You’ve no doubt seen multi-year financial projections prepared by accountants.  You get a plethora of spreadsheets showing income statements, cash flows, balance sheets, capex, assumptions, amortizations, tax estimates and all manner of interlinking numbers from sheet to shining sheet.  These thorough presentations have their place and are worthy products from anyone training in the accounting profession. 

However, at the startup or early stages of a company, this mountain of data often tends to obscure the numbers on which you need to base your decisions.  I spent time this past weekend wrestling with a workbook with 17 spreadsheets, many of them rolling out over 60 months and including summaries at the end, curiously enough ending up with the column “BS” – and I’m not making that up.  I was preparing for a potential investor an analysis with respect to trimming distribution methods and overhead, and I quickly found the expected issues with broken REF’s when I tried to make even minor changes.  Performing surgery on a model of that scale is almost more difficult than building it in the first place. 

I’m sure the original author of the model was following a template he or she uses often and probably could have manipulated it more easily than I could, but that person was not on the scene.   Just as coders have a natural instinct to throw out everything written before them and rebuild software from scratch, I believe spreaders of sheets have the same tendency.   We all develop our personal styles that are most comfortable and most easily managed through repeated iterations.

The core issue of an overly complex model for a startup is that it is just compounding from day one assumptions that have not been tested.   I can understand why many investors no longer look so closely at the numbers, particularly anything beyond the first year.  I heard a VC long ago say that with the invention of spreadsheet tools everyone could produce a pretty model with very little thought as to its core elements, and he thereafter paid relatively little attention to them.

I ended up boiling all this down to a new two-sheet workbook that addressed the investor’s core questions about runway he would be buying with his dollars.   The financial expert at that firm had been given the complex model but probably like me found it hard to tackle.  Seeing everything condensed made visible the actual cash needs under the newly trimmed format.  But, as a byproduct, it also revealed some flaws in the pricing of a strategic relationship that needed to be addressed as well.

We went from a mountain of GAAP data to a pretty simple picture of a real live situation.  We had something on which to make decisions that could lead to a better outcome.  And, some beneficial changes in the operation of the company will be a result.

Some of my core recommendations for creating a decision model are these:

1.  Keep the important assumptions highly visible and make them easy to change.   You should be able to see on one summary page the direct result of bumping prices or changing any assumptions about volumes.   You essentially need one dashboard that lets you manipulate the key variables and measure the consequences.

2.  Let your accountant worry about the GAAP presentations.  Your only interests are whether your plan has sound arithmetic and how much cash is needed when to enable you to execute that plan.

3.  Stress test that plan.  Try the extreme ranges of your input variables and see what happens.   You’ll discover how much cash cushion you may need if products get delayed or the revenue ramp doesn’t quite have the positive second derivative you estimated.

4.  Keep the presentation simple.   Staffing details, for example, can be in an ancillary sheet.  Investors will assume you can manage expenses to match the needs of the business, and their main focus with respect to staffing will be a reality check on head count versus revenues.  If you are showing 5 people generating $50M in revenues, you probably will have some explaining to do.

5.  Once you have gotten comfortable with your model, and you have raised your financing, try to lose it quickly.   It will change next month anyway, and you’ll be working on a new model after your first few tests on the battlefield.

For those of you youngsters who don’t recognize the image at the top, that’s the implement that put men on the moon and was the basic tool behind every financial model for startups in my early career.

<photo of teaching slide rule on author’s mantel>