I Fired 3 Marketing Agencies in 18 Months. Here's the Pattern.
Three agencies. Eighteen months. Same root cause each time. Here is the pattern that shows up before every agency relationship fails — and the 4 signals to catch it early.
Key takeaways
- All three agency failures had the same root cause: the agency optimized for the appearance of work rather than the work itself — vanity metrics, padded reports, activity theater.
- The pattern appears in weeks 6 to 10 of every bad agency relationship. By week 12, if it is not fixed, the relationship will not recover.
- The 4 early warning signals: reports shift from outcomes to activity, response times lengthen, revision requests increase, and the agency starts citing external factors for every miss.
- Firing an agency costs 30-60 days of dead-weight payments and 30-90 days of vendor transition. Starting the exit at the first red flag saves 3-6 months of sunk cost.
The honest answer
I fired three marketing agencies in 18 months. Each was different in size, specialty, and price. One was a boutique SEO firm at $3,500/mo. One was a full-service agency at $18,000/mo. One was a paid ads specialist at $6,000/mo plus 12% of spend. Same root cause in all three.
The root cause: the agency optimized for the appearance of work rather than the work itself. Vanity metrics. Padded reports. Activity theater. The engagement looked productive in the monthly deck and was not producing any business results. By the time I could prove it, I had paid for 4 to 6 months I should not have.
This post is the pattern, the early warning system, and the 30-day exit playbook. If you are reading this because you already suspect your agency is failing, you are probably 4 to 6 weeks behind where you should be. Start the audit today.
Agency 1: the SEO firm that ranked the wrong keywords
The SEO firm was competent by surface metrics. Monthly reports showed consistent ranking improvements. Domain authority up 7 points over 6 months. New keywords in positions 4 through 10. The deck looked excellent.
Month 7: I pulled Google Search Console myself. The keywords they were tracking were not the keywords we had agreed to target. The keywords they ranked were accurate — but they were long-tail variants with 10 to 50 monthly searches. Our agreed targets, the ones with 2,000 to 8,000 monthly searches, had not moved.
When I surfaced this, the agency said the target keywords were "more competitive than initially projected." That phrase — "more competitive than initially projected" — is the first of the four death signals. The agency found a path to looking successful without being successful. The monthly report tracked the keywords they could win. Not the keywords I hired them to win.
I gave them 60 days to show movement on the actual target list. Nothing changed. I issued termination notice on Day 61. The 60-day notice clause in the contract cost me $7,000 of dead-weight payments. My 6-month sunk cost was $21,000.
Agency 2: the full-service agency with beautiful decks
The full-service agency was expensive and polished. $18,000/mo covered SEO, paid ads, content, and social. The monthly call was 90 minutes. The deck was always 30 slides. The account manager was responsive and articulate.
Month 4: I started tracking output independently. I built a spreadsheet logging every deliverable received versus every deliverable promised in the monthly plan. Month 4 score: 11 deliverables promised, 7 received. The 4 missed items were rescheduled to month 5 without notification.
Month 5: 13 promised, 8 delivered. The missed items were absorbed into the month 5 promises without acknowledgment. When I raised the tracking data on the month 5 call, the account manager said we were "prioritizing quality over quantity" — the second death signal.
"Quality over quantity" is what agencies say when they are under-delivering and want to reframe the under-delivery as a strategic choice. Quantity was what the contract specified. Quality was an adjective added to avoid accountability.
I fired them at the end of month 5. The 90-day notice clause cost me $54,000 of dead-weight payments. My 5-month sunk cost was $90,000 plus $54,000 of exit payments. The full-service agency was the most expensive mistake.
The lesson: the 90-day notice clause in a high-ticket retainer is a $54,000 to $162,000 exposure depending on retainer size. Read it before you sign. Negotiate it to 30 days before the ink is dry. Do not learn this the way I did.
Agency 3: the paid ads specialist with attribution theater
The paid ads agency was the smallest engagement and the fastest to fail. $6,000/mo plus 12% of spend, managing Google and Meta ads for a product launch.
Month 2: ROAS reporting showed a 3.2x return. The number looked solid. But when I cross-referenced the ad-platform revenue attribution against actual sales in the CRM, the numbers did not match. The ad platform was attributing revenue that the CRM could not find. View-through conversions, assisted clicks, post-view engagement — all counted as revenue in the agency's reporting model.
When I built a last-click attribution report from the CRM, actual ROAS was 1.1x. Profitable. Barely. Not 3.2x.
The agency's defense: "Last-click attribution underrepresents the true value of upper-funnel advertising." That is technically true for brand awareness campaigns. We were running direct-response product ads. Last-click attribution is the correct model.
The third death signal: the agency chose an attribution model that made their results look better rather than the attribution model that accurately reflected business performance. The data was real. The interpretation was chosen for optics.
I fired them at the end of month 3. The 60-day notice clause cost me $18,000 of dead-weight payments. My 3-month sunk cost was $18,000 plus $18,000 exit. Total: $36,000 for a ROAS of 1.1x.
The pattern across all three
I did not identify the pattern until I was writing notes for the third agency exit. When I laid all three side by side, the root cause was identical.
In all three cases, the agency shifted from optimizing for my business results to optimizing for the appearance of results. The shift happened in weeks 6 to 10 of each engagement. By week 12, the pattern was locked in. The agency had found a stable equilibrium where they could satisfy the reporting requirements without producing the business outcomes.
This is not malicious in most cases. It is structural. The agency's incentive is to keep the account. Keeping the account requires satisfying the monthly review call. The monthly review call is satisfied by a deck that looks good. The deck that looks good does not have to reflect business outcomes. It has to reflect metrics the agency can move.
The structural fix — for both agencies and clients — is to define success criteria in business outcomes before the engagement begins and track them independently throughout. Not impressions. Not rankings on agency-selected keywords. Not attribution models the agency controls. Revenue. Leads. Cost per acquisition. Numbers that come from your systems, not theirs.
The 4 early warning signals
These appeared in all three engagements. They appear in weeks 6 to 10. If you see any of them, you have 4 to 6 weeks to turn the engagement around before the pattern is irreversible.
| Signal | What it sounds like | What it means |
|---|---|---|
| Metric substitution | "We're now tracking engagement rate instead of leads" | The agency cannot hit the agreed metric, so they found one they can hit |
| Capability exculpation | "More competitive than projected" / "external factors" | The agency is building a defense for missed targets before you notice |
| Activity inflation | "We published 14 pieces this month" (none converted) | Volume replaces quality when the quality conversation is too hard |
| Response lag | Emails taking 48 hours; calls rescheduled twice | Engagement is deprioritized because the account is unstable |
See two or more of these signals in the same month: call an emergency review. Not the regular monthly call. A standalone call with the account lead, the delivery lead, and any data you have collected independently. Make it uncomfortable. Comfortable calls at this stage cost money.
The 30-day exit playbook
You have decided to exit. The contract has a notice period. Here is the 30-day sequence.
Day 1: Read the notice clause. Note the notice period, unused-hour treatment, and work-product transfer requirements. Screenshot the clause. Time-stamp it.
Day 2-5: Build an inventory of every deliverable promised versus delivered, in writing. Email it to the agency and ask them to confirm it within 7 days. Their response (or non-response) becomes part of your exit record.
Day 6: Issue termination notice in writing. Cite the specific contract clause. State the termination date. Request that asset transfer begin immediately and complete within 30 days: ad accounts, analytics access, content files, source code, CRM integrations, credentials.
Day 7-21: Monitor asset transfer. Follow up in writing every 7 days. Productized agencies complete transfers in 48 hours. Traditional agencies drag for 60 to 90 days unless you push. If transfer stalls beyond 30 days, escalate to a formal written demand with a deadline.
Day 22-30: Review the final invoice. Reject any line items not explicitly required by the contract. "Wrap-up time," "knowledge transfer hours," and "final review" charges are typically not contractually required. Most agencies will remove them rather than escalate.
The exit will cost you the notice-period fees. It will save you 6 to 18 months of the same fees. Run the math.
How to avoid all of this on the next agency
Three practices that prevent the pattern from repeating.
First, define success in business outcomes before the first invoice. Not "improve our SEO." Not "increase brand awareness." "275 MQLs per month by Day 90, measured in our CRM using last-click attribution." Put it in the contract.
Second, track deliverables independently from day one. Build the spreadsheet on Day 1 of the engagement, not when you suspect something is wrong. Agencies that know you are tracking perform better.
Third, negotiate the notice period to 30 days before signing. A 90-day notice period on a $15,000/mo retainer is a $135,000 downside exposure if the relationship fails. That is not a contract term. That is a hostage clause.
See Striveloom's services page to understand how we structure delivery accountability from day one. We build outcome tracking into every engagement, not as a client request, but as a standard operating procedure.
What this means in practice
The pattern shows up in weeks 6 to 10. You will feel it before you can prove it. Monthly reports that look fine but leave you uneasy. A sense that activity is high and results are thin.
Trust the feeling. Pull the data. Cross-reference the agency's metrics against your own systems. Build the deliverable inventory. Do it in week 6, not week 16.
If the data confirms the feeling, call the emergency review. Give the agency 30 days to show movement on the agreed business outcomes. If they cannot, start the exit.
The best agencies welcome independent tracking. They want you to verify the numbers in your systems. The agencies that resist independent tracking are telling you why they resist it.
See Striveloom's pricing page to understand what outcome-accountable agency work costs in 2026. Compare it to what you are paying now and what you are getting for it.
Frequently asked questions
How do you know when it is time to fire your marketing agency?
Four signals: the agency substitutes easier-to-hit metrics for the ones you agreed to track (metric substitution), they cite external factors for every miss without a concrete recovery plan (capability exculpation), they inflate activity volume when outcome quality declines, and their response times lengthen. Any two of these in the same month warrant an emergency review. All four in the same month: start the exit. The pattern, once established in weeks 6 to 10, rarely reverses without a change in account leadership.
How much does it cost to exit a marketing agency mid-contract?
The direct cost is the notice-period fees. On a $10,000/mo retainer with a 60-day notice period, the exit costs $20,000 in dead-weight payments. On a $15,000/mo retainer with a 90-day notice period, it is $45,000. The indirect cost is vendor transition time: 30 to 90 days to onboard a replacement agency and rebuild institutional knowledge. The total cost of a failed agency relationship is typically 6 to 9 months of fees — which is why catching the pattern in weeks 6 to 10 instead of months 6 to 9 matters so much.
What should I do before firing an agency if I depend on their services?
Three steps before issuing notice. First, identify your replacement vendor and confirm they can take over within 30 days of notice. Second, inventory all assets the current agency controls: ad accounts, analytics access, CMS credentials, content files, automation workflows. Third, issue a formal asset-access request to the current agency before you issue termination notice — some contracts require the agency to transfer access only after the notice period ends, which creates a gap. Getting a head start on asset transfer reduces the risk of a service gap.
What is the most expensive mistake buyers make with marketing agencies?
Signing a contract with a 60 to 90 day cancellation notice period without negotiating it down to 30 days. This single clause turns a failed agency relationship into a 2 to 3 month penalty. On an $18,000/mo retainer, a 90-day notice period is a $54,000 exit cost. Most agencies will accept a 30-day notice period if the buyer asks before signing. Almost none will renegotiate it after the relationship has broken down. The leverage window is the signature call, not the exit call.
How do I track agency performance independently without their reporting tools?
Four sources of independent data: Google Analytics (direct access, not screenshots from the agency), Google Search Console (direct access via your Google account, not the agency's), your CRM (leads and pipeline attributed to marketing channels), and your ad platform accounts (Meta Ads Manager, Google Ads — the agency should never be the sole account owner). Build a simple spreadsheet on Day 1 tracking deliverables promised versus received and key business metrics from your own systems. Compare it to the agency's reports monthly.
Should I give the agency a second chance after discovering the warning signs?
One second chance, with specific conditions. Issue a formal written notice identifying the specific metric substitutions, missed deliverables, or attribution choices you discovered. Give the agency 30 days to show movement on the originally agreed business outcomes using your measurement systems, not theirs. If they show meaningful progress in 30 days, the relationship is recoverable. If they miss a second time or push back on independent measurement, the pattern is structural, not situational. Exit at day 31.
Sources & further reading
- 1Agency Pricing and Contract Trends 2025 — SoDA / Society of Digital Agencies, 2025
- 2Marketing Attribution Models: A Practical Guide — Google Marketing Platform, 2024
- 3$100M Offers — Risk Reversal and Client Retention — Acquisition.com, 2023
- 4Managing Outside Agencies and Vendors — Harvard Business Review, 2024
About the author
Founder of Striveloom. Software engineer turned operator, building the agency that ships like software — one team, one pipeline, one platform. Writes about AI agencies, web development, marketing automation, and paid advertising.
Continue reading
Why Most Digital Agencies Are Time-for-Money Traps (And How to Spot One Before You Sign)
Most agencies bill hours so they make more money when projects take longer. Here is the exact pattern, the four contract clauses to delete, and a checklist that has saved buyers $40K on average.
The $10K/mo Retainer Trap Nobody Warns You About
Retainers were a discount mechanism. They became Stockholm syndrome. Here is the math, the four clauses to renegotiate, and the exit playbook for $120K of sunk-cost retainers.
I Read 100 Agency Proposals. 92 Are Doing This Wrong.
92 of 100 agency proposals contained the same 8 flaws: padded hours, hidden scope, buried cancellation, vanity metrics. Here is the teardown.
Ready to work with us?
Book a free 30-minute call to scope your project. Fixed pricing, transparent timelines.