Wednesday, July 6, 2011

Product Experiments

Lately I've been thinking about business plans and product launches and how product managers can validate many of the assumptions that are made about their products. It is rare that a product manager can really know/survey potential customers before a product is launched or before making a major product decision. We all make some assumptions, is there a way to minimize the risk that we are wrong?

Some typical product assumptions:
  • Price elasticity of demand - This boils down to how price impacts demand. Should you go with a low price strategy or justify a higher price since you have a differentiated product. How will changing the price up or down by 10% impact quantity demanded?
  • Market Size/Potential - We can always assume that everyone is interested in our product, but sometimes those assumptions aren't true. This is especially difficult to assess when your market size is dependent on the adoption of a broader platform on which your product is built, etc.
  • Most Important Features - Often times product managers think they know which features will be used most often. The surprising thing to consider is how the feature is used with other features or what other features are being used to an increasing degree.
So how can a product manager mitigate the risk that one or more of his assumptions will prevent the product from living up to its full potential? Here are some ideas:
  • Experiment with pricing - One way to do this is to conduct a short pricing experiment where the price is increased by X% for 1 month and then decreased by 10% the next month. Compare and contrast the demand changes, if any. Another idea is to offer every other customer a different price (if possible) as this will remove the seasonality component of demand changes.
  • Try a Beta - Start with a beta and learn from that and then launch the real thing. This is a common method to help refine the feature while minimizing the likelihood of releasing something that is not very polished. You can even turn your "beta" into a special event by limiting who can participate and making those that you select feel special by giving them a sneak peek.
  • Offer the feature to your high end customers - Perhaps you have different versions of your product (e.g. Basic, Pro, Premium). Offer the feature to the premium customers as a sort of beta and then once the kinks have been worked out (and you have hopefully derived some additional revenue). For example, you could offer a feature in the high end version of your product and use those higher margins to recoup your expenses in building the feature. Once you have recouped the costs, consider moving it to the lower end version of your product to increase adoption and differentiation in that version of your product.
  • Include a Feedback button - Be prepared to get some constructive criticism :). I would recommend personally replying to every single one of them to encourage a dialogue.
Oh well...just some thoughts for today....I'm already thinking about my next blog post so that's probably a sign that I've exhausted my thoughts on this subject :)