Short answer: Switch marketing agencies when your own sales data shows flat results after a fair period (about three months for paid media, six to twelve for SEO), when deliverables keep slipping, when you cannot access your own data, when account leads keep changing, or when their approach no longer fits your stage. Give them a written 30 to 60 day improvement plan first, then decide.
Key takeaways
- Five signals justify switching: flat results in your own sales data after a fair period, repeated missed deliverables, no access to your data, constant account lead turnover, and a strategy that no longer fits your stage.
- Four things do not justify switching yet: a normal ramp-up period, seasonality, one bad month, and problems that start on your side.
- Before you leave, give the agency a written 30 to 60 day improvement plan with specific, measurable targets taken from your own data.
- Switching has a real cost: a new ramp-up of two to three months, lost account learning and the notice period you still pay for. Put numbers on it before you decide.
- Secure logins, data and creative assets before you give notice, not after.
Most owners who switch agencies switch either too early or too late. Too early, and you pay for a second ramp-up just as the first one was about to pay off. Too late, and you have spent another six months of fees on an account that stopped improving long ago.
This guide is only about the timing of that decision: which signals mean it is time, which ones do not, how to give the current agency a fair last chance, and how to weigh the cost of moving against the cost of staying.
What are the signs it is time to switch agencies?
It is time to switch when the problem is persistent, sits with the agency, and has survived at least one direct conversation about it. In our experience, five signals meet that bar.
Results are flat in your own sales data after a fair period
Judge on revenue, new customers and acquisition cost from Shopify or your CRM, not on the ROAS or click numbers in the agency's report. If new customers and contribution margin have not moved after about three months of paid search or paid social, or six to twelve months of SEO, and the agency cannot show a credible reason, the strategy is not working. Rising platform numbers alongside flat sales are a warning, not a defence.
Deliverables are missed again and again
One late report happens. A pattern of promised tests that never launch, creative that arrives weeks late and monthly reports that stop showing up tells you the account is not getting the attention you pay for. Keep a simple log: what was promised, by when, and what happened. Three or four misses in a quarter, after you have raised it, is a clear signal.
You cannot get access to your own data
If the ad accounts, analytics property or tag manager sit in the agency's name and requests for admin access are slowed down or refused, that is a reason to act now, regardless of performance. You cannot judge results you cannot see, and it only gets harder to fix the longer it runs.
Your account lead keeps changing
A new main contact every few months means each one spends weeks relearning your business, and decisions lose continuity. One change in a year is normal. Three is a sign the agency is stretched or that your account is low on its list.
Their strategy no longer fits your stage
The agency that took you from zero to your first steady revenue may not be the one to scale you across new markets or channels. If you now need things they do not do well, such as proper measurement, a second channel or international expansion, and they keep offering more of the same, the fit has run out even if nobody did anything wrong.
Which problems do not justify switching yet?
Some of the moments that make owners want to leave are normal parts of any agency relationship. Switching on these resets the clock without fixing anything.
The last row is the one owners skip most often. Before you blame the agency, check what changed on your side: did you run out of best-sellers, raise prices, change the website, take three weeks to approve creative or cut the budget mid-test? If so, a new agency would face the same problems. Be honest about it, and write down what you will fix too.
The opposite mistake is using these reasons to excuse a real problem. "It is seasonal" only holds if you have compared against the same period last year. "SEO takes time" only holds if technical fixes, content and rankings are moving in the meantime.
A performance marketing agency that reports on your sales data
We agree outcome targets from your Shopify or CRM figures at the start, keep every account in your name, and show you the tests that lost as well as the ones that won.
See our performance marketing agencyHow do you give your current agency a fair last chance?
Put the problem in writing and agree a 30 to 60 day improvement plan with specific measures, a named owner and a date for the verdict. That turns a vague feeling into a test the agency can pass or fail, and it protects you from switching on frustration alone.
Thirty days suits problems of service and process, such as late reports or missed launches. Sixty days suits performance problems, because paid media needs a few weeks to show the effect of real changes.
What to put in a written improvement plan
- The problem in numbers from your own data, for example "new customers flat at about 300 a month for four months while spend rose 25 percent".
- Two or three targets for the period, each with a number and a source, for example "new customer CAC below €60 in Shopify by day 60".
- What the agency will change, by when, and who on their side owns it.
- What you will do on your side, such as approving creative within two working days.
- Check-in dates (every two weeks works) and the date the plan ends.
- What happens at the end: continue as normal, extend once with new terms, or end the contract.
Ask for full admin access to every account in the same email. You need it whatever the outcome.
How the agency responds tells you as much as the result. A good one will push back on targets it thinks are unfair, propose its own, and treat the plan as a chance to reset. If the reply is defensive or vague, or the plan quietly stops being followed after two weeks, you have your answer before the 60 days are up. Our list of questions to ask a digital marketing agency works just as well in this conversation as it does before signing.
What does switching cost compared with staying?
Switching costs a notice period you still pay for, a ramp-up of usually two to three months while the new agency learns your account, and some lost learning from tests and audiences built up so far. Staying costs whatever an underperforming agency fails to earn each month. Put rough numbers on both before you decide.
The situation
Agency fee €4,000 a month, 60 day notice period. New customer CAC is €70 against a target of €55. About 300 new customers a month, each worth roughly €60 in margin over their first 90 days.
Cost of switching
Notice period: €8,000 in fees you pay either way. Ramp-up: assume CAC stays at €70 for three months while the new agency rebuilds, so no gain yet. Your own time for briefing and handover: a few days.
Cost of staying
If a better agency could bring CAC to €55 on the same spend of €21,000, you would get about 380 new customers instead of 300. That is 80 more a month, worth about €4,800 in margin, lost every month you stay.
The verdict
Over 12 months, switching gives up roughly five months of that gain to notice and ramp-up, leaving about seven months, or around €33,000 in margin. If the improvement plan had failed, switching wins clearly. If the gap were €55 versus €58, it would not be worth the disruption.
These numbers are illustrative, and the gain is never guaranteed. The point of the exercise is to stop the decision resting on mood. Two rules of thumb we use: if the expected gain does not repay the switching cost within about six months, fix the current relationship instead; and if the problem is data access or a strategy that no longer fits your stage, switch even when the numbers are close, because those problems tend to get worse.
A new agency ramping up during your busiest weeks costs the most. If the decision falls two months before peak, consider giving notice afterwards, or keep changes to the account minimal until it passes.
What should you sort out before you give notice?
Get admin access to every ad account, analytics property, tag manager container, email platform and website login, export historical reports and audience lists, and collect all creative files and copy, all before you send the notice email. Agencies are usually cooperative, but it is far easier to fix access while the relationship is still friendly. If you want an independent check that tracking and reporting are complete before you hand over, marketing analytics services can audit the setup so the next agency starts from clean data.
Frequently asked questions
As a rule of thumb, about three months for paid search or paid social and six to twelve months for SEO before you judge outcomes. Problems with service, honesty or data access do not need to wait that long. Raise them straight away.
Yes, in the form of a written improvement plan with targets and an end date. Most agencies would rather know and have a chance to fix things. Saying nothing and leaving suddenly also leaves you less time to secure your accounts and data.
It can be. If each agency failed for a similar reason, look at the common factors: unclear targets, limited budget for testing, slow approvals or a product and margin that make acquisition hard. Fixing those first makes the next relationship far more likely to work.
Yes. If one channel is working and another is not, moving only the weak channel limits the switching cost. The trade-off is that two agencies need clear ownership of shared tracking and reporting, or they will both claim the same sales.
Count the notice period as part of the switching cost in your calculation, and start the improvement plan early so its end lines up with your notice date. Use the notice period to secure access and export data, since you are paying for it anyway.
What to do next
Make the decision on evidence, not on the last frustrating call.
- Pull the last six months of revenue, new customers and spend from your own systems and check them against the five signals above.
- If at least one signal holds, write a 30 to 60 day improvement plan and send it with a request for full admin access.
- At the end of the plan, run the cost of switching against the cost of staying with your own numbers, then decide.
If you want an outside view before you decide, we offer a free written audit within 48 hours of your accounts, tracking and results against your actual sales. If you do switch, our guide on how to choose a marketing agency covers the next step, and a digital marketing agency that covers several channels can reduce the number of handovers you face.