I should say up front that I’m generally a huge supporter of AI tools and enabling people to be more effective and efficient at various aspects of their jobs. I’ve even hacked together small tools with Claude Code that have made a huge difference to my work and personal life.
But… I am a little worried that agencies are diving head first into coding up fancy tools and not remembering the basics of what makes their business actually function.
Let me explain.
Last month, I sat in on a management meeting where the founder walked through their new reporting framework. The framework had six sections that covered everything from new business pipeline to client retention metrics, through to utilisation and employee satisfaction scores.
It was shiny and very cool. The monthly management meeting now had clear agenda items to work through and all of the data was anchored in this new framework. Everyone felt like they were finally running a “proper” agency.
Then a five figure per month client gave notice. And the management team realised something: they could tell me their utilisation rate down to the decimal point, but they had no idea if that client’s organic traffic had actually improved in the last quarter.
The sophistication paradox
Agencies are getting better at measuring themselves and worse at measuring client outcomes. To be fair, marketing attribution is a mess right now and has been for a while, so it’s not an easy challenge to solve for most channels.
I see this pattern constantly. Founders implement proper business systems (EOS, dashboards, new frameworks) and it creates an illusion of control. You can answer questions about your business that you couldn’t answer six months ago. What’s your win rate on pitches? What percentage of revenue comes from your top three clients? How many days until your pipeline converts to cash?
All valuable questions. But here’s what’s missing: can you quickly answer “did we move the needle for Client X this month?”
An agency I’m working with recently reviewed their accounts. The founder could tell me the following:
- They had 78% utilisation
- There were three clients in the renewal pipeline
- They had two upsell opportunities worth low four figures per month.
But when we looked at their biggest account, they couldn’t tell me whether that client’s key metrics had improved quarter-on-quarter. They were tracking hours logged and tasks completed. Not results delivered. The team said that the client seemed happy, but how did they know?
This isn’t because founders are lazy or don’t care. It’s that internal metrics are genuinely easier to track than client outcomes. Your utilisation rate lives in your project management system or time tracking tool of choice.
Or pulled into a shiny new vibe coded tool.
Client outcome data is scattered across Google Analytics, their CRM, maybe their e-commerce platform. It requires access permissions, cross-platform reporting, attribution complexity.
Internal metrics also feel more controllable. If utilisation drops, you know what to do: adjust resource allocation, pause hiring, push harder on sales. If a client’s traffic drops, the action is murkier. Is it seasonal? Algorithm changes? Their pricing? Your strategy? Did they not do that thing you told them to do six months ago?
So we drift toward what we can measure and control. And we end up with sophisticated dashboards that tell us everything about our business except whether clients are actually getting better results.
When teams think activity equals results
If you’re not tracking outcomes, your team definitely isn’t.
Most agencies I speak with don’t even track top line retention, let alone the signals that feed into this such as results, happiness and engagement.
One agency I advise recently discovered a client was unhappy despite their account manager insisting everything was fine. The account manager joined a call and the client mentioned break clauses for the first time. Red flags suddenly went up.
The agency was measuring activity (hours logged, status updates sent, internal meetings held and even delivery of work) but not outcomes. Without quarterly outcome tracking, all they had was subjective “client satisfaction” versus objective “results delivered.”
Yes, the client was satisfied with the team and the quality of the work delivered. But the needle hadn’t moved on results yet and the lack of communication about this led to the question about break clauses.
This is where the “we tried our best” problem emerges. When a client churns and you ask the account team what happened, they’ll point to effort: we sent weekly updates, we responded to emails within 24 hours, we had monthly strategy calls. All true. But none of it answers: were their three key metrics better this quarter than last quarter?
I’ve seen senior team members actively shy away from tracking KPIs and setting targets because they’re worried about not hitting them or being held to something unrealistic. But this is actually making the situation worse.
Teams optimise for what’s measured. If you measure tasks completed, they’ll complete tasks. If you measure client results, they’ll focus on client results.
Well, you may need to work a bit harder than that, but you hopefully get the point!
What to actually track
You don’t need perfect attribution. You need monthly (or even quarterly) momentum indicators. You need leading indicators that things are moving in the right direction.
The basic question for each client: are the three metrics they care about better this quarter than last quarter?
Not “did we cause it” – that’s the attribution rabbit hole. Just – is it happening?
This should be a simple monthly or quarterly scorecard per client. Traffic up or down compared to last quarter. Conversion rate up or down. Revenue from organic up or down. Three to five metrics maximum, each with a simple direction indicator.
I’m not suggesting that you stop tracking internal metrics. Utilisation matters. Win rates matter. Pipeline value matters. But these should sit alongside client outcome metrics, not replace them.
And the point isn’t to prove causation. The point is to have a shared view of whether things are moving in the right direction.
Here’s what changes when you do this – suddenly you can see which accounts are “busy but not effective.” The client where you’re hitting 85% utilisation but their metrics haven’t moved in six months. That’s the account that’s going to churn, even if the relationship feels fine today.
It also highlights your stars. The account manager whose three clients are all showing improved results quarter-over-quarter becomes obviously more valuable than the one whose clients are flat.
The management meeting question to be asking
Instead of “What did we deliver this month?” ask “Which clients got measurably better results this quarter versus last?”
This immediately surfaces different conversations.
At first, this will be awkward and cause some tension. A lot of teams don’t like being held accountable to this level of detail. But this isn’t about them, it’s about the businesses that are paying you money to do a job for them and to get results.
The expectation isn’t that every number goes up every month or quarter like clockwork. The expectation is to know what’s going on with your clients so that you can take action if needed.
This doesn’t replace discussing utilisation or pipeline or team issues. It supplements them. And it shifts the frame from “are we busy?” to “are we effective?”
You might discover that your “problem” account with low utilisation is actually your best client outcome story. They’re efficient because the strategy is working, not because you’re under-servicing them.
You might also discover that your “profitable” high-utilisation account hasn’t moved the needle in six months. You’re busy, the client is paying, everyone assumes it’s fine. But without outcome tracking, you miss the early warning sign that they’re quietly shopping around.
The gap between knowing and acting
Here’s the uncomfortable bit – implementing this is genuinely harder than it sounds. I’m in the midst of it with two clients right now!
Building scorecards for a bunch of clients is admin work that feels less urgent than client delivery. Training account managers to have “results” conversations rather than “activity” conversations requires coaching.
And there’s a fear underneath all of this, which is: what if we track outcomes and they’re not improving?
That’s exactly why you should track them. Because if results aren’t improving, the client already knows. They’re not waiting for you to tell them. They’re either living with it silently or they’re already talking to other agencies.
The vibe coded dashboard that tells you everything about your business except client outcomes gives you false confidence. You feel in control because you have data. But you’re measuring your busyness, not your effectiveness.
The agencies I work with who track both internal metrics and client outcomes don’t necessarily have better results across every client. But they know where the problems are. And they can act on them before the client reaches for the break clause.
You’ve built sophisticated reporting systems because you want to run your agency professionally. That’s good. Just make sure one of those measures is labelled “are clients actually getting better results?” If it’s not there, your dashboard is telling you everything except what matters most.