Agency Accountability

Marketing Agency Audit: What to Check Before You Switch Agencies

Before switching marketing agencies, turn each suspected problem into an evidence record. Use the TAC Agency Audit Ledger to separate facts, responsibility, response and transition risk.

By The Agency Check Editorial TeamPublished October 5, 2026

You think something is wrong with your marketing agency. Maybe results have declined. Maybe the reports look better than the business feels. Maybe you keep hearing that campaigns are being optimized, but you cannot tell what has actually changed.

At that point, it is tempting to jump straight to: Should we switch agencies? That is usually one question too early.

The Agency Check
Before you decide whether to switch agencies, establish what the suspected problem is, what evidence supports it, who controls it, how the agency responded and whether it keeps happening.

The short answer

Before switching marketing agencies, audit each serious concern as a separate evidence record. For every suspected problem, document eight things:

The TAC Agency Audit Ledger
Ledger field	What you need to establish
Concern	What specifically do you believe is wrong?
Business Impact	Why does it matter to the business?
Evidence	What information actually supports the concern?
Control	Who could reasonably influence or control the issue?
Response	What did the agency do once the issue became known?
Recurrence	Is this isolated or a repeating pattern?
Fixability	Can the underlying issue realistically be corrected?
Transition Risk	What could be disrupted or lost if the relationship changes?

We call this the TAC Agency Audit Ledger. It is not a scorecard. Its job is to stop one vague concern from becoming a major business decision without enough evidence.

TAC Agency Audit Ledger
One suspected problem becomes one evidence record.

A marketing agency audit should investigate the concern, not start with the verdict

There is a difference between evaluating your agency generally and auditing a specific suspected problem. If your question is simply whether the agency is doing a good job, start with the broader performance standard. If you are considering a change because something specific appears wrong, use the Ledger to investigate that problem.

Suppose revenue has declined. That does not tell you whether the cause is weaker campaign execution, inaccurate tracking, lower website conversion, pricing, inventory, sales follow-up, changing demand or several factors together. Discovering a real problem also does not automatically prove the agency caused it.

What happenedWhat proves itWho controls itHow the agency responded

The TAC Agency Audit Ledger

1. Concern: What exactly do you think is wrong?

Make the concern specific enough that another person could investigate it. Write the observable problem, not your conclusion about whose fault it is.

Example
Weak: Our agency is not optimizing enough. Better: Paid-search cost per qualified opportunity has increased for three consecutive months, but budget allocation and targeting appear largely unchanged.

Replace statements such as “the agency wasted our money” with a testable concern such as “we spent more while the business outcome we care about deteriorated.”

2. Business Impact: Why does this matter?

Connect the concern to a meaningful business outcome such as revenue, qualified pipeline, customers acquired, profitability, customer acquisition cost, lead quality, conversion rate, wasted spend, lost organic visibility or operational time. The purpose is to distinguish something unusual from something material.

3. Evidence: What actually supports the concern?

A concern is not evidence simply because everyone in the meeting agrees with it. Use information as close as possible to the outcome you are investigating.

CRM or ecommerce records
Finance or qualified-lead data
Analytics and advertising-platform data
Search-performance data
Campaign change histories
Landing-page performance
Agency reports
Project-management records, emails or meeting notes

For example, Google Ads Change History can show what changed, when it changed and which user made the change. That can establish activity, but it does not prove the change was intelligent. Separate what the evidence shows from what you believe explains it.

Official documentation:Google Ads — Review your account history

If the evidence is insufficient, record that. “We cannot determine whether this problem is real because the measurement is unreliable” is itself a useful finding.

4. Control: Who could actually influence the problem?

Classify control as agency-controlled, client-controlled, shared, external or uncertain. This prevents everything that happens to marketing performance from being treated as agency performance.

An advertising agency may control targeting and campaign structure while the business controls pricing, inventory, sales response times, internal approvals and whether recommended website changes are implemented. That does not excuse poor agency work; it makes responsibility more accurate.

5. Response: What did the agency do when the problem emerged?

Did the agency notice it?
Did they raise it before you did?
What explanation did they offer?
What evidence supported that explanation?
What changed as a result?
Is there a clear next step?

The response is separate evidence. “Lead quality is down because the market is slow” is difficult to evaluate. “Sales-accepted leads declined most sharply from these campaigns, so we are changing how qualified outcomes feed back into campaign decisions” creates something that can be tested later.

6. Recurrence: One event or a pattern?

Record whether the problem happened once, occasionally, repeatedly or continuously. Use specific history where possible: “This occurred in March and May” is more useful than “This always happens.”

7. Fixability: Can the underlying issue realistically be corrected?

Document what would have to change for this specific problem to stop happening. That might mean repairing tracking, improving the CRM feedback loop, reallocating budget, clarifying responsibilities, adding specialist support or changing the underlying strategy. This field evaluates the problem's fixability; it does not decide the future of the agency relationship.

8. Transition Risk: What would change if the relationship changed?

Record what could be disrupted or lost during a transition: account access, campaign continuity, reporting history, analytics configuration, creative source files, testing history, unfinished work, tracking infrastructure or institutional knowledge.

Transition risk should never be used to excuse poor work. It simply tells the business what would need protection if responsibility changes.

A fully worked TAC Agency Audit Ledger example

The following numbers are illustrative, not TAC client data. Imagine a B2B company says: “Our agency says lead generation is improving, but sales says the leads are getting worse.” The Ledger turns that concern into a record that can be evaluated.

Illustrative TAC Agency Audit Ledger
Ledger field	Example audit record
Concern	Paid-media lead volume has increased, but the business suspects optimization is rewarding inexpensive form submissions rather than qualified prospects.
Business Impact	Over three months, paid-media leads increased 31%, while sales-accepted leads declined 22% and booked meetings declined 18%.
Evidence	CRM lead-status data shows the decline in accepted leads. Advertising reports confirm increased form conversions. Source data shows most additional leads came from two campaigns, while sales notes show more irrelevant or poor-fit enquiries.
Control	Shared. The agency controls campaign targeting, budgets, bidding and much of the optimization. The company controls qualification definitions, CRM discipline, lead follow-up and business-side constraints.
Response	The agency initially highlighted the lower cost per lead. After CRM qualification data was shared, it acknowledged that form volume and qualified demand were diverging and proposed changing campaign allocation and feeding deeper-funnel outcomes into optimization.
Recurrence	The lead-quality gap has appeared for three consecutive months rather than in one isolated reporting period.
Fixability	Potentially high. The issue may be correctable if both sides agree on the qualified outcome, improve the CRM feedback loop and reallocate spend based on business quality rather than raw form volume. The corrective approach has not yet been proven.
Transition Risk	Moderate. The business has access to the advertising and analytics accounts, but campaign rationale, testing history and current optimization context would still need to be documented if responsibility changed.

The original claim — “the agency is generating bad leads” — is now more precise. The decline in lead quality appears real and commercially meaningful. Responsibility is shared. The agency had been optimizing around a metric that no longer represented the business outcome well, but it has proposed a corrective response. The issue has recurred, appears potentially fixable, and transition risk is manageable but not zero.

TAC Agency Audit Ledger
That is an audit record. It is not yet a verdict on the agency relationship.

If you have more than one concern, create more than one record

Do not combine unrelated issues into one giant scorecard. Paid lead quality, unimplemented SEO recommendations and reporting that does not reconcile with CRM revenue may involve different evidence, owners and solutions.

Separate records also improve the agency conversation. Instead of saying “we do not think you are doing enough,” you can bring three specific concerns, the evidence behind each one, what remains uncertain and what needs clarification.

Use these questions to test unclear findings:12 Questions to Ask Your Marketing Agency on Your Next Monthly Call

Give the agency a chance to challenge the record

An agency audit should not be built only from the client's side of the story. Where appropriate, ask the agency to identify missing evidence, factual errors, work that was completed but not visible, constraints you attributed incorrectly and assumptions they disagree with.

This is not about letting the agency rewrite the audit. It is about testing the record. If credible evidence changes a finding, update it. The objective is accuracy.

Use monthly reporting as evidence, not as the verdict

A useful monthly agency report should make an audit easier. You should be able to identify the important business outcome, what changed, relevant spend, what the agency believes caused the change, what decisions were made and what happens next.

If months of reporting cannot help you reconstruct those basics, that is useful context — but the audit should still document the specific concern and evidence rather than treating a weak report as proof of every other problem.

What a completed Agency Audit Ledger should give you

At the end, you should have evidence records — not a list of complaints. A useful record separates the concern from the conclusion, business impact from annoyance, evidence from assumption, and responsibility from reaction.

It should also show whether the issue repeats, whether a credible correction exists and what would need to be protected if responsibility changes. That is enough for the audit.

The audit is complete. The agency decision comes next.

Once your concerns have been converted into evidence records, stop auditing. You now have the inputs for a different question: what should we do about the agency relationship?

That decision is intentionally outside the scope of this article. Do not force the Ledger to answer it.

Next step:When Should You Fire Your Marketing Agency?

The Agency Check
The Agency Audit Ledger has one job: make sure the next decision starts with evidence.

Frequently Asked Questions

What is a marketing agency audit?: A marketing agency audit is a structured review of evidence about how the agency's work is performing and being managed. For a specific concern, the TAC Agency Audit Ledger records the concern, business impact, evidence, control, agency response, recurrence, fixability and transition risk.
Should I audit my marketing before switching agencies?: If the reason for switching is uncertain, an evidence audit can help establish what is actually wrong before you make the relationship decision. If you have several concerns, create a separate record for each.
What if I cannot prove my concern?: Record the evidence gap instead of filling it with an assumption. If the business cannot reliably determine whether leads, revenue or conversions are being measured correctly, that uncertainty itself may be an important finding.
Does the Agency Audit Ledger tell me whether to fire my agency?: No. It intentionally stops before that decision. The Ledger creates the evidence record. Use TAC's separate guide on when to fire your marketing agency to evaluate what the completed evidence means for the relationship.

Ready for clarity?

Want clarity before you make a major agency decision?

The Agency Check reviews what is working, what is not, and what deserves action next so you can make the decision with better evidence.

Check my marketing

Keep reading

Agency Accountability
How Do I Know If My Marketing Agency Is Doing a Good Job?
A practical framework for evaluating your marketing agency without relying on gut feeling, vanity metrics or one good or bad month.
Read guide
Agency Accountability
When Should You Fire Your Marketing Agency?
Learn how to tell the difference between a difficult marketing period and an agency relationship that may actually need to end.
Read guide
Agency Accountability
What Should Your Marketing Agency Be Doing Between Monthly Calls?
A practical way to tell whether your agency is actively managing your marketing between monthly meetings, without expecting constant changes for the sake of activity.
Read guide

All guides