Understanding Innovation Accounting – a historical perspective
From Cost Per Unit to Cost Per Learning
Traditional accounting methods measure and manage innovation efforts but too often it is one of its biggest disablers. Conventional metrics can’t predict and measure progress in a meaningful way. Ironically therefore, current accounting tools, metrics and funding processes can work against a company’s goals of future proofing through innovation. The inefficiencies in corporate innovation ecosystems are significant. Accounting and the way we measure, analyse and fund projects, portfolios and teams play an important role in this.
Without historical data, traditional accounting is almost helpless in assisting the business in making decisions for innovation projects. The problem is, that the more innovative a project is, the less historical data it possesses. That is the very problem with investing in innovation projects of course and that is why they often fail. How can the variables constituting Return on Investment (ROI) predictions, for example even be close to true when the product and even the market doesn’t exist? Also, how can we hold teams and managers accountable while at the same time give them enough freedom to come up with breakthrough innovation? Well, that’s the job of Innovation Accounting.
To better understand the need for broadening the current toolset, let’s look at it from a historical perspective.
Once upon a time, there was no…
The double entry system of accounting is at least 650 years old and maybe even more than 1,000 years old according to some reports. Accounting is undeniably a very important part of running a successful business. However, accounting has evolved a lot since those early days. Similar to law, accounting is reactive in nature in that it adapts to changes in the environment only when the problems are too big to ignore.
The Industrial Revolution
The industrial revolution led to the creation of large factories, where most value creating activities happened in house. Through that, completely new accounting challenges arose. A company producing fabric, for example, did not have a system that helped them figure out the cost of one square meter of the cloth they were producing.
Calculating the Cost Per Unit (CPU) of producing something subsequently became very important and remains so today. CPU is essential in assigning a value to the inventory for the balance sheet, as we usually don’t value the inventory by its fair market value but rather, by its cost. Following from that, the calculation of Cost Of Goods Sold for the income statement can be made and so on. CPU is a measuring unit that we can hardly live without today but we didn’t always have it.
The birth of Management Accounting
The earliest widespread implementation of management accounting is credited to Alfred Sloan, who was the CEO of General Motors (GM) from the 1920s through to the 1950s. Before the introduction of management accounting, GM had problems with managing its various departments and factories, interstate and overseas. The problem was so bad that at the end of a given financial reporting period there were huge unexpected swings in inventory, expenses and revenue that the head office had no direct way of controlling or forecasting. Sloan’s introduction of management accounting allowed these departments and factories to be responsible for their own budgeting and reporting that could be funnelled into the corporation’s accounts.
To us, today, this sounds so obvious it hardly seems worth mentioning. Most of us have worked in a department where invoices come in, they get stamped and approved for payment and the approver adds the account code for the type of expense with a prefix that designates the department. This amount is then shown in the department’s P&L for the corresponding period and the department’s manager is held accountable against their budget. If the P&L shows an amount that is well over the budgeted amount, questions from a manager will undoubtedly follow.
Accounting is not the problem, but accounting has a problem
In this age of big data, extracting information is simple and cheap. The accounting tools we have available today are designed and work very well if we have historical data available. But when it comes to innovation you have little to no data – you have little to nothing you can use to make informed decisions.
The lack of relevant data is the very reason it is so risky to invest in innovative projects and indeed those initiatives and Startups are very likely to fail because traditional accounting methods are not efficient in handling this.
Another problem with standard accounting in the early stages of a risky project is the accounting department’s well intended risk management is to not waste organisational resources. Early-stage innovation projects are very risky and management discounts it accordingly. Whoever pitches knows this, and in order to demonstrate IRR that justifies funding, they ramp up predictions to the maximum of rationalisation. This process tends to hide assumptions in the plan, rather than unearthing and tackling them. The need to ‘prove’ IRR so early on also gives very innovative projects little chance as they often don’t draw on existing assets but want to create them. Using existing assets of course makes IRR look a lot better better. If IRR is the key metric for picking a project, it makes break through innovation indeed less likely to come about.
Further, today’s accounting tools, especially when combined with a business plan projecting far into the future, leave insufficient room for adaptation. Especially for innovation projects, this can create an incentive to execute a plan on time and on budget that creates no value and only burns organisational resources even though teams and managers knew this a long time ago. This is because we can’t really measure innovation processes and produce reports to inform about the past present and future of a project or portfolio with today’s methods.
This is just an example of how accounting can be a systemic problem for innovation.
The effectiveness of today’s R&D budgets
According to Strategy&, a business unit within PWC, the problem with innovation budgets is so big that after analysing the 1000 most innovative companies in the world for over 12 years they found – ‘no statistically significant relationship between how much a company spends on its innovation efforts and its sustained financial performance’. There are several reasons for this but what is certain, is that simply spending more while still using a traditional approach doesn’t pay off. Funding decisions are often made on the common belief in fiction not facts (as there are very little knowns). The communication abilities of individuals are more important than the truth. What is easy, is to retrospectively rationalise a denial of responsibility.
‘Without data, you’re just another person with an opinion’
Thanks, W. Edwards Deming
A system that doesn’t honour managers and teams clearly stating the known unknowns of an emerging business model and a clever way to tackle those, is very likely to return nothing. It should not be too surprising that the question ‘whether to invest and if so, how much’ for a project with no significant traction, can hardly be determined based on a Net Present Value (NPV) calculated for years into the future. One thing is certain, the calculation will be wrong.
How is it helpful to predict Cost Per Unit (CPU) when we are not even sure if the solution we are envisioning really solves a customer problem and won’t change anymore? At worst it holds people accountable towards building something that nobody wants.
Accounting should assist in assessing the past, present and future performance of a company/project and subsequently inform investment and management decisions. As predicting the future is very hard, we need a system that enables us to test the maximum amount of ideas to find the best ones.
The next re-evolution
Equally to Sloan’s need nearly 100 years ago, managing innovation requires us to gather information we didn’t need before. We need different information in accordance with a project’s stage within a product life cycle (more about stages within a PLC in another post). It is information that , for example, enables a metered funding process, informs about perceived customer value, and makes teams more agile yet comparable and accountable. We need a system that helps to make informed decisions about placing the maximum amount of well measured bets on the best ideas. We need different reports about innovation projects and portfolios.
Innovation’s main activity is learning. A manager’s main job is in fact effective learning management. We are producing knowledge about how to create, deliver and capture value. In order to manage this we need different KPIs. For example, we want to make sure that teams are not only learning the right thing at the right time but learning effectively and Cost Per Learning (CPL) (and Time Cost Per Learning) is a useful metric for that.
Information like what has been learned, how has it been learned and how much did it cost needs to be recorded in an ‘innovation ledger’. The more data we collect, the more accurate we can then become at budgeting future learning milestones and measuring performance. This is one example of how accounting can now do what it does best and draw on historic data to make better decisions for the future of innovation projects.
Information like what has been learned (Do they want it?), how has it been learned (Smoke Test), how much did it cost (Wages+Other) and how long did it take (i.e. a week), needs to be recorded in an ‘innovation ledger’. The more data we collect, the more accurate we can then become at budgeting future learning milestones and measuring performance. This is one example of how accounting can now do what it does best and draw on historic data to make better decisions for the future of innovation projects.
What to learn = How to learn = How much does it cost
NPV and Innovation Accounting
Of course, at some point in the innovation process, we want to include financial metrics such as NPV. It is important to note that the variables in the calculation should grow with the maturity of the product. No matter when we start demanding this calculation, we need a system that allows for NPV to be recalculated after an experiment delivers a new data point. Combining this approach with an iterative funding model ensures a highly effective innovation budget.
What are the most risky assumptions for a project that is scaling and how do real data impact NPV and subsequent IRR calculations? In order to answer this and to understand past decisions, it is essential to have a system that gathers these information and makes them auditable. Without learning report cards and standardised frameworks, this seems unachievable.
Management accounting doesn’t replace the company’s statutory requirement to report to government agencies, such as the ATO. Management accounting operates parallel to a company’s statutory obligations and serves the function of maintaining financial control and accountability to its managers. In the same way, Innovation Accounting does not replace Management Accounting; it is an additional tool.
In essence, we need to understand different things at different stages of a product life cycle; break the system down just like Management Accounting does to understand the past, present and future of projects and portfolios better so we can make better decisions.
We are entering the age where companies might create their own innovation ledger. Certainly, there will be a time where managers are able to understand and create new types of reporting and companies invest more like sophisticated Angels and VC’s.
Looking forward, Forensic Innovation Accounting will be a very important job that has not yet even been imagined.
It is exciting!