The deck says ROAS is up. Your CAC says otherwise. The agency celebrates a record month of impressions while your blended acquisition cost quietly climbs for the third quarter in a row. That gap between what they report and what you feel is not a misunderstanding. It is usually the first sign the relationship has stopped working.
Not every bad quarter means a bad agency. Campaigns need time, and honest testing can produce soft months. But when stale creative, recycled strategy decks, and rotating account managers show up alongside the flattering reports, the pattern is harder to explain away.
The seven signs below help you tell process failures from business problems, size the cost of switching, and protect your accounts and data if you leave.
The short version: Repeated missed work, unexplained results, and an agency that does not change course are reasons to review the relationship. Compare those failures with the agreed scope and give the team a chance to answer specific questions. If you switch, protect your accounts, data, and creative first. A weak product or offer needs a different fix.
What Are the Signs You Should Switch Marketing Agencies?
Consider switching when CAC rises behind positive platform reports, creative and strategy stop changing, or the account team keeps rotating. Watch for fees outpacing results, unanswered incrementality questions, and replies slowing from hours to days. One sign deserves a conversation. Two or three together form a pattern.
Here is the full list in one view. The sections below unpack each group.
| Sign | What it usually means | What good looks like |
|---|---|---|
| Reporting celebrates ROAS while CAC rises | Metrics chosen to flatter the work | Blended CAC and payback lead every report |
| Same ads running 90+ days | No creative testing engine behind the account | Fresh variants on a monthly cadence |
| Strategy deck unchanged for 2 quarters | The account is on autopilot | Quarterly strategy built on documented learnings |
| Account team keeps rotating | Agency churn, or your account got de-prioritized | A stable senior lead for 12+ months |
| Fees scale with spend, results do not | Incentives reward spending, not outcomes | Fees tied to scope or performance bands |
| Nobody can answer the incrementality question | Attribution is doing the agency's marketing | Holdout or geo tests at least once a year |
| Response time went from hours to days | You slid down the priority list | Same-day replies and a weekly live check-in |
Is the Reporting Telling You What a Customer Costs?
Healthy agency reporting leads with the numbers your CFO cares about: blended CAC, payback period, and revenue by channel. Failing agency reporting leads with the numbers the platform hands out: ROAS, CTR, impressions. When those two views diverge for more than a quarter, the report is protecting the agency, not informing you.
Rising acquisition costs are partly a market condition. According to Paddle, CAC has climbed roughly 60% across B2B and B2C over five years. A good agency names that pressure in the report and shows what it is doing about it. A bad one buries it under a record month of impressions.
Incrementality belongs in the same conversation. Ask your agency a simple question: if we turned brand search and retargeting off for 2 weeks, what would happen to revenue? A strong team has run a holdout or geo test and can answer with data. A weak one will change the subject to attribution windows.
Has the Work Stopped Evolving?
Stale work shows up in two places: the ads and the strategy. If the same creative has run for 90+ days untouched, there is no testing engine behind your account. If the quarterly strategy deck reads like last quarter’s with new dates, the account is on autopilot and you are paying retainer prices for maintenance.
Creative is the expensive one to neglect. Meta’s research on creative quality found creative drives more than half of ad performance variance on its platforms. An agency that stopped iterating has parked your single largest performance lever and kept billing you for the parking spot.
What good looks like: a monthly creative cadence where each variant tests a named hypothesis, and a quarterly strategy review that opens with what was learned, what changed, and what gets tested next. If you want a structured way to check this yourself, the 7-stage paid media audit walks through testing velocity in detail. That cadence is the baseline in how we manage paid media accounts.
Is the Relationship Itself Degrading?
Three of the seven signs concern the relationship: staff rotation, fees growing faster than results, and replies slowing from hours to days. Look at these alongside the work itself. They signal that your account has slipped down the agency’s priorities, even if nobody says so.
Team rotation is the most common. When your third account manager in a year introduces herself, your institutional knowledge walked out the door for the third time. Good agencies keep a senior lead on an account for 12+ months and name a backup who already knows the history.
The fee structure tells you about incentives. A percentage-of-spend model that scales fees while CAC drifts up rewards exactly the wrong behavior. Across $50M+ in managed paid media, the engagements that lasted were the ones where fees tracked scope or performance bands and got renegotiated as spend grew. That is how we structure our own engagements, because the alternative quietly taxes growth.
Response time often gives the first warning. Agencies answer their priority clients quickly, so a shift from hours to days suggests your account has become less important. The work usually suffers next.
What Does Switching Agencies Cost You?
Switching costs you 60 to 90 days of disruption, a temporary performance dip while algorithms relearn, and the internal time to manage a handover. Plan a 30 to 60 day overlap between agencies so the new team inherits live learnings instead of starting from zero. Budget for the dip. It is real, but survivable.
Use weeks 1 and 2 to audit access, review the contract, and onboard the new agency. In weeks 3 to 6, run both teams in parallel: the outgoing agency maintains campaigns while the incoming team rebuilds. Complete the handover in weeks 6 to 8, then offboard the old team and archive the documentation.
Be realistic about the dip. Restructured campaigns re-enter learning phases, and CAC often rises 10 to 20% for a few weeks before it settles. Founders who expect that ride it out. Founders who do not expect it often panic at week 3 and start doubting the new agency for inherited reasons.
How Do You Leave Without Losing Your Data?
Before giving notice, confirm that your company owns every account. Pixel history, ad account learnings, and audience data may sit in assets registered to the agency. If so, negotiate the transfer before ending the relationship. That order protects the data your new team needs.
Run this checklist before the notice email goes out:
- Admin access on Business Manager and every ad account, held by a company email
- Pixel and dataset ownership transferred to your Business Manager
- GA4 and Tag Manager admin held internally
- Domain verification under your own Business Manager
- Customer lists and saved audiences exported or shared to assets you own
- Creative source files and the testing log delivered
- Historical reports and learning docs archived on your drive
- Billing moved to your card or invoice account
Losing pixel history means the next agency starts optimization from a cold signal base, which extends the dip from weeks to months. It is the most expensive mistake in the whole process and the easiest to prevent. And before you sign with the replacement, run them through the questions to ask before hiring a growth agency so you are not rediscovering these signs in a year.
When Is Switching the Wrong Move?
Switching is the wrong move when the signs point at outcomes but the process is healthy. If the agency tests on cadence, reports CAC without spin, communicates fast, and results are still soft, the constraint is often upstream: a vague brief, a weak offer, or positioning the ads cannot fix. A new agency inherits the same constraint.
The useful test is to separate process signs from outcome signs. Stale creative, rotating teams, and slow replies are process failures, and those belong to the agency. Flat results despite good process usually mean the brief needs rewriting before the contract does. I have watched a SaaS client fire two competent agencies in 18 months when the real problem was an offer their market did not want.
Before switching in that situation, rewrite the brief around a clearer ICP, a real offer, and one primary metric. Give the current team a quarter to work from it. If nothing improves, you will have a clearer reason to leave, and the ownership checklist above will help protect the handover.
Weighing a different model entirely? Marketing agency vs freelancer vs in-house team compares the three paths.
That gap from the opening, the one between what the deck says and what your CAC says, has an answer. Sometimes it is a process failure the agency owns. Sometimes it is a business problem a new agency would inherit. Separate the two before you decide.
A recent report and the agreed scope give us a starting point for an outside view. Book a Free Strategy Call and we’ll work through whether a clearer brief, a process change, or a new team fits the problem.
Like this? Get the next one.
Short emails. New posts as they ship.
