I have data from a market research survey which covers behaviour of different customer groups. To be specific, I am looking at the number of channels that a customer will visit before they buy a product.

I would like to know whether the value of the customer (as measured by the value of the product they buy) is a meaningful way of splitting the data. I have calculated the average number of channels visited for the whole sample (let's call it X) and then for customers who bought only the most expensive products (Y) and separately for customers who bought only the cheapest products (Z).

I would now like to check whether Y and Z are statistically significant. i.e. what is the probability that in my whole population I would find a customer who has bought a cheap product having visited Y or more channels beforehand.

There are a number of statistical functions in Excel that look relevant (CONFIDENCE.NORM, CONFIDENCE.T, T.TEST, etc...). Which one should I use?

  • $\begingroup$ This may get a better response on the statistics SE site, which is more appropriate (SO is primarily for programming, algorithms, and the like). If you flag your question and ask for it to be moved, it will likely be accommodated. $\endgroup$
    – Iterator
    Nov 15 '11 at 23:49

Someone else will probably answer the precise question that you ask, but you might learn more if you explain what deeper questions you are interested in and get feedback on the best way to answer those questions.

I recommend graphing your data. There are many ways to do this. Perhaps a scatterplot of sales on the y-axis and number of channels visited on the x-axis... Or, similar idea, boxplots representing sales on the Y-axis separately for each number of channels a customer visits (not sure if excel can do this).

Or, I'm sure excel can do this, separate histograms of channels visited for each of your three categories of size of sales. Or reverse it, histograms of size of sales.

These visuals will complement, and probably be more important than, any statistical tests that you perform.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy