Analysing data -help with a presentation for a job interview

[TheGreatSoprendo]

I posted this in the Finance section but didn't get a response, so thought I'd try here....

I'm going for a job interview on Friday and I've been asked to analyse sets of sales penetration figures and prepare a presentation for the board of directors to explain the data. The data is over a period of 6 months, similar to the following:

	Jan	Feb	Mar	Apr	May	Jun
Product A	69%	71%	66%	67%	70%	68%
Product B	56%	56%	58%	59%	57%	58%
Product C	69%	71%	66%	67%	70%	68%

Now the obvious starting point is to graph the data, which I have already done. However, since the data sets don't neatly rise or fall over the period, but fluctuate up and down seemingly at random, I am wondering whether there is some kind of statistical formula that will allow me to quantify the change in penetration rates over the period?

Any help massively appreciated!

[davepoth]

Yup. Under the "chart" menu, there should be an option for "add trendline". Use a linear plot, and a two month moving average.

-edit-

Actually, don't bother with the moving average, that will make it look odd for six months.

[GadgeS3C]

There are severals stats techniques you could use (control charts for example). But with only 6 months data you haven't really got enough data points to draw any meaningful conclusions.

If it's a sale role then of course all that matter is if the last months number is bigger than the previous months

[Pints]

Depends what the BoD are wanting to know. Go back to basics.
The board will want to make decisions based on your analysis. Make it meaningful to those decisions.

Also, look at it from a quarterly cycle perspective. There is a spike in sales in the 2nd month of each qtr (certainly true for A and C). This could be marketing driven or otherwise. Discuss.

[oldbanger]

broadly speaking sales of A and C go up when B goes down (or vice versa)

why?

[steveatesh]

Not much to work on and really depends upon the job too, but for me they are either wanting trend lines as per the comment above ( trend, flat, up or down, is there any other information to go with this for you to off explanations?) OR they want a Control Chart doing, so you can demonstrate you understand them ( upper and lower control limits, special cause variations, whether the sales penetrations aw in control, anything that is statistically significant or not, etc) have not got a Minitab type app on my iPad so can't check the numbers out for you.

I think a basic trend is more likely than a control chart, but I can't say from the information you have given.good luck with it anyway.

[TheGreatSoprendo]

oldbanger said:

broadly speaking sales of A and C go up when B goes down (or vice versa)

why?

Don't get hung up on the actual figures. The figures I posted aren't the actual ones, but they're representative of the kind of data I have to work with (although I have more data than I posted)

The only reason A and C are the same is because I copied and pasted and then forgot to change the numbers!

I guess my main question was whether there was a statistical formula of some sort that could take account of each of the 6 monthly figures for a given product and give a numerical value to the compound increase or decrease in the penetration over the period?

[TheGreatSoprendo]

steveatesh said:

Not much to work on and really depends upon the job too

It's a marketing role and I don't have any data other than the penetration numbers (these refer to penetration of add ons to the main product, a bit like a car dealer selling things like extended warranty, gap insurance, paintwork treatment with a car). I have a set of these figures for the same products across 4 sites.