For more than a decade, I’ve been a fan of using granular cost-to-serve analytics to make better business decisions. Traditional accounting systems provide a distorted lens on the true costs of serving customers. In contrast, Activity Based Costing (ABC) solutions gather granular data on customer profitability based on the true costs of fulfillment for a stock keeping unit (SKU) as it moves through a supply chain. But until recently, the ABC market was served by relatively small niche software vendors. It was not clear to me that their solutions could scale to meet the needs of large companies.
More recently, SAP and Manhattan Associates have come out with solutions in this space. A few months ago, SAP briefed me on its new solution called SAP Net Margin Analysis. And I was recently briefed by Manhattan Associates about its new “Total Cost to Serve” solution (SAP and Manhattan Associates are ARC clients).
Interestingly, neither company is calling their application an Activity Based Costing (ABC) solution, but both have that flavor.
ABC views processes as the primary drivers of costs. Resources, including overhead, are mapped to specific activities. Using ABC, you can generate a granular customer Profit & Loss (P&L) statement based on the activity-based costs for each customer order. Customers do not behave the same way. Some place simple orders with long lead times. Others place complex multi-product and multi-ship orders, then call back to change order quantities, call back again to change the delivery time, and then, all too often, call to cancel the order.
Once costs are mapped to customer activities, it becomes possible to roll those order costs into monthly customer profitability reports. This allows management to understand which customers are profitable, and more importantly, why. What were the customer behaviors that made an account a profit-creating or profit-destroying proposition? When I was briefed by SAP, it seemed to me that their solution was more focused on customer segmentation, and would be more heavily used by sales and marketing folks than the supply chain group.
In contrast, Manhattan Associates seemed more focused on developing a solution that procurement and logistics teams would use. By having granular supply chain costs, buyers can make better decisions on who to buy from. Are the true net margins better if we source from Mexico or China?
The logistics team can also use this data to make better decisions. Should a distributor serve retail customers with direct store delivery or continue to fulfill to the retailer’s DC? Which DC should be used to fulfill a particular order, the one closest to the customer or the more highly automated DC? Should a retailer pay their suppliers to engage in valued added service activities, or would it be cheaper for them to do this at their own DCs? Should goods be delivered floor loaded or on pallets? Should inbound shipments come prepaid or collect? By adding up the costs associated with a SKU on a location by location basis as it moves through the supply chain, these kinds of questions can be answered accurately.
I have a few final thoughts.
First of all, ABC methodologies are great for applying costs to goods as they move through the supply chain. ABC is less effective for a granular understanding of manufacturing costs. For that, throughput costing tools should be used.
Secondly, these tools should be used to make short term, tactical decisions. But companies that have S&OP processes also need to make understanding true customer or channel profitability part of their executive review process.
Thirdly, mapping the costs to the process will take some time. Manhattan Associates argues that a company’s implementation team should contain roughly equal numbers of financial and supply chain members. In particular, if financial analysts dominate the implementation team, they will try to tie all costs back to corporate P&L statement. Manhattan argues, and I agree, that an 80/20 rule should apply. Don’t get too fixated about tying everything perfectly back to the P&L. Going too granular just drives up the implementation cost without improving decision making.
And finally, I am the strongest possible proponent of these types of solutions. I believe any company that sells mainly low margin products needs this type of solution.