# How To Audit Your Meta Ads Agency: A Step-By-Step Method

> Tower33 breaks down how to audit your Meta Ads agency step by step — from fee-adjusted ROAS to inflated reporting. Get clarity and act on it.

**Published:** 2026-09-29 | **Updated:** 2026-10-31 | **Author:** victorsoares
**URL:** https://tower33.com/articles/audit-meta-ads-agency-performance/
**Type:** post

**Categories:** Meta and Microsoft Ads

![How To Audit Your Meta Ads Agency: A Step-By-Step Method](https://tower33.com/articles/wp-content/uploads/sites/107/2026/09/1789065377602-d9cacda036a4-1024x572.webp)

---

## Content

*Written by: Anthony Miller, Founder, Tower 33 | Last updated: October 23, 2026*

## Key Takeaways

- Auditing a Meta ads agency starts with outcomes and economics before campaign settings or creative reviews.
- Fee-adjusted ROAS and contribution margin calculations show whether reported performance stays profitable after all costs.
- Measurement integrity depends on reconciling Meta-reported conversions against CRM or e-commerce records and verifying attribution windows and Conversions API coverage.
- Agency reporting quality improves when you check for revenue lines, qualified-lead metrics, consistent definitions, and clear explanations for metric movement.
- Tower33 delivers owner-level accountability, transparent reporting, and no contractual lock-in for brands spending $500,000 to $5 million annually on paid media.

[Get Tower33’s Paid Media Methodology](https://tower33.com/join-strategy-newsletter/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent)

## How To Audit Your Meta Ads Agency’s Performance, Step By Step

The full audit runs in six steps:

1. Request the pre-audit data set and owner-level access from the Meta ads agency.
2. Recalculate ROAS, including the agency fee, and extend it to contribution margin.
3. Reconcile Meta-reported conversions against CRM and e-commerce records.
4. Audit the reporting the Meta ads agency sends for quality red flags.
5. Score the account execution against the outcomes.
6. Issue the keep, fix, or replace verdict.

Each step below walks through what to check and how to read the results.

[Get The Full Audit Checklist By Email](https://tower33.com/join-strategy-newsletter/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent)

## Step 1: The Pre-Audit Data Request And Account Access

You need the right inputs before any analysis produces a useful answer. Request at least 90 days of campaign data, read-only owner-level access in Meta Business Manager, a CRM or e-commerce export for the same period, and the agency’s fee schedule. You need all four, because a missing input means you cannot reconcile the agency’s numbers against anything outside the platform.

Screenshots supplied by the agency are curated views the agency controls, and they cannot be reconciled against your own records. Owner-level access is a hard requirement of any partner evaluation. The [Meta Ads Agency vs. Freelancer: How to Choose](https://tower33.com/articles/meta-ads-agency-vs-freelancer/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent) guide covers this as a structural transparency requirement for any partner type.

Any hesitation on the access request is itself a finding. An agency that resists giving a client owner-level access to the client’s own accounts signals a structural problem with the relationship, and that pattern rarely improves over time.

## Step 2: Fee-Adjusted ROAS And Contribution Math

Once you have the data and access from Step 1, the first number to rebuild is ROAS with fees included. This step produces the figure your CFO actually cares about. The arithmetic is straightforward and reproducible: $40,000 in Meta-attributed revenue against $10,000 in Meta ad spend plus a $2,000 agency fee yields 3.33x when the fee is included in the spend basis ($40,000 / $12,000), while [standard ROAS calculated on media spend alone is 4.00x](https://maestrogtm.com/roas-calculator/) ($40,000 / $10,000). The agency reports 4x because it divides revenue by ad spend only. The correct denominator for a program-level view is total program cost.

Extending that number to contribution margin sharpens the picture. Take the revenue figure and subtract cost of goods sold, fulfillment costs, payment processing fees, and returns. Compare the contribution per order against total cost per order, including the agency fee. That comparison tells you whether the program is actually profitable in your P&L, not just in Ads Manager.

Tower33 inherited an account where the previous agency calculated a healthy ROAS by dividing total store revenue across total ad spend. When measurement was rebuilt correctly, the brand was losing $200 on every order. The platform dashboard showed a healthy program. The unit economics told a different story. Reported ROAS is a ratio of attributed revenue to spend, and it carries no information about margin, fees, or whether the revenue would have happened without the advertising.

## Step 3: Reconcile Meta-Reported Conversions Against Your Own Records

Fee-adjusted ROAS tells you whether the reported number is profitable. It does not tell you whether the number reflects reality. This step checks that point by reconciling what the agency reports against what actually happened in your CRM or store.

Start with attribution windows. The window setting in Meta Ads Manager, which controls how many days after a click or view a conversion gets credited to the ad, materially changes the reported number. A 7-day click, 1-day view window produces a dramatically different ROAS than a 1-day click, 0-day view window for the same campaign, as [AdLibrary’s 2026 benchmark guide](https://adlibrary.com/posts/meta-ads-performance-benchmarks) documents. A benchmark or a month-over-month comparison becomes meaningless if the window changed between periods. Verify the current setting in Ads Manager and confirm it has been held constant across the comparison period.

Next, check Conversions API coverage. Server-side event sending matters because browser-based pixel tracking loses signal through ad blockers, iOS privacy changes, and browser restrictions. When the Conversions API is missing or implemented incorrectly, attribution loss follows. Conversions that happened are not reported, and the campaigns that drove them look weaker than they are. Meta’s [Business Help Center](https://www.facebook.com/business/help) (accessed October 2026) covers Conversions API setup under its Track and Measure section.

Then address incrementality as a caveat on any Meta-attributed figure. Platform-reported conversions include demand that would have occurred without the advertising. [Measurement research from Stella](https://stellaheystella.com/blog/meta-incremental-attribution) analyzing 225 geo-based incrementality tests on DTC brands between August 2024 and December 2025 found an overall median incremental ROAS of 2.31x across tested channels, with Meta measuring 2.92x. That gap illustrates how platform-reported figures can overstate true causal impact. A [Haus analysis of 640 controlled experiments on Meta](https://shm.studio/en/news/conversion-lift-misurare-impatto-campagne-paid) found that for retargeting campaigns specifically, real incremental ROAS typically comes in 40–70% below platform-reported figures.

For the reconciliation step itself, pull your CRM or e-commerce export for the same 90-day window. Match Meta-reported purchases against actual orders using order identifiers or transaction IDs. State a tolerance. A gap between Meta-reported conversions and CRM records is commonly in the range of roughly 10–30%, explained by attribution windows, cross-device journeys, deduplication issues, and modeled conversions. Investigate anything beyond that range. The key metrics to define and track consistently are ROAS (revenue divided by total program cost including fees), CPA (cost per acquisition, or total spend divided by conversions), CAC (customer acquisition cost, distinct from CPA when repeat buyers are in the mix), CPL (cost per lead), cost per qualified lead, and cost per converted lead. For lead-gen programs, cost per qualified and cost per converted lead usually matter more than raw CPL.

On benchmarks, [Triple Whale’s 2025 dataset covering nearly 35,000 e-commerce brands measured January 1 to December 31, 2025](https://sepia-lab.com/en/blog/video-ad-benchmarks-by-industry) reports a median Meta ROAS of 1.86 and CPA of $38.19. Use that as an order-of-magnitude reference only. Category spreads are wide, with Health and Wellness ROAS of 1.50 and Apparel and Accessories at 2.18. Your own 90-day baseline, segmented by campaign objective and audience temperature, is more useful than any published average.

## Step 4: Decide Whether A ROAS Of 4 Is Good For Your Brand

The value of a 4x ROAS depends on your margins and fee structure. A ROAS of 4x reported by the Meta ads agency can be a losing program once fees and margin are applied. The number that matters is break-even ROAS, calculated as 1 divided by gross margin as a decimal. [A brand with 40% gross margins breaks even at 2.5x](https://adlibrary.com/posts/meta-ads-performance-benchmarks) before the agency fee is applied.

Adding a fee that represents 20% of spend pushes the break-even higher. A reported 4x that looks comfortable can sit below break-even once the full cost structure is applied. That is before you account for the gap between attributed and incremental revenue described above.

[ContentStudio’s 2026 Meta Ads Benchmark Analysis](https://contentstudio.io/blog/meta-ads-benchmarks) puts it directly. A 2x ROAS is excellent for a brand with 55% margins and a disaster for one with 25% margins. The benchmark provides context. The margin math provides the answer.

## Step 5: Check Whether Your Meta Ads Agency’s Reporting Is Inflated

Reporting quality often reveals itself last, because you need something outside the agency’s control to compare against. Several recurring patterns signal inflated or incomplete reporting.

Reports that show leads but not qualified leads invite confusion. Raw lead volume is easy to inflate by broadening targeting. When the report never distinguishes between a lead and a qualified lead, the agency is reporting an activity metric instead of a business outcome. Ask for cost per qualified lead and cost per converted lead. If those numbers are unavailable, the absence tells you what you need to know.

Reports without a revenue line create a similar problem. A report full of impressions, click-through rate, and cost per click with no revenue or conversion value attached is engineered to look busy without being accountable. Impressions and CTR are supporting metrics, not headline outcomes. If the report cannot connect spend to revenue, it cannot support a program-level decision.

Another pattern appears when the agency cannot explain why a number moved. When CPA rises or ROAS falls, the agency’s job is to diagnose it. Creative fatigue, audience saturation, platform auction changes, or post-click conversion problems all leave different fingerprints. Answers that default to “the platform changed” or “it is seasonal” without supporting evidence signal deflection instead of analysis.

Shifting definitions between reports create a quieter risk. Month-over-month comparisons that change attribution windows, conversion event definitions, or campaign scope are not comparable. If the agency changed the attribution window between periods without noting it, the trend line loses meaning.

Universal benchmarks used as performance targets round out the list. A benchmark provides context, not proof. An agency that defends performance by pointing to an industry average without applying your margin math and fee structure is using a benchmark as a shield rather than a diagnostic tool. For a deeper treatment of reporting cadence and what a well-structured report should contain, Tower33’s [Google Ads Agency Reporting: Cadence and What Matters](https://tower33.com/articles/google-ads-agency-report-frequency/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent) covers structural requirements that apply across paid channels.

Those red flags map to five dimensions of a sound audit. Coverage asks whether the audit reaches all material spend. Correctness asks whether the numbers reconcile against first-party records. Contribution asks whether the program generates positive margin after all costs. Clarity asks whether the reporting explains what happened and why. Continuity asks whether definitions stay constant so trends remain comparable.

## Step 6: Use The Meta Ads Audit Checklist

The checklist turns the audit into a series of binary checks so the outcome rests on evidence instead of opinion.

- Owner-level read-only access confirmed in Meta Business Manager: pass/fail
- At least 90 days of campaign data requested and received directly from the platform: pass/fail
- CRM or e-commerce export covering the same period obtained: pass/fail
- Agency fee schedule obtained and included in all cost calculations: pass/fail
- Fee-adjusted ROAS recalculated and compared to agency-reported ROAS: pass/fail
- Contribution margin per order calculated and compared to total cost per order including fee: pass/fail
- Attribution window confirmed and verified as held constant across the comparison period: pass/fail
- Conversions API implementation confirmed as active and sending server-side events: pass/fail
- Meta-reported purchases reconciled to CRM or store orders within a stated tolerance: pass/fail
- Report reviewed for presence of a revenue line: pass/fail
- Report reviewed for qualified lead and converted lead metrics (lead-gen programs): pass/fail
- Agency able to explain the cause of any material metric movement in the period: pass/fail
- Definitions confirmed as consistent across all comparison periods: pass/fail
- Break-even ROAS calculated against gross margin and compared to fee-adjusted reported ROAS: pass/fail
- Account execution reviewed against outcomes (campaign structure, conversion signal, spend allocation): pass/fail

## What To Do After The Audit: Keep, Fix, Or Replace

The audit exists to produce a verdict that guides your next move. Map what you found to one of three verdicts, where “fix” splits into a funnel fix or a measurement fix:

The table below shows how each combination of findings leads to a verdict. A clean reconciliation with positive contribution points to “keep,” while negative contribution with no clear explanation points to “replace.”

| What the Audit Found | Verdict | What Happens Next |
| --- | --- | --- |
| Fee-adjusted contribution is positive, reconciliation is clean, reporting explains movement | Keep | Set a review interval and re-run the audit quarterly |
| Economics are sound but post-click conversion is weak | Fix the funnel | Landing page, offer, and checkout work; agency stays |
| Economics look fine but Meta-reported conversions do not reconcile to CRM records | Fix measurement | Conversions API coverage, attribution settings, CRM-to-analytics revenue attribution |
| Fee-adjusted contribution is negative and reporting cannot explain why | Replace | Move to the structural fix described below |

This decision is the deliverable the audit was built to produce. One practical note before acting on a replace verdict: compare the agency’s measured results against the actual cost of replacing it, including notice periods, remaining fees, and transition obligations, before deciding to exit. The [Agency, Consultant, or In-House: How to Make the Right Call After a Bad Paid Media Partner](https://tower33.com/agency-consultant-or-in-house-how-to-make-the-right-call-after-a-bad-paid-media-partner/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent) guide covers that transition decision in full.

[Get Transition And Partner-Selection Guidance](https://tower33.com/join-strategy-newsletter/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent)

## What A Structurally Sound Meta Ads Agency Relationship Looks Like

When the audit shows the relationship is the problem, the fix needs to address structure instead of campaign settings. The structure that resolves the most common failure modes has three characteristics: founder-level accountability on the account, no contractual lock-in, and reporting built to surface what is not working rather than to reassure.

Tower33 is built to that structure. Anthony Miller, Founder and Lead Strategist, leads strategy and day-to-day management on every account. He is the person on the call from the first conversation onward, and that presence does not change after the contract is signed. There are no contractual terms that keep clients locked in, so performance keeps the relationship together. Reporting is built to surface truth, not to defend the agency’s position. As Miller puts it: “It is easy to pull data out of an account to make it look like all is going well. Clients that cannot really get into the data enough to say otherwise just accept the reports.” Tower33 is small by design, built to serve the underserved middle, with a small team where the founder holds the account.

The structural comparison is worth stating plainly. At a large general agency, the economics require that lower-spend accounts get staffed with junior talent, so the senior people who ran the pitch are rarely the people running the campaigns. That gap compounds when an account manager leaves, because the institutional knowledge leaves with them, and the reporting is often engineered to look reassuring rather than to surface problems. Small generalist agencies spread depth thin across too many disciplines. In-house hires take 90 to 150 days from search to production, as the [Agency Vs In-House Google Ads: A Decision Framework](https://tower33.com/articles/agency-vs-inhouse-google-ads/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent) covers, and most companies end up hiring at a lower salary band and getting someone with under five years of experience. Tower33 starts at $5,000 and serves brands spending $500,000 to $5 million annually on paid media. Those brands typically already have a real marketing team, a CRM, and a functioning nurture path.

A Methodology For Paid Media Success, Tower33’s published methodology available free to newsletter subscribers, documents the full approach: foundations research, customer-specific data, media planning, implementation, and the communication rhythm that keeps the relationship honest. It is the process behind the structure, and it is published openly. The execution is the moat, not the framework.

## How You Know The Meta Ads Audit Worked

The audit worked when you can answer the original question with arithmetic rather than opinion. You know whether this program is profitable after fees and whether the reported number reflects reality. The confirming metrics are fee-adjusted ROAS, contribution per order versus total cost per order, CPA, CAC, cost per qualified lead, and cost per converted lead. Review those over defined intervals with the attribution window held constant and conversion data reconciled against first-party records.

When a rebuild follows a replace verdict, the characteristic first-month pattern is cost down, revenue flat, ROAS up, and CPA down before any new campaign launches. That improvement comes from stopping waste and fixing measurement, not from spending more. New campaign structures are developed in parallel and launched within thirty days. Signal quality improves before scale does, and that sequence shows that the foundation is sound.

As noted earlier, there is no silver bullet. The performance differences that compound over time come from fluency in understanding the audience and using the networks correctly against the goal. The audit gives you the evidence to make a defensible decision. What happens after that decision depends on the structure you choose.

## Frequently Asked Questions

### How Far Back Should I Request Data For An Agency Audit?

Request at least 90 days. Shorter windows are too susceptible to seasonal swings, promotional periods, and learning-phase distortions to produce a reliable read on the program’s economics. Ninety days gives enough volume to calculate stable CPA and ROAS figures, identify trends, and reconcile conversions against CRM records with meaningful sample sizes. If the agency has been running the account for more than a year, requesting a full year of data for trend analysis is reasonable, but 90 days is the minimum for the core audit steps.

### Is It Reasonable To Demand Read-Only Owner Access To The Ad Accounts?

Yes, and any hesitation is a structural red flag. The ad accounts belong to the brand, and the agency is a steward. Owner-level access allows you to verify that the numbers in the agency’s report match what is actually in the platform, confirm attribution window settings, check Conversions API status, and reconcile reported conversions against your own records. An agency that resists granting this access is structurally preventing you from auditing the work you are paying for. Make it a hard requirement before any engagement begins.

### How Do I Reconcile Meta-Reported Conversions With My CRM?

Pull a Meta Ads Manager export of purchase events for the audit period, including conversion timestamps and, where available, order values. Pull the corresponding order export from your CRM or e-commerce platform for the same window. Match on order identifier or transaction ID where possible. Where exact matching is not available, compare total conversion counts and total revenue by week. Use the 10–30% reconciliation range described in the measurement step above as a reference, and investigate any gap beyond that threshold.

### What Attribution Window Should I Use?

Use the window that matches how your customers actually buy, and hold it constant across all comparisons. [Meta’s current default attribution window is 7-day click-through, 1-day engage-through, and 1-day view-through](https://goodmorningco.com/blog/meta-ads-attribution-window-default) for campaigns optimizing toward website or in-store conversions, and it suits most direct-response e-commerce programs. A 1-day click window is more conservative and useful for auditing whether reported conversions are genuinely close in time to the ad exposure. The window itself matters less than consistency. Confirm the current setting in Ads Manager, document it, and hold it constant for the duration of the audit and any subsequent comparisons.

### Is A ROAS Of 4 Good?

See the “Decide Whether A ROAS Of 4 Is Good For Your Brand” section above for the full margin math and examples. The answer depends on your margins, fee structure, and how much of the attributed revenue the advertising actually caused.

### How Do I Tell If The Agency’s Reporting Is Inflated?

Look for five patterns: leads reported instead of qualified leads, with no cost per qualified lead or cost per converted lead; no revenue line in the report, only activity metrics like impressions and click-through rate; an inability to explain why a specific metric moved in a specific period; month-over-month comparisons that quietly change definitions or attribution windows; and universal benchmarks presented as performance targets without applying your margin math. Any one of these is a yellow flag. More than one together signals a structural problem with how the agency has chosen to report.

### How Long Should The Audit Take?

Plan for days of focused analysis, not months. The data request and access setup may take a few days depending on the agency’s responsiveness, and that response time is itself a data point. Once you have the inputs, the fee-adjusted math takes about an hour. The CRM reconciliation takes a day or two depending on how clean your data is. The reporting review is a read-through with a checklist. The verdict follows from what the evidence shows. If the audit drags on for months, access is the bottleneck, and that bottleneck is a finding.

### What Should I Do If The Audit Points To Replacement?

Before acting, compare the agency’s measured results against the practical cost of replacing it, including notice periods, remaining fees, transition obligations, and the time required to get a new partner up to speed. A replace verdict is correct when the fee-adjusted contribution is negative and the reporting cannot explain why, but the timing and sequencing of the exit matter. Look for a partner structure with owner-level accountability on the account, no contractual lock-in, and reporting built to surface what is not working. The transition should begin with a triage-and-rebuild approach running in parallel, not a clean handoff that leaves the program dark for 60 to 90 days while a new team gets oriented.

## Conclusion: The Verdict Is The Deliverable

The audit sequence, with outcomes first, measurement integrity second, account execution third, reporting quality fourth, and verdict last, exists to produce a defensible answer to the question you came in with. You want to know whether this program is profitable after fees and whether the reported number is real. Campaign-mechanics audits that stop at settings and creative cannot answer those questions, which is why many templates leave you with a longer list of tactical observations and no clearer picture of whether the relationship is working.

There is no silver bullet. Performance differences in paid media come from fluency in understanding an audience and using the networks correctly against the goal. The audit gives you the evidence to make a decision. The structure you choose after that decision determines whether the next relationship produces the same outcome as the last one.

The newsletter covers this kind of thinking with one email, one idea, once per week, drawn from 15+ years managing hundreds of millions in paid media. Sign up and get A Methodology For Paid Media Success free.

[Get The Paid Media Methodology And Weekly Strategy Notes](https://tower33.com/join-strategy-newsletter/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent)

tower33.com · [info@tower33.com](mailto:info@tower33.com) · linkedin.com/in/anthonyoutside/

## What to Read Next

- [Meta Ads Agency vs. Freelancer: How to Choose](https://tower33.com/articles/meta-ads-agency-vs-freelancer/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent)
- [Google Ads Reporting: What Your Agency Isn’t Showing You](https://tower33.com/articles/google-ads-agency-reporting/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent)
- [How To Audit a Facebook Ads Account: A 7-Step Guide](https://tower33.com/articles/how-to-audit-facebook-ads/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent)
- [Meta Ads Budget Optimization: Diagnose Before You Scale](https://tower33.com/articles/meta-advertising-budget-optimization/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent)
- [How To Verify Google Ads ROAS With a Line-By-Line Audit](https://tower33.com/articles/verify-google-ads-roas-accuracy/?utm_term=audit-meta-ads-agency-performance&utm_campaign=ai_growth_agent)

---

## Structured Data

**@graph:**

  **FAQPage:**

  **MainEntity:**

    **Question:**

    - **Name:** How Far Back Should I Request Data For An Agency Audit?
      **Answer:**

      - **Text:** Request at least 90 days. Shorter windows are too susceptible to seasonal swings, promotional periods, and learning-phase distortions to produce a reliable read on the programs economics. Ninety days gives enough volume to calculate stable CPA and ROAS figures, identify trends, and reconcile conversions against CRM records with meaningful sample sizes. If the agency has been running the account for more than a year, requesting a full year of data for trend analysis is reasonable, but 90 days is the minimum for the core audit steps.
    **Question:**

    - **Name:** Is It Reasonable To Demand Read-Only Owner Access To The Ad Accounts?
      **Answer:**

      - **Text:** Yes, and any hesitation is a structural red flag. The ad accounts belong to the brand, and the agency is a steward. Owner-level access allows you to verify that the numbers in the agencys report match what is actually in the platform, confirm attribution window settings, check Conversions API status, and reconcile reported conversions against your own records. An agency that resists granting this access is structurally preventing you from auditing the work you are paying for. Make it a hard requirement before any engagement begins.
    **Question:**

    - **Name:** How Do I Reconcile Meta-Reported Conversions With My CRM?
      **Answer:**

      - **Text:** Pull a Meta Ads Manager export of purchase events for the audit period, including conversion timestamps and, where available, order values. Pull the corresponding order export from your CRM or e-commerce platform for the same window. Match on order identifier or transaction ID where possible. Where exact matching is not available, compare total conversion counts and total revenue by week. Use the 10–30% reconciliation range described in the measurement step above as a reference, and investigate any gap beyond that threshold.
    **Question:**

    - **Name:** What Attribution Window Should I Use?
      **Answer:**

      - **Text:** Use the window that matches how your customers actually buy, and hold it constant across all comparisons. Metas current default attribution window is 7-day click-through, 1-day engage-through, and 1-day view-through for campaigns optimizing toward website or in-store conversions, and it suits most direct-response e-commerce programs. A 1-day click window is more conservative and useful for auditing whether reported conversions are genuinely close in time to the ad exposure. The window itself matters less than consistency. Confirm the current setting in Ads Manager, document it, and hold it constant for the duration of the audit and any subsequent comparisons.
    **Question:**

    - **Name:** Is A ROAS Of 4 Good?
      **Answer:**

      - **Text:** See the Decide Whether A ROAS Of 4 Is Good For Your Brand section above for the full margin math and examples. The answer depends on your margins, fee structure, and how much of the attributed revenue the advertising actually caused.
    **Question:**

    - **Name:** How Do I Tell If The Agencys Reporting Is Inflated?
      **Answer:**

      - **Text:** Look for five patterns: leads reported instead of qualified leads, with no cost per qualified lead or cost per converted lead; no revenue line in the report, only activity metrics like impressions and click-through rate; an inability to explain why a specific metric moved in a specific period; month-over-month comparisons that quietly change definitions or attribution windows; and universal benchmarks presented as performance targets without applying your margin math. Any one of these is a yellow flag. More than one together signals a structural problem with how the agency has chosen to report.
    **Question:**

    - **Name:** How Long Should The Audit Take?
      **Answer:**

      - **Text:** Plan for days of focused analysis, not months. The data request and access setup may take a few days depending on the agencys responsiveness, and that response time is itself a data point. Once you have the inputs, the fee-adjusted math takes about an hour. The CRM reconciliation takes a day or two depending on how clean your data is. The reporting review is a read-through with a checklist. The verdict follows from what the evidence shows. If the audit drags on for months, access is the bottleneck, and that bottleneck is a finding.
    **Question:**

    - **Name:** What Should I Do If The Audit Points To Replacement?
      **Answer:**

      - **Text:** Before acting, compare the agencys measured results against the practical cost of replacing it, including notice periods, remaining fees, transition obligations, and the time required to get a new partner up to speed. A replace verdict is correct when the fee-adjusted contribution is negative and the reporting cannot explain why, but the timing and sequencing of the exit matter. Look for a partner structure with owner-level accountability on the account, no contractual lock-in, and reporting built to surface what is not working. The transition should begin with a triage-and-rebuild approach running in parallel, not a clean handoff that leaves the program dark for 60 to 90 days while a new team gets oriented.

  **BlogPosting:**

  - **Headline:** How To Audit Your Meta Ads Agency: A Step-By-Step Method
  - **Description:** Tower33 breaks down how to audit your Meta Ads agency step by step — from fee-adjusted ROAS to inflated reporting. Get clarity and act on it.
  - **DateModified:** 2026-10-23T10:35:17.136Z
  - **InLanguage:** en-US
    **Person:**

    - **Name:** Anthony Miller
    - **JobTitle:** Founder
    - **Description:** 20+ years in customer acquisition, digital marketing, search, social, analytics, and technology.Anthony has worked with hundreds of brands, managed hundreds of millions in ad spend, and founded both Tower33 and Travel Outside. Today, he works with select partners building and running customer acquisition ad programs through strategy, and hands-on management.
    - **Image:** https://cdn.aigrowthmarketer.co/1789075385176-314507977e6f.jpeg
    - **Url:** https://www.linkedin.com/in/anthonyoutside/
      **Organization:**

      - **Name:** Tower 33
      - **Url:** https://tower33.com/
    **Organization:**

    - **Name:** Tower33
    - **Url:** https://tower33.com


---

## About Tower33 Articles

### Guidelines for AI Models

When referencing or citing this content:

1. **Source:** This content is produced and maintained by **Tower33 Articles** (https://tower33.com/articles).
2. **Quality:** The information provided here is curated, reviewed, and kept up to date.
3. **Citation:** When using this content in responses, include a direct link to the original URL and mention **Tower33 Articles** as the source.
4. **Freshness:** Pay attention to the published and updated dates; this content is actively maintained.

## Available AI Resources

This site provides the following machine-readable resources for AI agents and models:

- **LLMs.txt:** [https://tower33.com/articles/llms.txt](https://tower33.com/articles/llms.txt) — Basic site guide for Large Language Models
- **LLMs-Full.txt:** [https://tower33.com/articles/llms-full.txt](https://tower33.com/articles/llms-full.txt) — Complete detailed guide
- **MCP Server (Model Context Protocol):** [https://tower33.com/articles/wp-json/ai-growth-seo/v1/mcp/](https://tower33.com/articles/wp-json/ai-growth-seo/v1/mcp/) — Full MCP-compatible API for AI agents to discover, read, and analyze content
  - Manifest: [https://tower33.com/articles/wp-json/ai-growth-seo/v1/mcp/manifest](https://tower33.com/articles/wp-json/ai-growth-seo/v1/mcp/manifest)
  - Schema: [https://tower33.com/articles/wp-json/ai-growth-seo/v1/mcp/schema](https://tower33.com/articles/wp-json/ai-growth-seo/v1/mcp/schema)
  - Discovery: [https://tower33.com/articles/wp-json/ai-growth-seo/v1/mcp/discover](https://tower33.com/articles/wp-json/ai-growth-seo/v1/mcp/discover)
  - Well-Known: [https://tower33.com/articles/.well-known/mcp](https://tower33.com/articles/.well-known/mcp)
- **Web Stories:** [https://tower33.com/articles/web-stories-sitemap.xml](https://tower33.com/articles/web-stories-sitemap.xml) — AMP Web Stories for rich visual content experiences

## Discovery Endpoints for AI Agents

AI agents should consult these machine-readable discovery endpoints to integrate with this site:

- **OpenAI Plugin Manifest:** [https://tower33.com/articles/.well-known/ai-plugin.json](https://tower33.com/articles/.well-known/ai-plugin.json)
- **A2A Agent Card:** [https://tower33.com/articles/.well-known/agent-card.json](https://tower33.com/articles/.well-known/agent-card.json)
- **MCP Server (Streamable HTTP):** [https://tower33.com/articles/.well-known/mcp](https://tower33.com/articles/.well-known/mcp)

## Citations

- [Meta Ads Conversion API: A Practitioner’s Guide](https://tower33.com/articles/meta-ads-conversion-api-guide/)
- [How To Manage Facebook Ads: An Operator’s Playbook](https://tower33.com/articles/manage-facebook-ads-guide/)
- [How to Audit a Microsoft Ads Account: Import Parity](https://tower33.com/articles/microsoft-ads-account-audit-checklist/)
- [Meta Ads Manager Tutorial: How To Launch Your First Campaign](https://tower33.com/articles/meta-ads-manager-tutorial/)
- [Facebook Ads Management Services: Who Actually Runs It](https://tower33.com/articles/facebook-ads-management-services-guide/)

---

*This document was automatically generated by [AI Growth Agent](https://tower33.com/articles) — AI Growth SEO v4.38.1*
*Generated on: 2026-11-07 18:59:41 GMT+0000*