Realizing the Potential of Activity Based Costing (ABC)

I was recently delighted, but surprised, to be booked to do a workshop at a fall ‘08, distribution association convention on the topic of “Activity Based Costing” (ABC). ABC is a cost modeling methodology for getting a better sense of the true profits from – customers, services/products, suppliers and other business decisions – that was first ‘productized” in 1987. Back then, it got a lot of “this-will-revolutionize-business” hype, but the adoption rate has generally been slow across all industries. Some industries – banking, casinos, large logistics firms – are heavy users, but activity within independent distribution channels is maybe about 10% or less.

Although I have been personally involved in some terrific, distribution-company turnarounds that were both guided and energized by ABC ranking reports (starting in 1976) for profitability of – customers, suppliers, branches and employees – most distributors have not tapped the full potential of ABC analysis or even philosophical thinking which raises at least these questions:

Why the low, slow adoption rate for ABC by distributors over the past 21 years?1

Why the weak follow through at firms that do get first or second-hand ABC insights?

How can we make ABC a more user-friendly and less threatening tool with simple starts?

What would be the first, easiest distribution-specific, ABC applications to try?

Is my recent booking a sign of broader interest within distribution channels?

And, from the non-user crowd: “First, what exactly is ABC anyway?”

ABC, simply put, is an accounting method (not software, although there are ABC software tools available) for assigning a firm’s “resource costs” for all activities that go into producing products and/or services by a more detailed splitting up of and re-assignment of indirect or overhead costs.2 (Readers can google ‘activity based costing” + “six steps” and find all that you need.)

By simple example, if two customers of a distributor each bought $1000 of goods with $200 of margin dollars in the order, how much profit did each order put to the bottom line? It depends, of course, on the number and degree of service cost activities. What if one order involved every (extra) service cost imaginable while the second was generated by a non-commissioned, house account which was ordered electronically for standard, in-stock, fast-turning goods and then picked up by the customer who paid in cash? We know intuitively that the second order is more profitable, and we are grateful to have it to cross-subsidize some of the losing orders that we know we have everyday. Our financial reporting systems have been shaped to pay our taxes on time and do asset-backed borrowing from banks; they don’t tell us which suppliers, products and customer combinations make us the most money, which really defines our competitive strategy.

To extend the two order problem to customer profitability analysis for an entire year’s worth of business, it is not uncommon for an ABC analysis to rank customers from the most profitable to the biggest losers to reveal the following magnitudes:

  • The top 10% of the most profitable customers yield 90-100% of company profits (The top 5 accounts are sensationally profitable; should we do anything extra not to lose them? Or, to sell even more to the core more effectively?)
  • The top 40% cumulatively yield 140-150%, which, let’s say, is the breakeven point for the ranked customers.
  • The bottom 60% start as negligible losers to finish with a few, shockingly-big losers at the bottom. The cumulative losses on the bottom 60% offset the extra profits from the top 40%, so that 100% of the accounts do total to 100% of the profit before interest and tax (PBIT).
  • But, the bottom 1-2% of the customers will, notably and amazingly, lose as much as 20% of PBIT (Should we try to turn these “lead” accounts into profitable “gold” accounts? How?)

What are the barriers to using ABC? Are there pitfalls? As we read about ABC on the internet, it becomes quickly apparent why there is low usage of ABC by firms that are small with less moving parts, owner-operated, sales-driven and without good accounting expertise on staff:

  • A would-be practitioner must learn new: vocabulary, building block concepts and spreadsheet ability before rethinking the overhead costs assigned to product or service process steps that thread their way through the business.
  • Mapping a process and interviewing everyone involved in that process to get time allocation assumptions to compute activity costs for an ABC model is tedious, although it can often reveal new process improvement ideas.3
  • The level of detail that goes into the model is a strategic-trade-off choice that must balance the cost of getting and maintaining new information with the immediate action value. Too much detail makes the model unwieldy and difficult to use on an on-going basis; too few cost-assumptions can make the model’s estimations of profitability less accurate. Because no model can simulate the dynamic reality of a business perfectly; where is the optimum balance? And, because no two similar distributors will make all of the same assumptions and detail choices even when using the same “best practice’ distributor ABC software tools, all models and results will vary to some degree.
  • It’s easy to get caught up in “analysis paralysis”, especially if we don’t know what our intended, disciplined strategy is. ABC analysis can help us optimize the past, but it does not directly reveal why the most profitable customers became that way or how others could too.
  • Any remarkable ABC conclusions will then be a direct challenge to the (unspoken) cultural rules that have been running the company and helped shape and deliver the current results. ABC will typically reveal huge, heretofore, hidden, internal cross-subsidies between: customers, SKUs, suppliers, employees, truck routes, sales territories, branches, etc. Expect big push back from the creators and leaders of the current system as well as any team members who are suppose to change themselves before they then try to change customer behavior, i.e., most of the management team and outside salesforce.

If a distributor principal decided to try an ABC experiment, what should be the first, easiest and most fruitful effort? It depends, I think, where the company’s industry and customer base are in a life-cycle sense. If we were selling mobile phones in 2000 and had exclusive product and service-contract franchises for a geographic area which we could maintain as long as we pleased the supplier with great annual growth rates, then we should apply ABC to our products to determine which ones we should sell, preferentially to (frankly) new-demand, first-time, somewhat ignorant customers. If we are selling experienced, consolidating customers who are buying commodity products, which every distributor has, for the zillionth time and wants to buy these products at the “lowest total procurement cost” (however they personally define that), then we should start with customer profitability ranking reports.

The choice of customer profitability ranking reports is supported by several more assumptions:

  1. If our simple, average, product mark-up practices have (hidden?) cross-subsidies built into them, then so might our competitors. This will create a pattern like we see in fast food restaurants where they breakeven or loss-leader-promote entrees to then make more than 100% of their profits on the drinks, fries and service items. What then becomes important is total margin dollars generated per customer transaction/visit and best retention rates for the best type(s) of customers. If a restaurant who tried to make the same true profits on all of their products – as a result of ABC studies, it would probably not do well.
  2. If the product lines we stock and sell are not exclusive and/or there are equally excellent alternative lines for all competitors, then we have no competitive advantage from the tangible products we sell, our profit power has now shifted to selling the best, one-stop-shop assortment of items (in-stock, fill-rates are THE foundational service factor) to the right, best, most-profitable customers within a given target niche.
  3. A simple, initial-model for customer profitability can be created without much work.4
  4. Once we can rank all customers by an estimated PBIT contribution, we can then manage the few (5-10) most profitable and unprofitable accounts with more manual analysis. If we then want to do more refined ABC model building to investigate greater numbers of customers in the middle of the ranking report, we can. Start simple. Learn, act and then do a next level ABC model analysis. Don’t get caught up initially in analysis paralysis and arguments about all of the cost assumptions when people don’t like the results of the study.

If necessity is the mother of invention, then today’s tough economy might be the spur that we need to try and look at our business in a radically new way through ABC profit reports. In the spirit of testing cheaply, it isn’t difficult or expensive to check out the support materials in this article’s footnotes and try to run a first, simple model for customer profitability. If any readers already have tried or subsequently create ABC modeling, then I would be glad to hear about your case studies. Please don’t hesitate to contact me via email at Otherwise, may remarkable ABC discoveries come your way!

D. Bruce Merrifield, Jr.

©Merrifield Consulting Group, Inc., Article 2.27

May 14, 2008


  1. Or, 29 years since the publishing of the Harvard Business School case entitled: “Paper Distributors, Inc.(E): The Small Order Problem”. The case was on one of my turnarounds and is now out of print.
  2. Acorn Systems ( is the #1 ABC software company. They have a number of very large distribution firm clients ($200MM and up). They hope to re-package their distributor-best-practices into a more affordable solution for smaller firms in the near future. I have helped a number of distribution software firms install a simple “customer profitability ranking report” option as described in footnote #3.
  3. For more on “service process re-engineering” go to and check “exhibit” #4
  4. Go to check out “slide show #10, slide #5; as well as “articles” 2.15 and 2.19