue.

Revenue targets tell you that you're behind. They don't tell you which of the eleven things feeding revenue is the reason.

Any metric

CVR, new-customer CAC, checkout completion, repeat rate, contribution margin: anything in the catalog carries a target.

Any cadence

Each target runs daily, weekly, monthly or quarterly and reports as on track, off pacing, or firing.

Pacing

A cumulative target tells you where you've got to. A pacing target tells you where you'll land if nothing changes, the only version you can still act on.

Target coverage (example account)
38of 62 funnel metrics have a committed target

Coverage is a number, because a taxonomy with gaps in it can group but not govern.

Net revenue · monthlyOff pacing-5.8% to plan
New customer CAC · weeklyFiring+16.1% to plan
Checkout completion · dailyFiring-3.1pt
Repeat rate · quarterlyOn track+1.1pt

Defined once. The same number everywhere.

Every metric in NorthStar is defined in one place and resolved through a governed semantic layer, so the number in your 8am report, the number in the dashboard, the number in Slack and the number your analyst pulls into the warehouse are the same number, computed the same way.

Axis 1

By funnel

The ordered steps, in the order they happen.

Axis 2

By source

Shopify, Meta, Google Ads, TikTok, GA4, Klaviyo, Adobe Analytics, your OMS.

Axis 3

By category

How your team already talks about them, not how we'd file them.

Axis 4

By pillar

Revenue and profitability, acquisition, conversion, active customers, lifetime value.

Now you can stop arguing about whose number is right, because every metric resolves through one governed definition, so the meeting is about what to do and not about the spreadsheet.

Your numbers, inside Claude.

Connect NorthStar as an MCP server and ask your own data questions from Claude, ChatGPT or anything else that speaks the protocol. Read-only, OAuth, scoped per user: access is resolved on every request against the same permissions you set in the product.

No export. No copy of your data sitting in someone's chat history.

// your assistant, your data

> Which step of the funnel is driving yesterday's revenue gap?

Checkout completion. 68.2% vs 71.3% baseline, worth $18.4k of the $25.2k gap, concentrated in mobile iOS since the express-pay change on 16 August.

Five workflows. They run whether you log in or not.

Everything above is produced by one of five workflows. Each has a trigger, a cadence and an output you can point at. Nothing runs because you remembered to open a tab.

WorkflowRunsProducesAvailable on
Morning reportDaily, at the time you setReconciled KPIs against your baseline, ranked callouts, and the dimension that explains the biggest one. Email or Slack.Every plan, including free
Target pacingPer target: daily, weekly, monthly or quarterlyOn track, off pacing, or firing, against the forecast landing point rather than today's value.Starter and above
Trend watchContinuously, per metricA deviation from baseline at the sensitivity and direction you chose, with its magnitude.Trend forecasting add-on, Growth and above
Period comparisonOn your comparison basisAny metric that moved further than your threshold against the previous period or last year.Starter and above
AskOn demand, from Claude or ChatGPTAn answer computed on the same governed metric definitions. Read-only.Starter, with limits. Unlimited Ask AI add-on removes them

An agent writes the morning report. The other four are a rule engine on a schedule. We'd rather tell you which is which than call all five agents. None of them spend your money.

How it works

Step 1

Connect your store

Shopify first, then your ad accounts and whatever else you run: Klaviyo, GA4, Adobe Analytics, your OMS.

Step 2

Pick your baseline

Same weekday last year on the NRF 4-5-4 calendar, previous seven days, or same date last year. Every report and every target uses it.

Step 3

Commit your targets

Set a goal on any metric in the catalog, choose cumulative or pacing, and pick the cadence. Coverage shows how much of the funnel is governed.

Step 4

Choose where it lands

Email, Slack, or both, at the time you set. Then it runs whether you log in or not.

Against the alternatives.

Taken from each vendor's own public pages. Verified 19 August 2026. "None found" means not published on their public pages on that date, not that it doesn't exist.

CapabilityNorthStarTriple WhalePolarNorthbeam
Platform-claimed conversions reconciled line-by-line against ordersYesNone foundNone foundNone found
Sample report published before you buyYesNone foundNone foundNone found
Targets with pacing on any metricYesNone foundNone foundNone found
Full funnel beyond ads: site, checkout, lifecycle, OMSYesPartialYesAds only
MCP endpointYesYesYesNone found
Charged per credit or per analysisNoYesNoNo
Money-back guarantee$10K+ in week one or you don't payNone foundNone foundNone found
Entry price publishedYes$0 free tier, $179+/moNot published$1,500/mo

"Before Consequential our discovery of revenue impacting issues was lagging. Now we are alerted immediately and can react quickly."

James Bertram, Head of Digital, Abbott Lyon

Connect a store. The first report follows.

We find $10K+ in recoverable spend in your first week, or you don't pay. Whatever you keep an agency for, it shouldn't be pulling the numbers together.

Frequently Asked Questions

Then you already have the numbers. What you don't have is a target on them, a comparison that survives a retail calendar, and something that has already done the analysis before you open it. Reporting tools tell you what happened. NorthStar tells you what it means and what it costs you if nothing changes.
An assistant answers the question you thought to ask. NorthStar runs whether you log in or not, against targets you committed to, and tells you the thing you didn't know to ask about. You can also ask it questions, by connecting it to Claude, but that's the second feature, not the first.
Own pixel, server-side collection, an identity graph, and every platform-claimed conversion reconciled against actual orders. One semantic layer, so the same metric means the same thing in the report, the dashboard and your warehouse. SOC-2 certified.
No. Shopify, Meta, Google Ads, TikTok, GA4, Klaviyo, Adobe Analytics, site, checkout and OMS, across 160+ connectors. A checkout error or a dead lifecycle flow surfaces the same way a fading creative does.
Not on this product. NorthStar diagnoses and forecasts. It doesn't touch your ad accounts.
Priced on your annual GMV, detected from your store. See the pricing page for the tiers.

Built to Replace Reactivity

If you're tired of reacting to revenue drops after the fact, Consequential gives you the edge.