The Limits of Quantitative Business Forecasting

As business owners, we use forecasts to understand macroeconomic and microeconomic factors contributing to the success or failure of a product and service.  We use these forecasts to evaluate product ideas and set overall product strategy.  Creating accurate forecasts is however, notoriously difficult.  And setting business strategy around grossly inaccurate secondary forecasts is downright disastrous.  Given the surfeit of industry and technology forecasts, how do we, as shepherds of corporate strategy, determine which forecasts are valid?  Being vaguely accurate is after all better than being gravely inaccurate.

How To Evaluate The Accuracy Of A Business Forecast

If forecasts allow us to implement long-term strategic objectives and quantify future risk, how do we know if our forecasts are accurate?  Making strategic decisions based upon an erroneous forecast could after all, materially and adversely impact the financial health of our business.  Take the cash flow forecasting methods previously discussed: If these projections were to determine your company’s future hiring, and if they were grossly inaccurate, you might find yourself with a serious cash flow problem.  So how do you determine the accuracy of business forecasts?

Using Markov Models To Estimate Accounts Receivable and Cash Collections

This article is a continuation of:

• Managing Your Small Business’s Cash Flow
• Translating Raw Sales Data Into A Cash Flow Statement
• An Overview of Sales Data and Markov Models: Finding Transition Probabilities

This is the final post in the series on managing cash flows. In the previous articles, I covered the impact managing accounts receivable can have on a company’s financial health.  I discussed two methods to forecast collection rates based upon current accounts receivable.  In this post, I will compare these methods (simple average & markov models) against the actual collection rate, and then use the Markov Model to create a cash flow forecast.

An Overview of Sales Data and Markov Models: Finding Transition Probabilities

When I first encountered Markov Models, I was a little overwhelmed.  The Wikipedia entry did an excellent job of adding to the confusion.  Thankfully, Markov Models are not nearly as complicated as they seem and are in fact, incredibly useful.  Fortune 100 companies, for example, use Markov Models to forecast gains or loss in market share.  They are also a valuable cash management tool.  As small business owners, we can use Markov Models to forecast the percentage of sales that will go uncollected.  This methodology yields forecasts which are more accurate than those obtained through taking the simple average.

Translating Raw Sales Data Into A Cash Flow Statement

In my previous post, Managing Your Small Business’s Cash Flows, I alluded to writing posts on using Markov Models to forecast cash flows and cash collection rates.  I started writing the post and realized that it requires a good bit of background knowledge.  This is the first post in a series of posts aimed at providing that knowledge so we can get to Markov Models. The series will be loosely structured along the following lines.

1. Translating Raw Sales Data Into a Cash Flow Model (this post!)
2. An Overview of Sales Data and Markov Models: Finding Transition Probabilities
3. Using Markov Models To Estimate Accounts Receivable and Cash Collections

Hypothetical Raw Sales Data

I created the data set below in Excel.  There are four columns: Invoice #, Sale Month, Sale Amount, Month Collected, Months Since Sale.  The Month Collected field tells us the month in which the cash from the sale was actually received.  The Months Since Sale tells us how many months it took to collect the cash.  This sales data covers a period of 4 months and has 4 sales per month.  Let’s pretend we took the data from XYZ Company, Inc.

Next Page

smbZen is a service by SMBZEN, LLC
Copyright © SMBZEN, LLC All Rights Reserved | info (at) smbzen (dot) com