What launching rockets can teach us about revenue management

Would you promise to send a rocket to Jupiter before you proved you could get one into space at all?

Many companies view optimizations–be they pricing, or anything else–as something that you need to get perfect at the beginning. When John F. Kennedy challenged America to send a man to the moon in a decade, NASA started with just getting a rocket into space. Executives should take note–low earth orbit first.

In revenue management, price optimization, and other deep business related optimizations, it is easy to expect perfection at the outset. This is a huge mistake, for several reasons.

Are you headed for the Jupiter trap?

Whether working with an internal team or an outside vendor, there are three questions you can start with to prevent disaster.

When is the soonest we can realistically launch a basic solution?

Ask your vendor or internal team this question and be skeptical–not that it's not soon enough, but that they are over promising. One major global car rental company did a 6 month project to launch revenue management; two years later it hasn't launched.

When is the soonest we can realistically launch our perfect solution, and what resources would be required?

Everyone likes perfect. So put a timeline and cost estimate on it–without being alarmed. For most companies it might take 5-10 years to get to a "perfect", custom solution, that addresses every nuance of your very unique enterprise. That's, assuming the market doesn't change drastically.

Have the same team or vendor give you an estimate for "perfect". Hint: good vendors might not be willing to.

How much are we losing by waiting?

Now, using both estimates at face value, estimate how much you'll lose by waiting. There are two concrete items here: the additional estimated cost of the bigger project every month, and the lost revenue, profit, growth (or whatever your goal is) during the intervening months or years.

Take an example. The basic solution–let's call it LEO for Low Earth Orbit–can be ready in 3 months, at $250,000 per year. You have a $300,000,000 business and expect a 1.5% increase in revenues which, because it comes at no incremental cost, will flow directly to your bottom line. That is $4,500,000 per year in incremental profit.

The perfect solution–let's call it Jupiter–will take 3 years, or 36 months. When it's live it will be amazing, and result in a 2.5% increase in revenues. That's $7,500,000 per year in incremental profit. Let's be generous and say it will take 2 engineers, which at current rates would equate to around $400,000-500,000 per year in cash and stock, to create, and let's be aggressive and say it's "free" to maintain once it's live.

See where I'm going with this?

After 2 years, LEO will have created a net $7,375,000 in value for you (that is, after fees and accounting for 3 months of waiting before it works). Meanwhile, Jupiter will have created - $800,000 in losses, and shown no benefit.

 

what_have_you_got_to_lose.png

Download the spreadsheet here.  

In fact, with these assumptions, as long as LEO provides at least 1.1% increase in performance, over 5 years it will beat Jupiter with no further improvement. Obviously, you might continue to build towards that Jupiter launch once you've established success with LEO.

If Jupiter under delivers or launches late, it only gets worse.

Why faster payoff matters

Companies are not run by rational machines. Companies are run by humans–highly emotional creatures, who conceptually understand math but prefer gut decisions.

If you were CEO and saw a 2 year old project that cost a lot of money, and is still 1 year from creating value, what do you think of that program and the people running it? No promotions for a year!

Meanwhile if you're CEO and saw a project that added meaningfully to the bottom line, increased revenue growth, and showed ROI after just 3 months–yeah, you promote that person. She knew what she was doing. And you give her more budget to do more of it.

Conclusions

You'd never launch a rocket to Jupiter before you tried to make it into low earth orbit first. So why hold out for a big, complicated perfect optimization before you optimize something simple? The cost of perfection is too high.

Related Posts