NPV doesn’t work for risky, high-potential projects. Here’s a better approach.

This article was co-authored with Claus Hirzmann of Strategic Finance 

“Never test the depth of the river with both feet.” Ironically, this is the very behavior that plagues many of the innovations that ended up in Rita’s “flops file.” From Disney’s Star Wars Hotel to Google’s Stadia to Anheuser-Busch-Keurig’s DrinkWorks and more, these projects all feature massive up-front investment and detailed years-long plans. There’s a smarter way to allocate resources to big bold things.  

 

Image Source

Innovation and Net Present Value calculations – two mistakes

The net present value (NPV) decision rule is widely adopted and often applied to decisions with respect to investing in innovative new ventures. The rule simply suggests that one project cash flows in and cash flows out for a given initiative over some period of time, then discount back to the present to account for the cost of capital. This is more than highly problematic for new ventures. There are many reasons why. 

Ventures are uncertain, so the idea that you can predict cash flows over the lifetime of a project is absurd. NPV further assumes that beginning a project means you will carry it forward to its conclusion. That may not be the case – in fact it will seldom be the case with truly innovative ideas. Most are not going to work out. Using NPV as a corporate decision rule also leads to people gaming it – overstating benefits, understating likely cost and time or otherwise misrepresenting what the venture’s advocates truly believe. And “winning” a pitch based on NPV almost certainly ensures a venture will be designed to accommodate that analysis – as though it were going to go through to market launch, meaning a big team and all the funding up front.

This creates two fundamental problems. The first is that many ideas worth exploring at the idea stage get too big too early. That in turn means that they gather momentum, and it becomes nearly impossible to stop them. The second is that, when looked at through the lens of NPV, many truly original projects will be rejected on the grounds that they don’t look promising enough.  

There is an alternative approach, based on real options in finance. In brief, what this approach suggests is that you think of investments in innovations similarly to investments in financial assets. Just as you might take out an option on the future performance of a stock, and decide to exercise it only if your option is in the money, so, too, you can make an investment in an innovation-related activity and only move forward with commercialization if information at a later date suggests this will be worthwhile. As with a financial option, if things don’t look promising, you stop the project. It’s important not to think of this as failure, but rather as tuition of sorts for the learning that can come with intelligent failures. So, options are small investments you make today that buy you the right, but not the obligation, to make more substantial investments in the future.   

Practically, what this looks like is breaking your innovation-related investments into concrete steps, with the option to stop at every step. This is the essence of discovery driven planning.  

But I still need some kind of number to justify an investment!

Although a lot of people will nod in agreement at this argument, when they face actual venture budgeting conversations, it is very difficult to leave the hunger for some kind of NPV calculation behind. This is why I (Rita) was so excited to see Claus’ approach to modeling option value at work. Indeed, NPV and real options are both financial metrics for expected value creation; NPV applies to predictable, linear projects, and real options applies to agile, discovery-driven initiatives. 

The first principle of real options is that investment commitments are always limited to just the next step. Instead of commitment to the whole program, the other steps are subject to discovery and will be allocated resources only if earlier assumptions are validated. Adjustments made after each step can be manyfold. For example, you may continue investing as planned, increase or decrease the investments, stop them with or without retrieving a residual asset value, or pivot the entire initiative. Beside investing in the next project step, later decisions will concern Go-To-Market investments and operational right-sizing. 

Importantly, these investment decisions can be taken quickly, because the real options business case can be built fast. Speed of discovery reduces risk of execution. Financial risks are minimized by a stepwise and adaptive investment strategy. Moreover, to further reduce risks, bigger investments are deferred until the most critical assumptions have been verified. Fast investment decision-making translates to fast Time-to-Market, which is crucial for innovation.

Disposing of a suitable financial valuation rationale also drives innovation focus and efficiency. As Steve Jobs put it, focus is about the ability to reject most ideas and keep only the very most promising. This becomes possible in a truly rational way.

Using Claus’ software, and the best information you happen to have at hand, one can come up with a real options based valuation that far more accurately gives a sense of what a project is truly worth – solving both the error of not investing in truly promising projects or investing way too much in projects that look great but are nonetheless highly uncertain.  

A real life example.

Figure 1 shows an example from the software of a multinational IT company that considered the creation of an innovative IT solution for the US market. The way the software is constructed is that it models the possible paths a project might take and attaches financial assumptions to each decision. Each project step (see blue bricks) provides an option on the next step and ultimately on the financial returns (ROI, see yellow bricks). The investments in each step are shown as red figures with the possible returns as green figures. There were three major outputs from the planning process envisioned – an early proof-of-concept version of the solution, similar to a minimum viable product (MVP), version 1 of the software (V01) and the more evolved version of the software that would fully deliver the vision for it (V02). Each step forward is assigned a probability, which allows the software to calculate the value of the option of moving forward.  

Figure 1: Tree structure of the possible trajectories of a discovery-driven business case together with main inputs.

The leaders of this business opportunity were excited by the potential business benefit. Reaching this benefit, however, was subject to three assumptions: (1) The solution would be technically feasible, (2) The general market feedback would be positive, and (3) It would be worthwhile to create a full-fledged solution, which we named v02.

To investigate each assumption, the team contemplated the following major steps: (1) Build a prototype or minimum viable product (MVP), (2) Show the MVP to prospective customers to gather market feedback, and (3) Perform pre-sales efforts in parallel with the R&D needed to develop v01, such as to understand the concrete market outlook.

The “success” pathway was: If the MVP is feasible (step #1), then continue and show the MVP to prospective customers (step #2). If both steps are positive, then invest in both the R&D for v01 and pre-sales efforts (step #3), else stop. Upon completion of step #3, explore the market outlook. If it is good, proceed with sales of v01. If it is very good, invest in the R&D for v02, while selling v01 in the meantime. The idea would then be to switch to sales of v02 when development is complete. If, at any point along the way the evidence does not support continuing the development, funding would be discontinued and the team would disengage from the project.

Structuring a sequence of decisions this way allows the software to calculate a real option value. In Figure 2, the Real Option values are labelled on the respective bricks, net of the eventual investments (investments are shown as the red figures next to the blue bricks).  

 

Figure 2: Real Option valuation of each project step

The net Real Options value of taking the very first step – the investment to create an MVP is +$7.2. That means it is worthwhile to start by investing in the first step. Essentially, you are making a small investment (that is affordable to lose) for the promise of accessing a substantial upside. The investments in the subsequent steps are open and subject to the results learned from the preceding steps. The investment may be adjusted or even stopped. This investment adaptability reduces financial risks. As you’ll see, the real options value increases substantially by the second step, the third step and so on.  

Importantly, if this initiative had been managed in a linear, rigid manner, with all the possibilities entered into a total-project NPV calculation, the expected NPV would have been -$23.3. The initiative would have been rejected right from the outset. This indicates the strategic importance of taking a real options approach to highly uncertain ventures – the ones with greater substantial upside are likely to be under-valued when taking an NPV approach and are likely to be rejected.  

The under-valuation by the NPV results from these attitudes: 

• Rigidity: The initiator considers making all necessary investments to fully implement the innovation. However, this is like gambling, as the ROI is uncertain and the investments may or may not be justified. This leads easily to a negative expected NPV (here: -$23.3). 

• Fear of over-commitment: The ROI projections in NPV calculation are like promises to the sponsor. While the upside potential of the innovation is exciting (here: NPV = +$39.1), it is too uncertain to be committed to. Therefore, innovators will announce some safer, lower ROI numbers, leading to an unattractive or even negative NPV.  

In both cases, the mistake is to neglect an agile, discovery-driven investment strategy that would decrease financial risks and thereby would enable access to an exciting innovation opportunity. However, using the probability-weighted NPV as a way to account for an agile, discovery-driven approach would be misleading. This is because the project specific risk levels would be under-estimated by presuming average, Business-As-Usual risks, leading to an over-valuation (here: +$14.7) instead of the correct real options valuation of +$7.2.  

Connecting these calculations to your larger portfolio of innovations

Successful corporate innovators approach managing their initiatives through a portfolio approach. By using the software, each time new discoveries are available, new opportunities arise or the strategic context changes, entering new information into the software allows for dynamic re-calculation of portfolio value. This valuation can drive focus and efficiency, and further can provide financial prioritization metrics, useful for allocating constrained budgets.

A point we can’t emphasize enough is that to be successful, this approach requires a ‘funding’ approach to the allocation of resources, not a ‘budget’ approach. So, while budget envelopes might be predefined, the allocation of resources to steps in the learning process is not, at the outset. You might, for example, establish a budget for “robotics projects” and have, say, 10 initiatives that fall under that umbrella in the works. Allocation of resources to each specific project only occurs as it meets key checkpoints and only if the projects’ prospects merit it.  

Another way in which the software and resulting calculations can be helpful is by incentivizing the right way of managing project disengagement (or discontinuation). Under NPV logic, failing to meet specific project plans amounts to, well, failing. This is a huge disincentive to people in typical companies because such failures (and they are often huge, large-scale failures) represent career risk. You’ll never get people to be willing to work on an uncertain venture if they feel they could be blamed for not having all the answers up front. With this approach, the unknowns are spelled out, and it’s quite clear that management doesn’t expect everything to work out as originally envisioned.  

Showing the discipline

Innovation leaders are often accused of operating without discipline, of wasting money, and of going off on tangents when funds and talent could better be allocated to an existing business. Using disciplines such as real options modeling combined with discovery driven planning, leaders can demonstrate to the rest of the organization that thoughtful planning and execution has indeed happened. That is very different from promising specific results and is much more suitable to the high uncertainty / high reward investments in innovation than standard issue planning is.  

If you are interested in learning more about real options and discovery driven planning, contact us for a free case study.

*Originally published by Thought Sparks