# Win/Loss Data Analysis for Pricing

Pragmatic Marketing emphasizes the importance of win/loss analysis. You can learn so much by listening to your buyers, especially right after they have made the decision on what to purchase. Regardless how valuable these win/loss visits are, they are not the topic of this blog.

Today, we are looking at win/loss data, which has nothing to do with visiting buyers after each closed deal. Rather the data typically come from normal operations. For this to be useful, you need to be collecting the right data, constantly.

Every opportunity we quote is a data point. For each quote we want to collect the price quoted and the quantity quoted. Later, you will go back and determine if you won or lost the business as an additional datapoint. If possible, you also want to collect data on who the competitor was and how much they charged. Then collect data on anything that you think might influence your buyers’ purchase decisions. For example, day of week, weather, time of day, where they are in their budget cycle, etc. We never know what is valuable, so the more data we can systematically (and automatically) collect the better our analysis will be.

Once you’ve put programs in place to collect data, then you periodically (quarterly?) want to analyze it. This is easiest if you have a statistician on staff, but if not, here are some hints to get you started.

You are going to run something called logistic regression. All statistics software packages have this as an option (although it is not available in the analysis add-in on excel for the mac). In logistic regression, the dependent variable needs to be a 1 or a 0. Since we want win/loss to be the dependent variable, make every deal you won a 1 and every deal you lost a 0.

Then your independent variables are everything else you’ve collected. When you run the logistic regression function, you are looking for what variables have a statistically significant impact on whether you won or lost. Price should be one of your independent variables and it is almost always statistically significant. If you’re able to include competitor’s price, it too will most likely be significant.

What’s really important here though, is looking for the other variables that impact our win ratio. We know we can lower price, but what else might we be able to do? Maybe we put on big sales pushes on Fridays because that’s the day we win most often. (There is probably a reason if this is true.)

When it comes to price, we want to watch the power of price over time. If our competitor comes up with great new features for their product and we haven’t kept pace, that will show up as us having to discount more in order to win new business.

To be honest, I’m a geek and love statistics. What I described is what I would do. It is also possible to get a huge amount of this understanding using pivot tables and graphs in excel. You are looking for trends. What is correlated with winning? Does this change over time?

There is so much power and information in understanding what helps us win and lose. Regardless of what analytics you choose to use, you have to collect the data before you can do any of this. Start collecting data.