You know the feeling. You just had an amazing onboarding call with your new agency. They asked all the right questions. They wanted to know about your competitors, your margins, your sales cycle, that weird seasonal pattern where leads dry up in August because your buyers are all at the same industry conference. You hung up thinking, these people actually get us.
Fast forward 90 days. You're on a monthly performance call and your strategist suggests running ads against a competitor you explicitly told them isn't a real competitor during onboarding. They mispronounce your product name. They recommend a landing page strategy that you already tried two quarters ago and told them about in week one.
This isn't a bad agency problem. This is an every agency problem. And after ten years of running one, I can tell you exactly why it happens.
The Onboarding Call Is a Performance
I don't mean that in a cynical way. Mostly. The onboarding process at a good agency is genuinely thorough. Someone senior — sometimes the person who sold you, sometimes a strategy lead — runs a 60-90 minute deep dive on your business. They take notes. They ask smart questions. You feel heard.
But here's what happens next. Those notes go into a Google Doc or a Notion page or a shared drive somewhere. The senior person who ran the call hands the account to the person who's actually going to manage it day to day. There's an internal briefing. "Here's the client, here's what they care about, here's the login info." Maybe 20 minutes. Maybe 30 if the manager has follow-up questions.
And then the clock starts ticking on how fast all that context evaporates.
Because the person managing your account has seven other clients. Maybe ten. They're in the weeds on campaign builds, bid adjustments, creative reviews, and reporting deadlines. They looked at your onboarding doc once. They'll look at it again if something goes wrong. Otherwise? It sits in a folder, gathering digital dust.
By month two, they remember the broad strokes. By month three, they're working from what they've observed in the account data, which is useful but incomplete. The nuance from that onboarding call — your founder's strong opinions about brand voice, the reason you pulled out of a specific market last year, the fact that your "competitor" is actually a company you partner with on certain deals — that's gone.
Turnover Makes It Worse. Way Worse.
The average tenure at a digital marketing agency is somewhere around 18 months. Some studies say less. Think about what that means for your account.
You sign with an agency and get assigned a strategist. For the first three months, they're learning your business. Months four through twelve, they're actually performing well — they know your account, they know your preferences, they're making smart decisions based on accumulated context. Then around month fourteen, they start interviewing for their next job. Month eighteen, they're gone.
The agency assigns someone new. And guess what? That 60-minute onboarding call you had a year and a half ago? The new person gets a 15-minute internal handoff. If you're lucky. If the departing strategist didn't just quietly leave everything in their head and walk out the door.
I talked to a prospect last year who'd been with the same agency for four years. In that time, she'd had five different strategists. Five. Each one started by asking her the same questions. What's your target audience? Who are your competitors? What does your sales process look like? By the third time, she stopped giving detailed answers. "Just read the notes from last time," she told them. They couldn't find the notes.
That story isn't unusual. I hear some version of it on almost every sales call.
There's No System. There's Just People.
Look, I've been inside agencies of all sizes. From five-person shops to global holding company networks. And the dirty secret is that almost none of them have a real system for preserving client knowledge over time.
They have project management tools. They have reporting dashboards. They have CRMs that track the sales process in granular detail. But actual client intelligence — the accumulated understanding of a client's business, market, history, preferences, failures, and lessons learned? That lives in individual brains. Maybe in scattered Slack messages. Maybe in the "notes" field of a task that nobody reads.
I'll give you a specific example. We took over a paid social account from a large agency. During our audit, we noticed they'd been running the exact same audience targeting for 18 months. When we asked the client about it, she said: "Oh, the first strategist tested a bunch of audiences and found those worked best. But she left, and nobody after her knew about the testing, so they just kept running what was there."
The original strategist had done great work. She'd tested 12 different audience segments, documented which ones converted and why, and built a targeting strategy based on real data. But that documentation was in her personal notes. When she left, the agency didn't transfer it. The next three strategists just saw the existing audiences and assumed that's how things had always been. Nobody questioned it. Nobody improved on it. Eighteen months of potential optimization, frozen in time.
This is what happens when knowledge depends on people instead of systems. People leave. People forget. People get busy with their other nine accounts and stop investing mental energy in the details of yours.
The "Discovery Call to Handoff" Pipeline
There's a structural problem in how agencies are organized that makes this almost inevitable. The person who sells you on the agency is not the person who does the work. Sometimes they're not even in the same department.
The sales team's job is to close the deal. They're really good at asking questions, building rapport, and making you feel understood. But their incentive is to move on to the next prospect once you've signed. The "handoff" to the execution team is usually a one-page brief and a quick call.
I watched this happen from the inside at a previous agency. The business development director would have these incredible discovery calls. I mean, the prospect would share their entire business strategy, their growth plans, their pain points, their budget flexibility. Gold. And then the BD guy would write a two-paragraph summary in Salesforce and tag the account manager. Two paragraphs to capture a 90-minute conversation.
The account manager would read those two paragraphs, build a campaign plan based on best practices for the industry, and get to work. Ninety percent of what the client shared on that discovery call never made it into the actual strategy. Not because anyone was lazy. Because the system didn't have a place for that level of detail.
Knowledge Should Compound. Instead It Resets.
Here's what drives me absolutely crazy about this problem. Every month that an agency works with a client, they learn something. Which ad copy resonates. Which audiences convert. What time of year performance dips. What the client's board cares about. What their competitors are doing that's working.
That knowledge should be additive. Month six should build on everything learned in months one through five. Year two should be dramatically more effective than year one because you have a full year of accumulated intelligence informing every decision.
But that's not what happens. What happens is: strategist leaves, knowledge resets. New strategist spends three months getting up to speed, performs well for six months, then leaves. Knowledge resets again. The client is essentially paying for the same learning curve over and over.
I ran the numbers once for a prospect. She'd been spending $25K/month on PPC for three years with an agency that had turned over her strategist four times. By my estimate, she'd "paid for" the initial learning curve at least three times over. That's roughly $50,000-$75,000 in suboptimal performance during transitions — money that should have been driving results instead of funding someone's on-the-job training.
And this doesn't even account for the ideas that were lost. The strategist who left in month 14 had a theory about a new campaign structure that she never got to test. The one who replaced her didn't know about it. It was gone. How much was that untested idea worth? We'll never know.
Why Nobody Fixes This
If this problem is so obvious — and it is, everyone in the industry knows about it — why hasn't it been solved?
A few reasons. First, building systems to capture and preserve knowledge is hard. It requires discipline, time, and tools that most agencies don't invest in because the ROI is long-term and agencies tend to think in monthly billing cycles.
Second, it's not in the agency's short-term interest. If knowledge resets every time a strategist leaves, the client becomes more dependent on the agency as an institution, not any individual. The client can't easily leave because they'd lose whatever context the current person has. It creates a kind of accidental lock-in that agencies aren't exactly motivated to solve.
Third — and I think this is the real reason — most agencies don't think of client knowledge as an asset. They think of people as the asset. Hire good people, put them on accounts, let them do their thing. Which works great until those people leave. And they always leave.
What Should Knowledge Actually Look Like?
I've spent years thinking about this. What would it actually look like if an agency treated client knowledge like the critical asset it is?
It would mean every insight gets captured somewhere persistent and accessible — not in someone's head, not in a Slack DM. Every test result. Every client preference. Every competitive observation. Every piece of feedback from a sales call about lead quality. All of it, organized in a way that any team member can access and build on.
It would mean that when a new strategist takes over an account, they don't get a 15-minute handoff. They get the full accumulated knowledge of every person who's ever worked on that account. The learning curve goes from 90 days to 90 minutes.
It would mean that knowledge compounds instead of resets. Month 18 is built on everything from months 1 through 17. Year three is genuinely smarter than year one, not just a repeat of the same learning.
It would mean the client owns that intelligence. It doesn't walk out the door when someone quits. It doesn't disappear when you switch agencies. It's yours.
I don't think this is a fantasy. I think it's a basic operational discipline that the agency industry has just... never prioritized. Because it's hard, and it's easier to just hire another strategist and hope they stick around.
But the agencies and teams that figure this out — that build systems where intelligence about a client's business only grows and never gets lost — those are going to have an unfair advantage. Because while everyone else is resetting the clock every 18 months, they'll be compounding.
And compounding always wins.