My Great Lesson in Cash Flow

One of the most important experiences I’ve ever had as an entrepreneur, or as a business person, for that matter, was when the first company I founded, DataWare Logic, crashed full-speed and completely unexpectedly (to me) into the wall of an empty bank account. I fondly remember the experience not because there was anything pleasant about it at the time, but because it was among the most profound learning experiences I’ve ever had. I’m not a fast learner, but the kind of pain that comes from virtually going bankrupt is not forgotten quickly. Needless to say, I never ran out of cash in any business again.

I joined a startup right out of college (actually, I quit college after my sophomore year to join this startup, but that’s another story . . . ).  Health Care Computer Systems was based in Jenkintown, PA.  I was its sixth employee.  The company wrote billing and forms management software for pharmacies, nursing homes and hospitals.  At the time, virtually all insurance forms, including Medicare, were done by hand and the company saw a huge opportunity to automate this process.  It made loads of sense to me.  Besides, I had a chance to write software under RT-11 on a PDP-11 using a VT-52 terminal – this was very exciting since I was trained in school to program in FORTRAN on punch cards fed through a slot in a wall to some grad student who fed them into a mysterious CDC 6700 black box that I never actually saw.  Yeah, I’m afraid that I’m that old.  But I digress . . .

Health Care Computer Systems failed after I had been there for a little less than two years.  I learned a lot about what not to do in running a business from that company but I was way too young and naive to learn half of what I could have.  You can’t learn from what you don’t see or understand.  One thing I didn’t learn was the difficulty in trying to automate a government process that the government wasn’t ready to have automated.  So, thinking I could do it better, I set off to make my mark on the world.  Little did I know that my mark would be more like a crater than a monument.

I started DataWare Logic out of a bedroom with one part-time partner.  I was flush with excitement when I signed on a couple of Health Care’s customers to support contracts so I emptied my bank account and bought a brand-spanking new Tandy TRS-80 (hey, who knew IBM was gonna’ enter the PC market?); a used PDP-11 with a 5MB RL01 platter drive; and I splurged on a new VT100 terminal.  I was so excited I barely slept for 6 months while I worked on code and supported my two customers.

I eventually realized that it was harder to acquire new customers than I had originally expected.  They trickled in, though, and I redoubled my efforts every time I opened an envelope with a new check.  In the mean time, PDP-11s, while fairly studly machines, were not maintenance free.  And, since there were no PDP-11–World stores in strip-malls, self-maintenance was far from a slam dunk.  I also had to buy another RP01 disk to back up my massive 5MB of storage – let me tell ya’, these things weren’t cheap.  My angst in writing the big checks was offset, though, by the shear excitement I got from having all this glistening hardware at my disposal and from the money that kept rolling in.

One day, I got a call from one of my original customers who told me that everything was working great.  So great, in fact, that he saw no reason to continue paying me maintenance.  The underlying forms weren’t changing and the government and insurance companies hadn’t kicked one back in months.  This call was soon followed by one from another customer who thought that I was way over my head and would never get his system working (he was probably right).  He not only told me he would stop paying, but he demanded his money back (that never happened).  As I scrambled to bring on new customers, I had a few hardware failures.  I lost a bunch of code I had written and I wasn’t able to support my customers – the hardware needed to be repaired fast.

As I wrote the DEC (Digital Equipment Corporation, the makers of the PDP-11) service guy a check, I realized that my balance was precariously low.  What I didn’t realize (who knew what a spreadsheet was?) was that my costs were now way above my revenue.  Even though my DSO was almost 0–days – it was essentially a cash business – I all of the sudden had flipped from a cash generating business to a cash consuming one.  The concept of high fixed costs (relatively speaking) was foreign to me.  My head was down and I was driving forward.  That’s all that’s important, right?

Even more shocking was the speed at which my business flipped from success to failure.  One day I watched my bank account grow at rates that blew me away and the next, I sat looking at all my cool computer equipment deciding whether it was more important to me and my business than paying rent that month was.  I didn’t have time to get new customers while supporting my existing ones and my company virtually augered itself into the planet in an instant.  I shut down the company just a month or so later and ran to work at a big company to lick my wounds.

After that miserable experience, I changed the way I did things when it came to cash flow.  I not only started to model my businesses moving forward in as detailed a process as possible (automated spreadsheets really helped, of course), but I learned to do sensitivity analyses, varying several key assumptions in an attempt to better predict a range of potential outcomes.  It’s comforting to know the likely scenario but, to me, that information has little value without knowledge of the potential negative and positive outcomes. 

Over the years, I have become much better at determining the most important assumptions to vary in my models and understanding these is a high priority for me when I make a new investment or join a new board.  Ultimately, you can’t accurately plan your business without understanding what these critical factors are.  Obviously, the factors are different for various businesses, selling models and markets.  To get a handle on them, though, think carefully about the fixed and variable costs in your business.  Consider how to keep fixed costs as low as possible; what is your real cost of sales; what are your key out-of-pocket expenses in the business; how to continuously monitor your development costs; and how your spending can be rapidly changed if revenue falls off fast or doesn’t materialize as soon as expected.

In my experience, over time you develop a better feel for the business and you’ll be able to minimize the list of key variables in your model to the few critical ones.  While there is no panacea for cash flow issues, careful planning and modelling of potential upside and downside business scenarios goes a long way to making sure that you don’t have to learn this lesson the hard way.