Pipeline Management - Sensitivity Analysis

October 16, 2006 · · Posted by Jordan Frank

In the management of a business obstacles may arise at any moment. One of the better lessons I learned at business school (the one on the other side of the river) is to recognize that the past does not predict the future (I have also learned that in a casino or two). A related lesson was to do sensitivity analyses on our data models and business plans.

VC Confidential offers a really good take, and working model, for pipeline management. He makes two key points which we do in practice:

1. Only include deals at the top of the funnel. Anything that is below "developed" (the customer states that your product meets their needs, but the product is not yet selected) is generally too iffy to predict. In his model, each deal has a percentage likelihood of closing in a specific month.

2. Include a handicap. Month by month, record your % miss and apply that moving average as a handicap to projections.

Two further factors, which served us well on the course from early adopter to early market development, should be considered in pipeline management and financial modeling.

1. Likelihood of any given deal closing in any given month. Jason introduced this method when he joined our team, and it has added a great deal of accuracy to our funnel predictions. If a given deal may close with a probability of 60%, the model should indicate a fraction of that 60% across an array of months: 10% in November, 30% in December, 20% in January. Sometimes percentages may stretch out over as much as 8 months (I may close a deal by November which the customer originally said should close last March or April).

2. Plan for droughts and disasters. Most businesses face cycles where revenue dips considerably or growth slows down for a period. Retail, for example, has an obvious spike between Thanksgiving and Christmas. Businesses that sell to the US Government may spike in September and October, at the end and beginning of the fiscal year. If you are planning to close a big deal to a European company in July, plan for it to close in September, as the last approving manager may go on vacation in late July or August. I recall that the Iraq War threw off my projections in 2003, setting government and business purchases off by several months. A lot of variables play into the timing of droughts and disasters.

I have read my share of early stage business plans and have yet to see one suggest the number of months of cushion required to survive droughts, or the variability as to when their sales ramp may occur. These variables should come into play in early stage planning and in every day decisions, as you never know what positive or negative event may lurk around the corner. Every business should be ready with a party for a positive event, and a plan for a negative event.


Page Top