I usually get a chance to talk to few new product managers and I am always surprised how little value they give to statistics in general. Yes doing regression analysis on variables may not sound cool but if you master a few techniques you can actually go far. In this blog I document 5 areas where a little bit of statistics can help
If you are product marketer or product manager this is one activity you have to do. Segmentation is critical activity even in the context of a startup. If you need to create a new niche in the marketplace or focus on a particular type of client archetype or experience this is key. One statistical technique you can use effectively is regression analysis to see which independent variables influence the dependent variable. If you don’t know what I am talking about I would recommend the following books as a great primer on statistics
Heads first Statistics or Statistics in a nutshell both book happen to be from O’Reilly media because they are actually useful books
Every once in a while you are asked what is value or how do you know what is valuable to your client base. In statistics there is a tool you can use called Conjoint Analysis. Conjoint analysis let you look at different aspects or features and figure out how to maximize and identify the right features and function to deliver by looking at the data (which you should have after meeting your clients). One of the best explanations of Conjoint Analysis is given in the book Marketing Metrics. Conjoint Analysis is very powerful tool and can also give you broad insight into managing your requirements better for the various products you bring to market
Trends are the anathema of product managers especially if the trend has already taken a foothold in the marketplace. Obviously the most simple way to spot a trend is a to plot on a graph and see the trend (if you are doing that, it means the trend has already taken hold and you are late in the game). In order to stay ahead of trend and if you are constantly engaging with your clients, you should be able to see what variables are important to them. There are some tools like binomial distribution that can help in identify a trend manifesting in a sample survey you with your client base. Binomial distribution can always provide a good proxy for a full blown research effort but they can offer a quick an dirty way to get an idea of what is going on.
Client Satisfaction analysis
I am big fan of Pareto Charts or most commonly known at the 80/20 rule. You can identify the top issues that matter to clients and focus your efforts in remediating those top concerns in your overall user experience. There are plenty of credible examples of how to develop a pareto chart but the simplest explanation that I have read is on a blog written by blogger Duncan Haughey. Please check out his blog at the following URL: http://goo.gl/4E2VV
You can contact me @ kkanakas on twitter with your comments