Managerial Accounting for Product Management

Ever since I studied accounting in college I’ve been fascinated by problems related to measurement & quantification of the world. Ultimately, accounting is the practice of quantifying activity (often complex and messy) that takes place in the real world, and boiling that activity down to dollars & cents in a standardized format.

As a product manager, I often found myself doing “managerial accounting” exercises in the form of “costing” and “profitability measurement” during sprint planning + annual planning processes. And as a growth product manager on a subscription product there was an entire discipline around “growth accounting”, looking at measuring cohorts of users and quantifying customer LTV & CAC.

However, time and time again, I saw these analyses and decision making processes break down across a number of different failure modes, and I wanted to think about how concepts from the managerial accounting toolkit could help.

So, I started having a series of conversations with my former college professor and thesis advisor Sajay Samuel about Product Management and how we might be able to apply concepts from managerial accounting. These conversations were extremely fun, and took us down a winding road of concepts from the philosophy of measurement, to systems theory, industrial engineering, and risk management to name a few. I also provided Sajay with an overview of the product development lifecycle. You can see a mind map of the topics we covered through our conversation:

There were a lot of different areas that we were interested in thinking about:

• Costing: Resources consumed are engineering head count/development weeks, cost objects typically thought as the software features built, but could also be the “outcomes” delivered (e.g. customer reviews solicited). Different costing strategies used depending on the planning horizon/frequency (annual planning vs sprint planning). Lots of good examples about challenges in allocating indivisible capacities (e.g. tech debt, operational excellence, shared infra, analytics, on-call, long term tech Capital project budgeting). Factoring short & long term opportunity costs into costing, differential analysis (opportunity cost is the primary cost in decisions for PMs, and relies on accurate estimates of cost/impact among alternative options, which are almost always inaccurate).
• Profitability measurement: Calculating “profitability” of products when "revenues" come in non-financial measures (these often vary across products within the same portfolio), and need to be allocated across many products/features. There are some other abstract ideas about “product accounting” I’ve been thinking about dealing with how to prioritize resources within product functionality itself (solicitation attempts to maximize reviews for example).
• Performance management/Control: Tracking head count and dev estimate variance (as well as performance variance) could be a mechanism to improve, it would be great for dev estimates. Agency problems exist, mainly at the annual planning level incentivizing managers to engage in “empire building”, but also on a more micro level in “away team” development. Many traditional MA tools are already being used across the board for managing “direct” server costs/budgets/etc, but not for indirect costs or engineering resource allocation.

However, we decided it would be best to pick one very specific problem/use case to investigate and go deep on that one thing. So we looked at different problems across the product lifecycle:

Many of the problems across both long and short term planning horizons come down to failures in accurately assigning costs and evaluating “profitability” of various investment decisions. We ended up deciding to think about how projects that are indirect drivers of business results (e.g. tech debt, analytics, infrastructure/devops, testing, etc) are often chronically under-prioritized. You can read more about how these problems manifest here.