How To Make Your Financial Modeling Accurate

Successful business man. Cartoon character.

I’m going to let you in on a secret. The key to developing accurate financial models has nothing to do with your ability to understand financial planning or business plans.

The key to accurate financial models rests on your ability to launch your products on schedule. I’ll explain why later.

[Do you want to grow your business? Maybe I can help. Click here.]

Estimating your expenses is really easy.

“How’d you do that?” I asked my friend and mentor Dave.

“When you’ve looked at enough companies, you kind of know how much money they’re spending,” he said.

Dave had asked me what our headcount would be at the end of year one. Then Dave did some simple math. “At $200K, fully loaded, per head, you’re going to burn around $4M (in your first year of operation).”

The actual number was $3.74M, but Dave’s estimate was pretty close. So, if you want a simplistic model you can use $200K/head to determine how much money you’re going to spend each year.


Sadly your revenue plan is likely to be way, way off.


“Most startups are horrible at forecasting their revenue, and you’ll find us to be no different,” Barry, our chairman of the board said to a prospective investor.

Everyone laughed because there was a lot of truth in what Barry was saying.

Startups are usually horrible at forecasting revenue for three reasons:


A. Founding CEOs are way too optimistic.


You think your business is just going to take off when you launch your first product. Reality of course is way different.

Growth is slow, very slow. You may have to change your pricing, your business model, your go to market strategy and more before you gain traction. Plus…


B. You have no data.


Your first few attempts at forecasting are just calculated guesses because you don’t have any data. You’re just grasping at straws really.

But, over time, you do get better at forecasting.

Your forecasts get more realistic because your board (at some point) starts putting pressure on you to be accurate. Plus you start getting some data.

We used two simple methods to improve forecast accuracy:


A. Bottoms up forecasting.


Bottoms up forecasting is as old as dirt. You simply forecast by each customer what you expect them to do for whatever time period you are forecasting.

Focus on the 20% of your customers that make up 80% of your revenue. Then extrapolate the rest. Then…


B. Use “The Graham Technique.”


The Graham Technique is something we borrowed from Benjamin Graham. Graham wrote The Intelligent Investor an incredible book on value investing.

Graham would analyze the growth rate of stocks by taking the average of the last three years of a company’s earnings and comparing that to the average earnings 10–12 years previously. He would then have an average earnings growth rate for a company.

You can apply the same methodology to forecasting revenue. Take the last three months of company’s revenue and compare that the revenue 12–14 months earlier.

Now you have an average trailing 12 month growth rate that averages out any large swings up or down.


C. Then you compare your bottoms up forecast to your Graham forecast.


The nice thing about the Graham forecast is it’s just numbers. So you are forced to explain why you are expecting higher or lower growth to the trend.

It’s not perfect (nothing is with forecasting), but comparing the bottoms up to the Graham results keeps you honest.


The biggest problem you’re likely to miss your forecast is you’ll miss your original product launch date.


Let’s say you’re planning on launching your first product or service in January. Your original first year revenue plan is:

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

10 100 1.1K 2K 4.0K 6.0K 8K 10K 12K 16K 20K 25K = 104K

But, there’s been a six month delay because your team underestimated how long it would take to develop the product. Now your first year revenue is only 31K

The problems don’t end there because product delays cascade through your revenue plan forever.

Now your year two revenue plan drops from $1M to $200K. And your year three revenue plan drops from $3.5M to $1M.

See the problem. And we haven’t even added in that you probably assumed revenue would ramp faster than it did.

If you want to get your financial model reasonably correct, manage your product development, so your products come out out on time.

We had a plan of announcing 5 product sets in our first year. If we were late announcing these products, we would miss our revenue plan by a long distance. So, I went into my bag of tricks and suggested to Jeroen, our VP Engineering, that we use a warboard methodology. Warboards, for those of you not familiar with the concept, are a great tool when absolutely, positively have to get something done by a certain date.

Jeroen’s warboard was a daily schedule for each of the five product sets. We knew exactly where we stood on every product every day.


You need a great team if you expect to execute well.


I don’t care what operational plan you have if you don’t have a great team around you to execute your plan. I learned this early on in my career at Micrel.

Micrel had operational plans but its execution was mediocre at best. But, then again, many of the people weren’t that great either. So the plans were irrelevant.

The sooner you learn the lesson of how important your team is, the better off you will be. Interestingly enough, really good teams develop better and more detailed operational plans too.


You need to relentlessly follow up if you are going to stay on plan.


You have a great team and you have a detailed plan of what’s going to happen. Now this is where good management comes into play.

Jeroen, for that first set of products, held a daily meeting with his engineering team. He knew where we were ahead of schedule and where we were behind schedule.

The good management part came in not only knowing where we were, but rallying the team to help anyone that was behind schedule. The result was we taped out all 5 product sets within 8 months of launching the company.

We were well ahead of schedule for receiving our tranche, and we got our funding without any hangups.


You can’t follow up daily forever or you will burn out your team.


Warboards are great when you have a short time, life or death, time crunch like we did. But you can’t operate with that amount of stress forever or you’ll burn out your team.

As we transitioned to the longer term slog of building our company I again reached into my bag of tricks. This time I suggested to Jeroen that we follow the “Critical Chain” methodology.

The Critical Chain methodology was developed by Elihu Goldratt. Goldratt wrote a best selling book called “Critical Chain” that goes into a detailed explanation of the format.

It’s a great book that I highly recommend you read. Jeroen had never read the book, so he said he would read it, and I said I would re-read it.

Jeroen became a true believer in the Critical Chain format. The two sentence explanation of Critical Chain is you develop a median case schedule for each task in a project and you determine the buffer for each task. Instead of putting buffer between each task, you have an overall project buffer at the end of the schedule.

Jeroen implemented Critical Chain, and the result was we actually improved our time to market. We weren’t perfect, but the results proved the concept worked.


When you execute your plan on schedule, your financial planning gets a lot easier.


As I said, the easiest way to become an expert on financial modeling is to get your products out on time. If you screw up your development schedules it will not matter how good you get at financial modeling. You will find yourself missing plan after plan and redoing plan after plan.

For more, read: What's A Warboard And Why Do You Need it?


Do You Want To Grow Your Business?  Maybe I Can Help.  Click Here.