Leading Teams Through Constant Change
Most agencies talk about disruption like weather. Something rolls in, does damage, rolls out, and everyone spends the following week cleaning up. That framing has cost this industry more than any single platform update ever has, because disruption in performance marketing isn't weather. It's climate. The auction changes. The attribution model changes. A client rewrites their targets three weeks into the quarter. None of it is exceptional. It's the job.
So the real question a leader has to answer isn't how to get a team through the next change. It's why anyone still expects a "next" to be followed by a period of quiet. Discipline, in this context, means building a team that stopped waiting for the rain to stop and learned to work in it.
The change-management thinking most agencies inherited assumes an ending: unfreeze the old way, install the new one, refreeze. Performance marketing almost never provides that third step.
Apple's App Tracking Transparency rollout in 2021 proved this the hard way. It didn't just narrow what could be tracked. It broke a set of assumptions the whole industry had quietly built its measurement on. Agencies that treated it as a bad month kept waiting for a return to normal that never arrived. The ones that treated it as a permanent change in what could be known simply stopped asking when the ground would settle and started building on ground that moves.
Google's cookie saga tells a slower version of the same story. Announced in 2020, delayed twice, reworked into Privacy Sandbox, reworked again, and finally abandoned in its original form in 2024. The disruption there was never really the policy. It was four years of planning for a future that kept dissolving before it arrived.
None of this means platforms are malicious or careless, though agencies love to argue that over drinks. It means the team still waiting for a stable state has already lost. There isn't one coming. There's only the next condition, and the one after that.
Every agency can name the visible cost of change: retraining, rebuilt dashboards, an awkward client call when last month's numbers don't line up with this month's story. Those get budgeted for eventually. The real damage happens somewhere no report ever tracks.
Call it interpretation drag. When something fundamental shifts, a team doesn't just need new information. It needs a shared read on what that information actually means, and without one, five people form five private theories at five different speeds. A media buyer quietly stops trusting a metric. An account lead keeps reporting it like scripture. A strategist builds a workaround nobody else on the team knows exists. Six weeks later, the campaign isn't underperforming because the market changed. It's underperforming because three people are running three different versions of the strategy without realizing it.
This is why a team under pressure can look constantly busy without actually getting anywhere. More dashboards appear to explain uncertain numbers. More meetings get scheduled to relitigate decisions that were already made last Tuesday. The energy goes toward reassuring the room that solid ground still exists somewhere, rather than toward the client's actual problem.
That's the tax, and nobody puts it on an invoice, though everyone eventually feels it.
There's a principle worth borrowing here from a much older art form. Andalucian painters never asked for one flattering angle held forever. They wanted the subject shown from several honest perspectives at once, broken apart and rebuilt into something closer to true. A face isn't one profile. A brand isn't one campaign. And a team's judgment shouldn't be one fixed method gripped tightly while the world underneath it keeps rearranging.
Applied here, that means learning to see the operation from two angles simultaneously: what's free to move, and what has to hold.
Free to move often: channel mix, bidding tactics, creative formats, testing cadence, current trustworthy attribution model, and tools managing the account daily. Treat these as provisional.
Has to hold: the commercial outcome the work actually serves. The customer problem sitting underneath the campaign. The bar of evidence a team requires before it trusts a result enough to act on it. Who gets to decide what, and when. The point where an initiative gets scaled or gets killed.
When a platform rewrites its auction overnight, a team with clear anchors only needs to ask: how should this tactic change based on what's important now? That takes an afternoon, unlike starting from scratch each time, which takes a quarter and risks team exhaustion.
Failure occurs in both ways: teams that see everything as fixed become brittle when pressured, while those that see everything as negotiable lose clarity and can't define their stance. The key is knowing the difference and not blending the two, especially under pressure, which most teams do.
Agencies love a playbook. Three ad sets, five creatives, kill the losers after seven days. It reads like competence on a slide. In practice it's often a liability, because a playbook is only a record of what worked under conditions that have already changed by the time anyone reads it back.
A decision framework survives what a playbook can't, because it hands down reasoning instead of instruction. Not "refresh creative weekly." Instead: refresh when performance decay clears a defined threshold, when that decay traces back to the creative itself rather than audience fatigue or the calendar, and when a replacement has actually earned enough confidence to deserve the slot. A platform can change how creative gets served overnight and the framework still holds, because it never depended on a fixed schedule to begin with.
This reshapes team expertise: the most valuable person isn't just someone who remembers the approved method, but someone who understands its purpose and can create new ones when the old fails. Hiring should follow this shift, not lag a year behind.
None of the above survives if adaptation only lives in one strategist's head. That knowledge walks out the door the day they take another job, and in an industry with this much turnover, it walks out often.
A few habits are worth putting on the calendar rather than leaving to whoever remembers. Reviews of finished tests that ask one consistent question: what did we believe going in, and what do we believe now that we're out the other side. Honest checks on platform changes that ask whether this actually touches an anchor, or only a tactic dressed up to look important. A running list of the assumptions a strategy leans on, checked on a schedule instead of only after something has already broken.
The line that matters here is between reading about a change and actually learning from it. A team can track every product update the day it lands and still fail to adapt, because reading isn't the same as changing what happens next Monday. A lesson that gets discussed once and then filed away was never learned. It was only noticed, briefly, and then forgotten at the pace of everything else that week.
Results move in real time here, which makes urgency the easiest posture to fall into and the hardest to walk back once it's taken hold. A platform announcement, a rough week of numbers, a client forwarding an anxious headline at nine at night: each one arrives asking to be treated like a fire.
Most of them aren't. Borrow a distinction from statistical process control: most week-to-week movement in a dashboard is just the ordinary noise of a system that was never going to sit perfectly flat. Only some of it marks a real break in the pattern, the kind actually worth investigating. A team that can't tell these apart ends up meeting a routine Tuesday dip with the same alarm as a genuine tracking failure, and after enough Tuesdays like that, the alarm stops carrying any information at all. Everyone's just tired.
A leader's real job in that moment is sorting, not projecting confidence into a room. Some changes need action today. Some need a week of quiet watching before anyone touches a setting. Some deserve a small, contained test. A lot of them deserve nothing more than a note and a return to whatever actually pays the bills. Saying that out loud, calmly, while a client's own nerves are pushing the other way, is harder than it sounds and worth more than most agencies credit it for.
Agencies carry a version of this weight that in-house teams don't. Disruption shows up twice: once from the market, and again from the client's reaction to reading about the market before the agency has finished its own read on it.
A client doesn't need to watch the uncertainty resolve in real time. They need to see, clearly, that there's a process for resolving it. That's a smaller thing to deliver than perfect foresight, and it matters more. Trust in this business rarely comes from never being wrong. It comes from a team that visibly knows what to do with being wrong, quickly and without drama.
The agencies that come out ahead this decade won't be the ones that guessed every platform shift correctly. Nobody's luck holds that long. They'll be the ones that meet an unexpected change, hold it up against what already matters, test what needs testing, and keep moving without losing their nerve or their standards in the process.
Call it something closer to the old idea that a disruption, handled against the right structure, leaves a team sharper than it found them. Each shift adds to a framework that gets better with use, instead of adding to a pile of crises that got survived and quietly forgotten.
Performance marketing spends its days optimizing everything outside itself: bids, audiences, creative, budgets down to the cent. The harder work, and the one that actually compounds, is optimizing the team doing all that optimizing. Build the anchors. Build the reasoning underneath the tactics. Build the habit of turning change into knowledge instead of noise. Do that, and the next disruption won't be a crisis to manage. It'll be familiar terrain, seen from whatever angle it takes to read it clearly, before anyone else on the account even notices it moved.
