Picture an agency leader whose firm grew 9% last year. He is proud of it, and he brings it up in every meeting, mentions it to every carrier rep, works it into every recruiting pitch. On paper he has earned the victory lap. The number clears the benchmarks MarshBerry, Reagan, and Sica Fletcher publish and puts his firm in the top league.

Walk to the other side of the agency and the picture changes. New business is flat. Retention is slipping. Account managers keep churning out the door. The 9% is real, but it is being carried by rate, not by anything the firm is building. The headline says high achiever. Underneath, the foundation is crumbling. And the KPI report never flagged it, because he read it one number at a time.

That gap is the whole problem. A single metric is a headline with the story cut off. Ask what actually drove the 9% and you find three different businesses: new business from producers actually selling, rate where the same accounts renewed higher and the commission came along for the ride, or an acquisition being counted as organic growth. All three print the same 9%. One compounds, one runs out, and one you already paid for.

The story is in the bottom-up build of each KPI, its component parts, and how the metrics move against each other.

Most agencies fail at this twice. First, they read each KPI on its own instead of together, so the numbers never add up to a story. Second, even when the dashboard is clearly saying something, nobody acts on it. A polished report lands every month, gets a few minutes of attention in a meeting, and goes from dashboard to dumpster before lunch. Next month a fresh one arrives and the cycle repeats. This series is about doing the opposite: reading these KPIs together so they tell you the truth about the business, and then actually acting on what they say.

Most brokerages track somewhere between twenty and forty KPIs across four families:

Business metrics describe the book: organic growth, revenue retention, revenue per employee, EBITDA margin, book mix by line.

Producer metrics describe who sells and services it: new business written, book size per producer, producer age, pipeline coverage, validation rate.

Client metrics describe who buys: client retention, average account size, policies per client, account concentration, client tenure.

Carrier metrics describe who underwrites it: premium placed per carrier, loss ratio, contingent and supplemental income, carrier concentration.

Each family answers one question. Put them side by side and they answer the question that actually matters, which is whether this business is growing for reasons that will repeat.

And none of it reads off a single snapshot. One month, or one quarter, tells you almost nothing. The same KPIs tracked month after month and year after year are what actually show the health and longevity of a business, because the signal is in the trend, not the reading.

Here is the logic that connects them, before we spend a part on each.

Growth has a source, and the source decides both the multiple and your next move. Organic growth is the headline. New business, lost business, and price/exposure change are the sources. A firm growing on new business from a deep producer bench is worth more than a firm growing on rate it didn’t earn and can’t keep, even when the growth number is identical. You cannot tell those two apart from organic growth alone. You need the bridge underneath it. And the bridge is not only for a buyer pricing the firm. It tells the operator where to act: if the growth is coming from rate, the new-business engine is stalling and needs attention now, not at renewal.

Every business metric traces back to a person. Retention isn’t a number the firm produces. It’s the sum of what each producer and account manager holds onto. New business isn’t a target. It’s what specific people on a specific bench actually wrote. When you see retention slipping, the producer KPIs tell you whether it’s broad or whether one aging book is bleeding out. The business layer tells you what happened. The producer layer tells you who.

Every producer metric traces back to a client base. A producer’s book is only as durable as the clients in it. Client concentration, average account size, policies per client, and cross-sell depth decide whether a producer’s revenue is a fortress or a house of cards. A top producer whose book is three large accounts in one industry is a different risk than a top producer with two hundred rounded-out mid-market relationships, even at the same revenue.

And every client sits on top of a carrier. The broker doesn’t manufacture the product. Carriers and wholesalers do. Carrier concentration, wholesale dependency, and the premium expiring each month decide how much market leverage the firm has, how exposed it is if an appetite shifts, and how lumpy the renewal workload gets. A great client book placed entirely through one carrier or one wholesaler is a great book with a single point of failure.

Read top to bottom, the four families form a chain: carriers make the product, clients buy it, producers sell and keep it, and the business metrics total it all up. A weakness anywhere shows up two layers later as a number someone misreads as a sales problem or a market problem when it’s really a concentration problem or a succession problem.

That’s why the diagnostic question is never “what is this KPI.” It’s “what does this KPI look like next to its neighbors.” A retention rate of 92% is good or alarming depending on what new business, producer age, and client concentration are doing around it. The same number means different things in different company.

The rest of this series takes each family in turn. For every KPI I’ll cover what it actually tells you, the metrics it has to be read against, and then the part everyone is circling right now: what happens to it when AI moves from the pitch deck into the workflow. Some of these KPIs get more important under AI. Some stop measuring what they used to measure. A few may stop being worth tracking at all.

The takeaway

A brokerage’s KPIs are a system, not a list. Growth has a source, the source traces to a producer, the producer traces to a client base, and the client base sits on a carrier. Read any one number alone and you will misdiagnose the business. Read them as a chain and the real story, good or bad, is hard to hide. That system is also what AI is about to reshape, which is where this series goes next.