Cisco, Are You Spending the Right Amount on R&D?

Technology companies all face similar issues when it comes to their product portfolio:

  • Relatively short product life-cycles
  • Need to continually innovate
  • Uncertainty of outcomes
  • Build or Buy decisions

We took an outsiders view of Cisco, and its major product lines, excluding its services business which is becoming more important, but will not require major resource investment decisions.

Product information we considered while developing the model included:

  • Size of market
  • Point in product life-cycle (when it was released)
  • Typical life-cycle length for the type of product
  • Investment required to develop new product with similar profile
  • Latest date when a product would still be viable
  • Acquisition cost of companies/products in the market

The data is not exact, but strategy development does not require precise forecasts to give you useful directional guidance.

If we look at the existing portfolio and new product development (NPD) by R&D (or acquisition cost) required (on horizontal axis), and the risk-adjusted Net Present Value (NPV) of the cash flows for each opportunity, it looks like this:


The potential products would be pre-vetted with the usual process to see that they fit within the overall company strategy, that the company has or can acquire the necessary organizational capabilities to develop and deliver the product, and uncertainty has been considered.  In other words, all of the considered projects would be done if resources were available. Also note that the values are risk-adjusted, with at least three cases made for each forecast.

The big provisional note there was, “…if resources were available.

There are no companies with infinite resources, and if there was one, it would have acquired all the others and owned the entire market.  Even a company as large as Cisco has to control its spending on new products. In the most recent quarter, it spent $5.6B on R&D, which was 15% of its product revenue of $37B.

I wanted to see if they were spending the right amount, if they wished to achieve near and long-term growth goals.  Using npv10’s Portfolio Optimizer, it was simple to set constraints on R&D spend as a percentage of the resulting cash flows.  We ran seven different cases, with R&D as a percent of product revenue ranging from 0% to 16%. Once data was in, it took a little over an hour to run this set of analyses.

Each scenario selected a different set of new products with different start times (within the time-constraints allowed).  Sometimes the scenario decided an acquisition would be more appropriate, and other times not. An acquisition may speed things up, but they also come at a higher cost because much of the development has been de-risked.

The results were interesting.  Obviously cutting off all R&D spending helped near-term cash flow, but also dramatically impacted the future as existing products reached the end of their competitive life-cycle.  Spending the most on R&D constrained cash flow over the first few years, but then dramatically increased as all of the new products entered their markets and generated revenue (which in turn provided more cash flow for future R&D).   It also showed how almost all of the cases resulted in the same cash flows 2 ½ years out, but had vastly different numbers leading up to that time and afterwards.

We did not consider additional goals (such as growing revenue, paying dividends, improving margins, etc.) or constraints (headcount resources below some level, number of simultaneous product launches, etc.) but those can rapidly be layered onto the analysis.

In the end, your company needs to decide what your corporate goals are, and what’s holding you back from achieving them.  It’s up to us to help you quickly and optimally determine various paths forward and the trade-offs among them.

Why AI and Goal-Based Planning?

Artificial Intelligence (AI) and goal-based planning go together like peanut butter and chocolate.  It really is an unbeatable combination.

When we first applied our goal-based planning process to a Value Engineering (VE) fund, we increased the returns by an additional 20% and we showed our client that they could either make that much more, by selecting different projects, or reduce their capital outlay by almost that much, and continue making the same overall savings.  

But for many companies, the main problem is not selecting from among a large number of VE opportunities.  It’s coming up with the cost-reduction opportunities in the first place.

Enter Artificial Intelligence.  Our Quanta AI seeks out the opportunities from among your many products or components, based on some obvious information (number of units you expect to sell after reducing the cost, types of components used, etc.) and some not-so-obvious ones, which for now remain proprietary.  We’ll take in whatever data you have though, and find opportunities.

But now you may have too many opportunities, and do not have the budget or people resources to implement them.  Enter goal-based planning.

We take all of the value engineering projects you have available, including some you may have started, add in those that Quanta AI has identified, mix in your available resources, set your targeted goals for savings, and out comes a scheduled list of projects.  If you include uncertainty in your forecasts for one or more of the projects (maybe sales will be below/above target, or cost savings will be below/above estimated, or the project takes longer than expected), we show you the uncertainty across your entire portfolio of VE projects.

Now that you know which projects to pursue and what their total  risk-adjusted outcome should be, you can start working on addressing some of the constraints that are keeping you from saving even more.  Perhaps you add more capital to your internal VE fund, hire additional people to implement the projects, or we can help you identify outside resources for the implementation phase, while your internal development teams focus on new products.

Once we help get your fund set up, and Quanta AI scouring your product and components list for opportunities, we know you’ll agree that AI and Goal-based Planning belong together.  #QuantaAI

Best Alternative Resource Uses (BARU™)

If you have taken any course on negotiation theory, you probably heard the acronym, BATNA, or the “Best Alternative To a Negotiated Agreement.” It simply means you should know what your fall-back alternatives are, if you do not come to agreement with another party. If the “value” (however you define it) of an agreement is less than you could get elsewhere, you probably should not make the agreement.

For example, if you are offered $500k for your home, and then get into negotiations with another buyer, you know your fall-back is to accept the first offer for $500k, and any offer below that from the new buyer is of no value to you.

The same is true for the various resources in your business. I just now coined the term, BARU™ (I like to pronounce it Baaa-Rooo!), or the “Best Alternative Resource Uses.” If your company is going to use various resources to develop a project or product, you should at least consider the alternatives to see if there are more “valuable” (again, however you wish to define value) uses for those resources.

For example, if you have a machine that in one hour can make 100 units of Product A, which will net you a profit of $100, or 50 units of Product B, which nets you a profit of $75, you should probably be making Product A. Your resource, the machine, creates more profit in an hour making Product A. Maybe there’s pricing pressure on Product A, and if the net profit for those 100 units ever drops below $75, you should switch to making Product B. (all else being equal)

Your Manufacturing Manager tells you that one of the parts needed to make Product A is now in short supply, and he will only be able to make 300 units per day. You can probably figure out in your head that you would want to make 300 units of Product A, using up the limited part supply, using 3 hours of the machine, and then switch to making Product B, which is your BARU™ for the machine.

If your company is large, you may have tens or hundreds of different products, all requiring different parts and company resources. You can no longer make all those trade-offs in your head, so you probably try to get some complex software and some person with an Operations Research background.

With Quanta AI, it’s far simpler. You would describe each product you can make with the resources it requires (parts and machine time in this case), and at least the net profit (to simplify things). Then you would describe each of the resources (maybe you have 100 such machines, operating 8, 16, or 24 hours per day, and parts supplies), and then optimize for total net profits. The software will tell you how many of each product to make. It will find the BARU™ for all of the various resources, in total, and recommend your operational strategy.

If you want to then include your corporate goals, that indicate you want to net $100M, the software will either tell you it’s possible, and how, or it’s not possible. And you would then discover which resources are constraining you. Machines or parts. And THEN you would start to address those constraints, either adding more machines, or finding additional suppliers for the parts.

In this way, you now tie your operational strategy to your financial goals. You know what you want to accomplish, you know what alternatives are available to you to accomplish them, you know what resources you have, and Quanta AI recommends one or more ways you can achieve them.

We will come to your company, and quickly help you build the model describing your business, and then you can run all the analyses you want, update the model as things change, and all for a fraction of the cost of alternatives for finding your BARU™!