Big Data

Big Data Not Required for Making Big Decisions

“Big Data” has been all the rage for a while.  Everyone is trying to figure out what to do with all that data that your business is collecting about customers and operations.

I want to propose the “Big Data Decision Value Theorem™” :  The value of a decision is inversely proportional to the amount of data you have available to make it.

In other words, if you have billions of data points, any individual decision made is probably not worth a lot.  Collectively, all those tiny value decisions may add up to be worth a significant amount, which is why companies like Amazon, Google, and Netflix collect so much information about each individual using their services.  The value of a single ad delivered to a single user isn’t worth much to Google, but scaling it to billions of ads served to a billion users, and we’re talking real money.

Conversely, making large strategic decisions requires less data than you would imagine.

Okay, so you probably wouldn’t take this to the extreme, and say you only need 1 (or no!) piece of data to make a “bet your company” decision, but the idea of managing the businesses in your company does not require a billion pieces of data.  You probably have a good idea of how a business is doing by looking at a few charts that show its current state, past trends, and forecasts.  Reviewing the competitive landscape rounds out the information you need to make decent decisions about a single business.

But what if you have 30+ businesses or 100+ properties and are trying to view them as a cohesive whole?

While you can’t begin to understand the interactions of all the businesses by looking at all their separate charts, computers can aid you.  They can notice that 7 businesses generate excess cash flow during times when 14 other businesses need cash to grow.  Or that adjusting when new opportunities are executed can keep you from needing additional capital.

These are the kinds of tools that npv10 is building.  They aid your analysis, and help make decisions about what to do, and when.  They only require the same data you probably see in reviewing each business separately.  And then you can start thinking strategically about what your company’s overall goals are, and the tools will help you figure out how to get there.

We will be glad to do a quick analysis for you, using what data you have, to show you the value your company can create by using our products.  As the tools are not yet released, we can do some number of analyses free of charge.

$300M in One Morning!

As I continue talking with companies, and trying their real data, the value of the software is proving itself.

One company allocates $25M/quarter to invest in opportunistic projects.  They provided ~400 such projects, ranging in investment size from $100k to $35M total, with the investment spread over 1-16 quarters, and the return coming back over time (up to 8 years).

“How would you select them?” I asked.

They would typically rank them by decreasing NPV, and select projects until their available capital was allocated.  This is a typical “rank and cut” approach that many companies use, and seems perfectly reasonable.  It’s straightforward and relatively simple to understand and implement.

I manually replicated this process, investing in projects (starting with the greatest NPV project first, and continuing down the list) each quarter until the available capital was exhausted.  This approach invested $500M over 20+ quarters in the 80 most valuable (by NPV) projects.  As these are forecast to be very valuable projects the NPV of the resulting cash flows was forecast to be ~$3.6B.

Wanting to see if the software could improve on that, I limited the quarterly investment spend to $25M and a total $500M over the total time-frame.

Honestly, I didn’t expect to get a result much different from their approach, given that they were already picking the 80 most valuable projects.

In 10 minutes, I was able to add $300M to their portfolio’s value!

It again invested the same $500M over 20+ quarters, but instead picked 226 different projects, and again never spending any more than $25M per quarter.

The optimization approach identified the optimal set of projects and their timing to maximize the NPV of cash flows, given the various constraints. (you’re actually able to maximize or minimize any of the key metrics you provide or calculate).

Also, in this case, picking a larger number of projects reduced the risk because the portfolio was no longer dependent upon relatively few of them to succeed.

As I look forward to using these tools to help clients, and eventually getting the tools in their hands, I expect there to be many more such successful mornings.

Hello world!

It is finally time to hang my shingle out, and let the world know what I have been up to for the last 5 months.

I’ve been attempting to solve that formula at the top of the page…

Okay, that’s actually just PART of the formula.  The actual formula is far longer, and it is just one of many other formulas that must be created and calculated.  This goes part of the way towards explaining why not everyone is doing this. It is not easy.  But solving thousands of simultaneous equations is not what I am ultimately trying to accomplish.

I am making it far easier for companies to quickly evaluate their strategic alternatives, in terms of financial and operational performance.

As the software progresses, I will share more details.

For now, just know the answer to the equation was 728. That wasn’t so hard, was it?