Every new client relationship starts with a dashboard. What's in that dashboard tells you more about the agency than any case study they've ever sent you. The metrics an agency chooses to report are a direct signal of what they're willing to be held accountable for.
Most agencies report what's easy to collect and difficult to argue with in isolation. Impressions are real — they happened. Engagement rate is measurable. Follower growth is visible. None of these numbers close a loop to revenue, but all of them go up when content is being produced, which makes them useful for justifying a retainer.
Impressions in isolation tell you that the content existed and was technically visible to a human being at some point. They tell you nothing about whether that person was a qualified buyer, whether they formed an opinion about your brand, or whether they ever came back.
Engagement rate measures the algorithm's decision to amplify content within its own system. A 6% engagement rate on a post that reached 400 people is 24 interactions. That's a Tuesday afternoon on a niche Slack group. Reporting it as a performance metric is technically defensible and strategically meaningless.
Follower growth is perhaps the most durably misleading metric in the stack. It measures an audience that opted into hearing from you — but nothing about that audience's intent, their proximity to purchase, or whether your content is actually serving the people who matter most to the business.
The report that admits "this campaign underperformed, here's exactly why, and here's what we're testing next" builds more trust than any 300% engagement spike we could show you.
We organize reporting into two tiers: revenue-connected metrics that tell us whether the work is actually driving business outcomes, and leading indicators that give us early signals about trajectory.
Revenue-connected metrics:
Leading indicators we trust:
Good reporting has a rhythm that matches the speed at which decisions need to be made. We operate on three distinct cadences:
Weekly pulse: Three to five numbers. Trends only — no explanation unless something changed dramatically. The purpose is to stay calibrated, not to justify the week's work.
Monthly review: Channel-by-channel breakdown, optimization decisions made and the reasoning behind them, performance against benchmark, and one honest section titled "what's not working." This section is non-negotiable.
Quarterly: Full attribution analysis, channel mix review, and strategic recommendations for the next quarter. This is the only cadence at which we discuss whether we're targeting the right channels at all, rather than just optimizing within them.
A campaign underperformed last month. You have two choices: write a report that contextualizes the miss with seasonal factors, competitive noise, and platform algorithm changes — or write a report that names what specifically failed, why you believe it failed, and what you're changing in the next cycle as a direct result.
The first report maintains the relationship in the short term. The second report builds the relationship that survives a bad quarter. Senior clients don't fire agencies for bad results. They fire agencies that hide bad results behind language designed to make the miss sound like a considered strategic pause.
If you want to know whether your current agency is worth keeping, ask them to show you a report from a month where something didn't work. The quality of how they communicated that failure tells you everything about what you can expect when the stakes are higher.