Most agency relationships that get fired should have been renegotiated. A smaller number that get renegotiated should have been ended six months earlier. The difference between the two is rarely about how bad things feel; it’s about which problem you’re actually looking at, and whether the agency is capable of fixing it.
The short version: if the issue is drift in output, scope or account attention, a structured renegotiation is usually the better first move. If the issue is lost trust, a capability ceiling, repeated missed commitments or senior people disappearing without a credible plan, it’s time to go.
Should you renegotiate before switching agency?
A working agency relationship is an asset that took time to build. The senior account person knows your brand voice. The developers know which integrations break when you push them. The strategist remembers the campaign that flopped two years ago and why. Replacing that institutional memory costs more than people expect, and the replacement isn’t guaranteed to be better.
So the default, the position you should argue yourself out of rather than into, is that the relationship is worth saving. That doesn’t mean tolerating bad behaviour. It means starting from the assumption that a frank conversation, properly run, will fix more than a switch will.
The exception is when the problem is structural rather than performative.
What’s gone wrong?
Before deciding anything, name the problem. Buyers tend to arrive at ‘we should switch’ through a stack of small irritations rather than a single rupture, and the small irritations often point to different underlying causes that respond to different treatments. Six common diagnoses, drawn from what we see most often. Most real situations are some combination of two or three.
The work has slowed but the invoices haven’t. Output drift. Possibly a resourcing problem at the agency end, possibly that your account is being de-prioritised in favour of newer or larger clients.
You’ve stopped hearing from the senior people you signed with. Account drift. The strategist or director who sold you the engagement has handed you off to a junior, and nobody told you that was happening.
Every request now comes back as a change order. Scope hardening. Often a reaction to a previously over-scoped retainer, but it can curdle into a defensive posture where the agency treats you as an adversary.
The agency is good, but you’ve outgrown them. Capability ceiling. Common when you started with a generalist agency and your needs have specialised, or your scale has tripled and they’re still running the same playbook.
Nothing’s wrong, exactly, but you can’t point to anything excellent. Mediocrity drift. The most common one. Months of competent-but-forgettable work, no obvious failure to flag, no obvious success to celebrate.
You don’t trust the numbers on the invoice any more. Trust drift. The serious one. Once a buyer starts checking timesheets line by line, the relationship is on a clock.
Each of these maps to a different answer. Output drift, account drift and scope hardening are usually renegotiable. Capability ceiling and trust drift usually aren’t. Mediocrity drift is the hardest call.
What switching actually costs
The cost most often underestimated by people considering a switch is the cost of the switch itself. The new agency’s headline fees may not look wildly different from your current ones. The real cost is everything else.
In our experience, a new agency rarely reaches full productivity immediately. Expect several months of lower output while they learn the brand, the stack, the stakeholders, and the history of decisions that shaped where you are today. During those months, you’re paying a full retainer for partial output.
The internal cost is real too. Briefing the new agency, transferring documentation, reviewing initial work that misses the mark because they don’t yet know your context, managing the handover from the outgoing supplier: it can easily consume days of senior internal time. That time has to come from somewhere, usually from the in-house owner who already feels stretched.
Expect things to break that weren’t broken before. There is a new-supplier rebuild bias: the incoming team will often want to redo work they didn’t build, partly because that’s how they understand it, partly because rebuilding is easier to sell than maintaining. Some of those rebuilds will be improvements. Some will reintroduce bugs your current agency quietly fixed two years ago and forgot to document.
None of this is an argument against switching. It’s an argument for not switching unless the gain is large enough to clear the cost. A 15% improvement in service quality doesn’t clear it. A 50% improvement does, if you can predict it; the honest answer is you usually can’t.
Signals that mean renegotiate
If the underlying problem is one of the three renegotiable categories above (output drift, account drift or scope hardening), the right move is a direct conversation, not a search. A few specific signals point this way.
The agency still has the right people; they’re just not on your account any more. Ask for them back, or ask why they’ve gone.
The retainer was set up two or three years ago and hasn’t been revisited since your needs changed. Most agencies will quietly admit a retainer has drifted out of shape if you raise it before they have to. We’ve covered why that drift happens in day rates, retainers and project fees.
The agency’s senior leadership still answers your emails within a day, and admits problems exist when you raise them. The willingness to engage honestly is the single best predictor that a renegotiation will work.
You can describe what you want differently in one sentence. ‘More strategic input, fewer reactive deliverables.’ ‘Two clear initiatives a quarter, not 12 half-finished ones.’ If the change you want is articulable, it’s negotiable.
Signals that mean it’s time to leave
A smaller set of signals mean the relationship is structurally finished and a conversation won’t fix it. These are the ones we’d treat as decisive.
Repeated missed commitments after a frank conversation. A first slip is a mistake. A second slip after the agency has explicitly acknowledged the first and committed to a fix is a pattern. We’d give one structured conversation, a written agreement on what changes, and 90 days. If the same problem recurs in that window, the agency either can’t fix it or won’t.
Loss of senior staff with no credible continuity plan. If the strategist, technical lead or account director who knows your account leaves and the agency’s response is ‘we’ll find someone’, that’s a capability hole, not a transition. Watch what they do in the first 30 days. Real continuity looks like a named replacement, a structured handover and a discount while the new person ramps. Vague reassurance is the signal.
A capability ceiling the agency can’t credibly clear. Your needs have moved past what the agency can deliver, and they know it. The honest agencies will tell you. The less honest ones will keep selling you a level of service they can’t execute. Either way, staying is a sunk-cost decision.
A trust breach on billing or transparency. Invoices that don’t reconcile to work done. Hours that appear without explanation. A refusal to break down a retainer when asked. work was subcontracted without disclosure. Each is, on its own, usually enough to end the relationship. In most cases, this is not a fixable category. Once trust on numbers is gone it rarely comes back, and every subsequent interaction will be contaminated by the suspicion.
If you recognise more than one of these four, you’re not deciding whether to leave. You’re deciding when.
How to run a renegotiation that works
If you’ve diagnosed a renegotiable problem and the agency is the kind of operation that engages honestly when challenged, here’s the conversation worth having.
Write down, before the meeting, three things: what specifically isn’t working, what good would look like in concrete terms, and what you’re prepared to change on your side. The last one matters. Most failed agency relationships are mainly the agency’s responsibility, but rarely entirely, and starting the conversation by acknowledging your share tends to draw out the other side’s faster than anything else.
Ask for the meeting with the most senior person you have access to, not your day-to-day account contact. The day-to-day person often can’t authorise the changes you’re asking for, and putting them in the middle wastes everyone’s time.
Bring specifics. The campaign brief took 11 days when we’d agreed on five’ is workable. ‘Things feel slow’ isn’t. The agency can only respond to evidence, and the act of assembling it usually clarifies your own thinking too.
Agree changes in writing, with timeframes. The output of the meeting should be a short document (an email is fine) that names what’s changing, who’s responsible, and when you’ll review. Without that, the conversation tends to produce two weeks of better behaviour followed by a return to the same pattern.
Set a review point 90 days out. Long enough for changes to take effect, short enough that you’re not letting another six months drift past while hoping. If the review point arrives and the agreed changes haven’t happened, you’ve already made the decision to leave. You just haven’t actioned it.
If you’ve decided to leave: what to do before you start the search
Two things to do before sending the first RFP.
First, work out what you actually want this time. The natural instinct after a bad relationship is to write a brief that’s a list of the things the current agency got wrong. That produces a new agency that’s good at not being your old agency, which isn’t the same thing as being right for you. Spend a week writing a brief that describes what good looks like in its own terms, with the current agency’s failures used only as anti-patterns rather than as the spec.
Second, get your house in order before the new agency arrives. Our agency handover checklist covers the operational side: credentials, accounts, repositories, documentation. The strategic side matters as much. Who internally owns the relationship? Which decisions did the old agency make that you should reclaim? Which of their recommendations did you ignore that you now wish you hadn’t?
Then run the search properly. Comparing your current retainer against new quotes is harder than it looks; the pricing structures rarely line up cleanly, and we’ve covered that in why agency quotes vary so much. The headline number is almost never the comparable number.
The discipline through all of this is honesty about the problem you’re actually solving. If you can name the problem in a sentence, you can usually tell whether your current agency can fix it. If you can’t, the next agency probably can’t either.
