For media planners & performance strategists

You have to recommend what to do with the budget.
The information isn't always perfect.

AdCrux helps you plan media investment using each brand's real history of spend and results — a clearer starting point than intuition or generic assumptions.

Set up in minutes · Works with the spend and results data you already track

The situation

Meta says one thing. Google says another. The CRM says a third.

And you still have to recommend what to move. That's the job — regardless of whether the data lines up.

Google Ads 87 leads
Meta Ads 74 leads
Sum of claims 192 leads
Actual leads generated 120 leads

Every source tells its own version: Google says 87, Meta says 74, your CRM says 120 total. Each platform is just measuring a different slice of the same activity.

Looking at channels one at a time doesn't show how the mix works as a whole, and that's what you need before recommending the next move.

This isn't about which platform is lying. It's about needing a single, comparable read before you can recommend anything with confidence.

What every planning conversation runs into

One channel looks like it's working The dashboard is green and the ROAS looks fine. But looking good in isolation isn't the same as being worth the next dollar.

Another looks expensive High CPA on paper. But expensive compared to what: the rest of the mix, or what it used to cost?

A third one, nobody's sure It might be driving new demand. It might just be riding along on what the other channels started. Nobody on the team can say for sure.

And the client still wants an answer "What should we do with the budget" doesn't wait for certainty. The recommendation still has to be more than a hunch.

A recommendation is only as good as what it's compared against

Without a way to compare channels side by side, the decision ends up depending on who argues best in the room, not on the numbers.

The recommendation doesn't survive the first hard question

Answering "why this channel and not that one" takes something more comparable than a quick dashboard read.

Every stakeholder is comparing different numbers

The client is looking at one report, the media team another, finance a third. Nobody's arguing from the same baseline.

Testing becomes a last resort, not a plan

Testing without a clear read going in amounts to "let's try something and see," with nothing to compare the result against.

The same debate repeats every planning cycle

Each quarter starts the argument over from scratch, with no record of what was tried last time.

Plan the whole mix, not one channel at a time

AdCrux uses each brand's spend and results history to help you compare scenarios, understand contribution by channel, and plan tests.

Compare scenarios: See what a reallocation would do to the whole mix before you propose it.

Understand contributionSee how much each channel is adding, separate from baseline and from what the other channels are already driving.

Plan the test: Turn "let's try something" into a specific, bounded test with an expected range to compare the result against.

Build on what you knowEvery run, scenario, and test adds to what's already known about this brand — so the next recommendation starts ahead of the last one.

Contribution by channel
Google Ads
31%
Meta
22%
Email
19%
YouTube
12%

Organic baseline
16%

Modeled from the uploaded spend and results, not an industry benchmark

The platform

See inside AdCrux

The workspace for comparing scenarios, reading contribution by channel, and planning the next test.

One comparable read across the whole mix.

app.getadcrux.com
One comparable read across the whole mix
Acme Agency · Q1 2024
Performance Overview
Jan 1 – Mar 31, 2024
Model Leads
120
Independent estimate
Total Spend
$84,320
Across all channels
Blended ROAS
3.4×
Revenue / spend
Contribution by channel (model estimate)
Google Search
37%
Meta Ads
26%
Email
22%
YouTube
15%
Organic baseline
38 leads
Would occur without media
Incremental (paid)
82 leads
Driven by your spend
What each platform claims, next to what the mix model attributes
Attribution Gap Analysis
+60% overreported
Channel Platform claim Model (MMM) Gap
Google Search 87 leads 31 leads −64%
Meta Ads 74 leads 22 leads −70%
Email 31 leads 19 leads −39%
YouTube — leads 15 leads
Total 192 120 −38%
Platform vs Model
Platform total
192
Model total
120
Platforms are claiming credit for 72 leads that didn't happen, or happened organically.
Where each channel sits on its own return curve
Saturation Curves
2 channels past optimal
Google Search High · 78% saturated
Email Moderate · 42% saturated
Meta Ads High · 85% saturated
YouTube Moderate · 31% saturated
The reallocation, compared to what you're spending today
Budget Optimizer
+18% projected leads
Channel Current Recommended Change
Google Search $5,000 $7,200 ↑ +44%
Meta Ads $8,000 $5,100 ↓ −36%
Email $2,000 $3,400 ↑ +70%
YouTube $1,500 $1,300 ↓ −13%
Same budget
$16,500
Projected leads
142 vs 120
CPA improvement
−15%
Compare reallocations side by side before recommending one
Scenario Planning
Test decisions before committing
Conservative
-10%
108 leads · −10% budget
Balanced
+0%
120 leads · Current budget
ACTIVE
Optimized
+18%
142 leads · Same budget
Aggressive
+30%
156 leads · +25% budget
Leads comparison
Conservative
108
Balanced
120
Optimized
142 ← active
Aggressive
156
vs current budget
+22 leads
Optimized scenario
CPA change
↓ −15%
More leads, same spend
Turn the recommendation into a bounded test
Incrementality Test Planner
Feasibility: Good
Channel
Meta Ads
Objective
Increase spend
Risk tolerance
Moderate
Duration
4 weeks
Expected outcome
132–148 leads
Expected over the 4-week test window
33–37 / week
Baseline weekly rate
Guardrails
  • Increase capped at 40% — core channel, not a full pause
  • Recommends reverting the change if the read is still inconclusive by the end of the window
  • Compares the result against an expected range, not a single guess
Save as Test Plan →
The record of what was tested and what happened
Test Tracker
Brand-level · survives model re-runs
1
Planned
1
Running
3
Completed
0
Cancelled
Recent tests
Test Channel Status Verdict
Increase Meta +40% Meta Ads Completed Supported
Pause YouTube YouTube Running — pending
Reduce Email -20% Email Planned — not started
Every completed test adds to what you know about the channel, so the next plan starts smarter.
How much to trust this read
Model Diagnostics
High confidence
Confidence Score
82
/ 100 · High confidence
Model fit metrics
R² (variance explained) 94.7%
MAPE (prediction error) 4.2%
Holdout RMSPE 6.8%
Actual vs predicted
Actual
Predicted
Confidence breakdown
Data quality Good · 26 weeks
Channel variation Sufficient
Negative coefficients None detected

Sample dashboard · Your data stays in your account · Export to PDF

How it works

From history to a recommendation you can defend

Every step below runs on the spend and results you upload.

01

Upload the history

Spend and results by channel and week, straight from what you upload.

CSV upload
02

See how the mix works

The model reads the full mix together: contribution, saturation, and baseline for every channel.

~30 seconds
03

Compare scenarios, pick one

Try different reallocations side by side, see the projected outcome of each, then save the one you'd recommend.

Compare & save
04

Plan the test, add to the record

Turn the recommendation into a bounded test, log what happened, and it becomes part of this brand's planning history.

Build the record

What to increase. What to decrease. What to hold. What to test before scaling.

Four calls, each grounded in the uploaded numbers for this brand.

What to increase

Channels with room left in the mix, worth the next dollar based on how they're performing now.

What to decrease

Once a channel passes the point where more spend stops adding, it's a candidate to scale back, even if the platform report still looks fine.

What to hold

At the current level, some channels are simply doing what they should. No case yet to move either way.

What to test before scaling

A specific, bounded test to validate the move before you commit the full recommendation.

Not enterprise-only. Not a black box.

Built for the person who has to make the call.

Traditional approach
AdCrux
MMM consulting: $50K+ and 8 weeks
Results in under 5 minutes
Platform dashboards: each silo optimizes itself
One model across all channels
Black-box tools: outputs you cannot explain
Transparent model diagnostics

Accessible from day one

Marketing Mix Modeling used to require data scientists. AdCrux makes it available to any team with a CSV file.

Independent by design

Built independently from any platform attribution model. No pixel, no SDK, no API key connecting back to Google or Meta.

Transparent outputs

Every result comes with a confidence score and model diagnostics. You know when the model is reliable.

Pricing

Less than 1% of your media budget

If you're already planning budget for $10K–$30K/month in spend, this pays for itself the first time it changes a recommendation.

Starter
$249/mo
~1% of a $25K/mo media budget

For brands and small teams running paid media and looking for their first independent view of performance.


  • 1 active model
  • 4 analysis runs/month
  • Full analysis suite, incl. Test Planner & Tracker
  • Budget optimizer & scenarios
  • Confidence score + diagnostics
  • AI insights
Book a demo
Agency
$999/mo
For agencies managing multiple clients

For agencies that plan budget, defend recommendations, and track tests for their clients — not just a one-time measurement report.


  • Up to 10 active models
  • 40 analysis runs/month
  • Everything in Growth
  • Client → Model workspace
  • 5 user seats
  • Dedicated support
Book a demo

All plans include a 14-day free trial. Cancel anytime. Need more? Contact us about Enterprise.

FAQ

Common questions

See what your own numbers say about the next move.

Book a demo — we'll walk through it using a case close to your own.

Book a demo →

Set up in minutes · Works with the spend and results data you already track