Media Agencies: How to Prove Real ROAS to Your Clients

Every Agency’s Dilemma

You are presenting monthly results to your client. Google Ads shows a 5.2x ROAS. Meta claims 4.8x. TikTok reports 3.1x. Your client pulls out a calculator, adds up the revenue claimed by each platform… and gets twice their actual revenue.

The inevitable question: “So what is the real number?”

If you don’t have a solid answer, you have a credibility problem. In a market where advertisers increasingly challenge their agencies, transparency on real numbers is no longer optional — it is a competitive advantage.

Why the Numbers Never Add Up

Every Platform Claims the Conversion

The fundamental problem: every ad platform takes credit for a conversion as long as it had a touchpoint in the journey. A user sees a Meta ad, clicks a Google ad, then converts — Meta counts the conversion (view-through), Google counts it too (click-through).

Result: the sum of platform conversions always exceeds actual conversions. It is mathematically inevitable.

Attribution Windows Differ

Platform	Default Window	What It Counts
Google Ads	30-day click	Last Google click
Meta Ads	7-day click + 1-day view	Click AND view
TikTok Ads	7-day click + 1-day view	Click AND view
Google Analytics 4	30 days	Last non-direct click
Reality	—	What the client actually deposited

When Meta counts a “1-day view” conversion (the user saw the ad but did not click), did Meta really generate that sale? The answer is: maybe, but you cannot state it with certainty.

View-Through Inflates Results

View-through (post-impression) conversions are the most slippery metric. Meta is particularly aggressive here: if a user simply saw an ad in their feed (without any interaction), and converts within 24 hours, Meta claims the conversion.

On broad audiences with high impression volumes, view-through can represent 30-50% of Meta’s claimed conversions. That is enormous — and largely overstated.

5 Solutions to Prove Real ROAS

1. Unified Dashboard with Deduplication

The first and most impactful step: centralize all data in a single dashboard with deduplication logic.

The principle:

Source of truth = actual sales data (CRM, e-commerce backend, ERP)
Attribution = unified logic (last non-direct click, or custom model)
Comparison = platform ROAS vs real ROAS displayed side by side

Example structure:

Channel       | Spend    | Platform Conv. | Deduplicated Conv. | Platform ROAS | Real ROAS
Google Ads    | $6,000   | 120            | 95                 | 4.8x          | 3.8x
Meta Ads      | $3,500   | 85             | 52                 | 5.1x          | 3.1x
TikTok Ads    | $2,500   | 40             | 28                 | 3.2x          | 2.2x
──────────────────────────────────────────────────────────────────────────────────────
Total         | $12,000  | 245            | 175                | —             | 3.2x (MER)

2. Server-Side Tracking for Reliable Data

Without reliable input data, no attribution model will produce accurate results. Server-side tracking solves a large part of the problem:

Bypasses ad blockers: data routes through your server, not the browser
Resists ITP: server-set cookies with extended lifespan
Feeds Conversions APIs: Meta CAPI, Google Enhanced Conversions, TikTok Events API
Improves matching: first-party data (hashed email, phone) sent server-side

Measured impact: +25-35% conversion recovery compared to 100% client-side tracking.

3. Incrementality Testing

This is the gold standard for proving a channel’s real value. The principle: measure what would have happened WITHOUT the advertising.

Geo-lift test: pause advertising in selected geographic regions for 2-4 weeks and compare with control regions. If sales drop 15% in regions without ads, then advertising generates 15% incremental.

Holdout test: exclude a percentage of the target audience (10-20%) from all ad exposure and compare conversion rates.

These tests are the only ones that provide causal proof (not just correlation) of advertising impact.

4. Blended ROAS (MER)

The Marketing Efficiency Ratio (MER) is brutal in its simplicity:

MER = Total business revenue / Total ad spend

No attribution window. No deduplication. No view-through debate. Simply: how much the business took in vs how much it spent on ads.

This is the number your client’s CFO understands. And it is the most honest number you can present.

MER does not replace attribution (it does not tell you where to allocate budget), but it provides the big-picture view that per-channel metrics cannot offer.

5. CRM-Linked Attribution

For advertisers with longer sales cycles (B2B, lead gen, subscriptions), connecting ad data to the CRM is essential:

Track GCLID/FBCLID through to final sale in the CRM
Import offline conversions into Google Ads and Meta
Calculate ROAS on actual customer value (not just the first transaction)

Impact: for one B2B client, switching from “lead” attribution to “closed deal” attribution revealed that Google Ads’ real ROAS was 40% higher than standard tracking showed — because Google generated higher-quality leads.

How to Present This to Clients

Transparency Builds Trust

Paradoxically, showing that the real numbers are lower than platform numbers strengthens the agency’s credibility. Here is why:

The client already knows the numbers are inflated (or suspects it)
An agency that shows “real” numbers proves its integrity
Optimizations based on real data produce better results
The client retains the agency longer (trust beats promises)

ROAS Reconciliation Report Template

Here is the structure we recommend for a monthly client report:

Page 1 — Overview

MER (Blended ROAS): month-over-month trend
Total revenue vs total spend
6-month trend line

Page 2 — By Channel (Platform Data)

ROAS per platform (native figures)
Note: “These figures include cross-platform double-counting”

Page 3 — By Channel (Deduplicated Data)

Deduplicated ROAS per channel
Platform vs real gap
Each channel’s contribution to overall MER

Page 4 — Actions and Recommendations

Suggested budget reallocation
Planned incrementality tests
Ongoing tracking optimizations

The Result: +28% ROAS Post-Audit

Across tracking audit engagements we have delivered for media agencies, the finding is consistent: after implementing reliable tracking and deduplicated reporting, optimizable ROAS increases by an average of 28%.

Not because campaigns got better, but because algorithms finally receive complete data and can optimize correctly.

Where to Start

Identify the gap: compare the sum of platform conversions vs actual sales
Implement a MER: the simplest and most honest reference metric
Audit your tracking: verify that collected data is reliable before analyzing it
Centralize in a dashboard: one source of truth, with deduplication
Plan an incrementality test: to prove the causal value of each channel

Managing media budgets for clients and the numbers never match? We audit your tracking setup and build reporting that shows the real ROAS — the one that holds up to tough questions.

Discover our media tracking audit →