Voodoo Benchmarking: Overhead Rate Benchmarks

Benchmarks for overhead rates are part of  "Voodoo Benchmarking," a set of business practices preached by Procurement Departments and their consultants, believed in by those who have a such a need  -- but invalid from any scientific, mathematical, accounting or logical basis. Voodoo Benchmarking is so well-established in business that even smart people accept its existence and quibble only about the details. It's part of the dumbing down of business decision-making, liberating executives and staff people from having to think. Instead, they compare ratios to "benchmarks" and make their decisions accordingly.  

Overhead rate benchmarking is one example (among many) of Voodoo Benchmarking. Overhead rates are used by many advertisers in the calculation of agency fees. In many cases, the overhead rates of a client’s agencies are compared to “benchmarked overhead rates” as provided by benchmarking consultants. Differences between actual overhead rates and benchmarked overhead rates lead to much hand-wringing and tough negotiations. Fees are driven down this way. Both parties have, over the past decade, grown used to this drama and accepted it as a normal way that the annual fee-setting game is played. 

There are understandable arguments between the parties over the way that the actual overhead rates are calculated and the way that the benchmarked rates have been determined. What is less understandable is why either party thinks that there is any science in this. The simple fact is that there is no such thing as a benchmarked overhead rate. It is a pure invention based on bad assumptions and questionable mathematics. It’s made a lot of money for the benchmarking consultants and cost the agencies a lot in the way of fees. It has also misled CMOs and Procurement Executives into thinking that benchmarked overhead rates represent managerial sophistication on their part and give them transparency into agency operations.  

Even the ANA has given its tacit endorsement of overhead rate benchmarking by giving benchmarking consultants a prominent place in Webinars, committee meetings and speakers' platforms at conferences. 

The overhead rate mathematical logic goes like this: all advertising agency costs are assumed to fall into one of two buckets: the Professional Services Cost bucket or the Overhead Cost bucket. Professional Services costs are the salary and benefits costs of the agency’s client-facing people. Overhead costs are everything else. The two buckets account for 100% of the agency’s costs. (We’ll ignore, for the moment, the different ways that agencies may decide into which bucket to put different types of costs, like benefit costs or the salary costs of unassigned people. That’s a separate and distracting subject.)

The overhead rate is simply Overhead Costs divided by Professional Services Costs. It is a number, like 100% (the two buckets are equal in size), or 125% (overhead costs are greater than professional services costs), or some other number expressed as a percentage. The overhead rate is “overhead costs per dollar of professional services costs.”

So far, so good.

Benchmarking consultants will tell you that there are standard proportions of these costs in the marketplace, and that these proportions (“rates”) should be used to evaluate an agency’s overhead costs. The benchmarkers have the industry data and proof, they say, but their database cannot be revealed for confidentiality reasons.

Because an overhead rate involves a numerator and a denominator, it is a funny animal. A “high” overhead rate could mean either “high” overhead costs or “low” professional services costs. Which one is it? In the negotiation between client and agency, the foregone conclusion is that a high rate means high overhead costs. Period. This is one-sided and wrong.

Overhead costs are relatively fixed costs for a given scale of agency, and professional services costs are variable costs – they vary with the amount and complexity of agency work.

These costs do not vary naturally with one another. Overhead costs are driven by their own logic: property costs, physical location, square footage per person, level of training costs, amount of unreimbursed client travel, amount of recruiting expenses, pitching costs, etc. Overhead costs reflect, to a considerable degree, the cultural legacy of an agency and the choices it has made in the way it wants to do business. Big global agencies tend to have more relative overhead than small local agencies. Some of this is a matter of choice and heritage.

Unquestionably, there are high overhead agencies and low overhead agencies, just as there are young men and old men, short women and tall women – there is a variety of experience in the real world, and the different levels of overhead costs are just another dimension by which agencies differ from one another. Benchmarked overhead costs are like average height or average age. If you happen to be taller or older – you are expected to change!

Professional services costs are another matter indeed. They should vary in some way with agency workload and the choices that agencies make about how to staff their agency and the kind of work they do. First, there are decisions about the number of Client Service, Planning, Creative and Production people to assign to the variety of briefs carried out by an agency. Second, there are the decisions about seniority and cost for the various people. Different agencies go about these headcount and seniority choices in different ways. They make these choices relative to workload and tradition, not to overhead costs.

We’ve been looking at this for some time in our consulting business. One thing is clear: agency workloads have been growing, while agency fees have been under downwards pressure. Consequently, price for agency services (fees divided by workload) have declined. We calculate that the decline has averaged 4.5% per year for the past 20 years, under the relentless negotiation success of Procurement.

Agencies have reduced their Professional Services costs in response to fee declines. Agencies have also tried to reduce overhead costs, but this is not as easy to do and has not been as successful. Variable Professional Services costs have declined faster than fixed Overhead costs. This has affected overhead rate calculations over time. Agencies have become more fixed-cost intensive, and this has tended to increase average overhead rates.

The number of ads created and produced per head by agencies has increased steadily during the past several decades. David Ogilvy wrote in 1983 that the average copywriter completed 3 ads per year. Today, the completion rate per equivalent ad is more than double that amount, according to our data. Agency productivity has doubled since Ogilvy’s day. This takes into account the changing mix of media. Output per head on an equivalent ad basis has risen.

Logically, then, the agency’s Professional Services costs are becoming more efficient. In any given period of time, overhead costs may stay constant, but professional services costs will decline relative to workload. In terms of the overhead rate, this means that overhead rates must rise.

The benchmarking consultants claim that benchmarked overhead rates are in decline. Their declining benchmarked rates are then used by clients to force down agency fees.

When fees decline, agencies downsize. The downsizing comes from reductions in Professional Services costs. As a result, the overhead rate for subsequent years goes up.

This is not a theoretical critique. Our work shows that the long-term trend is for agency overhead rates to rise with increases in agency productivity. Advertisers, though, believe what they are told -- that benchmarked rates are going down and so should agency fees, even though workloads are on the increase. Well, if fees go down, overhead rates will rise. That’s the real mathematical logic of the industry, in stark contrast to industry practice or the claims of the benchmarking consultants.  The problem is not only that Voodoo Benchmarking is invalid -- the results from this Voodoo Benchmarking are false.

We don't dispute that there are overhead rates in the advertising industry.  What we object to is the belief that these rates can be benchmarked to a standard. 

Advertisers beware! Agencies beware! There are overhead costs out there, and professional service costs as well, but their relationship to one another is pretty random -- there are no industry overhead rate benchmarks, any more than there is gold on the alchemist's worktable. The sooner the parties realize this the sooner they will seek a more rational way of dealing with costs, workloads and fees. One place to start would be to include workload metrics and agency productivities as part of their dialogue.

This change in approach can’t come fast enough, because Voodoo Benchmarking and the myth of overhead benchmarking are causing manslaughter on Madison Avenue, as agencies downsize with their reduced fees. Sacked men and women don't write great ads, nor can the enfeebled agencies they leave behind.