Data driven management is a popular concept and with good reason. CEOs, management teams, and board of directors gain tremendous value in defining and managing to key performance indicators (KPIs).
Why?
KPIs provide insight into the underlying mechanics of the business model that help teams manage departmental functions and focus the company on the building blocks of future success. The concept of KPIs as powerful intra-period (day, week, month, quarter) barometers of company performance is well understood.
However, traditional mechanisms for financial management, GAAP financials and the operating plan, are necessary, but not sufficient tools for data driven management.
Start-ups are exercises in prospective thinking and both GAAP financials and the operating plan generally fail to provide insight into the upstream metrics that drive financial success.
GAAP financial statements are backward looking records of the past, and they often fail to provide insights into how the company will perform in the future. Start-ups all develop operating plans, however, in my experience 3 year plans, while necessary, are often highly abstracted summaries of a hoped for future that tend to be more academic than operational. They tend to be the product of the CFO's office with little day to day value for the team.
How then should one bridge the gap between GAAP financials and the high level financial projections? As a CEO, VP, director, or individual contributor what data points matter to you? When you come in in the morning, how do you know what to focus on with some sense of certainty that your particular KPI is a key part of the broader company's goals?
In 1919, DuPont's F. Donaldson Brown was tasked with turning around GM after DuPont bought a 23% stake. In order to help drive clarity and transparency into the state of GM's finances, Brown developed a model that broke down the company's ultimate goal, high return on assets, into an easy to visualize set of critical building blocks.
The standard DuPont model follows:
Return on Equity = Net Profit Margin * Total Asset Turnover * Equity Multiplier
Each component can then be broken down into its constituent parts.
For example,
NPM = Net Income/Net Sales
TAT = Net Sales/Total Assets
EM = Total Assets/Common Equity
One can now see that higher profitability, higher asset utilization, and higher debt levels can all lead to higher ROE. Further, each child node can be further analyzed to understand the key levers that drive the parent node.
Net Profit Margin can be influenced by unit volume, unit price, fixed costs, variable costs, and so on.
Now, how does this apply to start-ups?
Well, start-ups can develop a custom version of the DuPont model that 1) transparently states the formula for value creation and 2) makes visible key value-creating levers that are themselves the "Is" in KPIs.
For example, the search business can be defined by the following formula:
Revenue is no longer an abstract concept but a goal with clearly defined indicators that departments can execute against, such as driving more queries per user, maximizing ad attach rates per query, optimizing click through rates, driving higher ecpms....
A team's ability to develop a relevant DuPont model that maps out the key components of revenue and profit is critical in developing material KPIs. To bridge the gap between Quickbooks' financials and Excel operating plans, I suggest that a team whiteboard a model that drives both revenue and costs.
As a planning exercise:
- develop DuPont formulas relevant to the business that define the material components of revenue, costs, and profit
- drill down on each node until there is no marginal benefit of further granularity
- analyze the impact of moving each indicator, or formula argument, on the desired result and identify the most impactful indicators to manage
- assign each department indicators, arguments in the equation, they can control, effect, manage, and report against
- the formula's arguments are now the organizations KPIs
- With the KPIs extracted from the company's DuPont model, data driven management is now possible. The company, departments, and individual contributors now understand how their daily work contributes in the aggregate to the model's efficacy
Are you struggling with identifying material KPIs? Is the operating plan a set of forecasts with seemingly little relevance to day-to-day operations?
If so, take the time to develop a company specific DuPont model and agree as a team on the formula's key arguments; assign each team member a set of arguments to optimize, and come to the team meeting with weekly snapshots and trends of the arguments' execution.
In may ways, laying out a DuPont formula is a precondition to building an operating plan, ie it identifies the key business model drivers. Moreover, aligning strategy with a DuPont model allows for orchestrated execution where the component pieces, ie departmental specific activities, sum to a larger whole.