Pricing research tips
Pricing research brings with it a number of potential pitfalls for the unwary. Even within a business, price research can mean different things to different people. For instance, at a seminar we ran for marketers, price was often talked about in terms of product positioning. Among financial directors they'd see pricing to be about income, profits and return on investment. Below are some hints and tips based on questions and queries we've received.
Can you do qualitative pricing research?
It is possible to investigate individual's feelings and perceptions towards prices in a qualitative research setting in order to establish a basic sense of what customer's think is fair and to look at the different mechanisms for pricing. But for the most part qualitative research is not a good way to study pricing. The purpose of pricing research is to optimise the price and not to maximise market share. That is for some customers the price will be too high - you need to know how many. And in general for most products, customers simply don't like higher prices, so the results from qualitative research tend to be relatively predictable while not providing the hard measures of demand needed to assess what buying rate would happen at each price point.
Revenue or profit optimisation?
There is a general assumption that optimising prices to maximise revenue (eg looking at price elasticity) gives the same answer as looking to optimise prices according to profitability. It can be true, but often there is a difference between the revenue optimisation point and the profit optimisation point. The profit optimisation point is almost always higher. To give a trivial illustration, anyone can maximise revenue selling $5 bills at $4.99 - but I only need to sell one $5 bill at $5.01 to make a profit. So if you are looking at prices it's very important to try to include costs - both fixed and variable costs if possible - when modelling. It's also true there are situations where revenue or profit optimisation are not the right measure - strategically the business might not want to maximise revenue or short-term profit - for instance to gain market share or to seed a market such as selling ink-jet printers at low prices in order to earn more money later on the inks.
Frequency and volume effects
Most consumer research looks at counting the number of buyers, so demand is measured in terms of number of heads and implicitly assumes that each individual buys the same volume in a given time. This isn't necessarily the case. If prices increase it is possible that the individual's mitigate the price rise by choosing to purchase less volume, or to buy less often. For instance if the price of newspapers rise, readers may decide not to buy every issue. And in business markets, if a customer stops buying on price this can be a large potential revenue because of the volume involved (this is one reason why business markets are much more negotiable and open than consumer markets). Care is therefore needed to factor in volume effects where possible.
Competitive versus uncompetitive pricing
Some of the basic forms of pricing research (Gabor Granger and van Westendorp) do not include any competition. They simply ask for views of certain prices or price points. The problem is that though the research point to a certain optimum price, in practice the ability to set that price depends on how the competition react. In a market for A4 paper for instance where there is little differentiation, there might be an optimum price available at £15, but if competitors are offering the product at $10 it doesn't matter if customer's would pay more. The price itself has to be competitive, so it's possible that not all the value is available unless better brand-differentiation can be established.
Bundled versus unbundled pricing
For many offers such as clubs or subscription-based products, the company can choose to bundle products or offers hoping to achieve a greater take up, or to deliver higher value. The difficulty in establishing a price for a bundle is that for larger bundles the consumer has a tendency to simplify the offer and only to look at the core parts that seem to provide the most immediate value. Including more elements is therefore not necessarily directly additive in value terms - that is add a new item the customer values at $2 to the bundle, doesn't mean the value price for the bundle goes up by $2. So in terms of research the value of each item discovered individually need not be the same as the total of all the items when put into a group. Understanding bundling group size and issues of offer complexity are also important. In some instances, make-weight offers (eg 20% off vouchers) can even devalue the offer.
In transport economics, yield management refers to the division of journeys into different ticketing groups - for instance tickets for those who book a long way in advance are often cheaper than tickets for those who book on the day. Yield management relies on modelling against the demand curve to make the optimum set of pricing 'buckets' - a few cheaper seats for the fastest bookings, to very expensive seats for those with no travel flexibility.
Real price testing
One of the things that can be seen in real prices is that the product itself is not the only source of price differential. A can of Pepsi may have many differing price points in a small geographic area - different in a small shop to a supermarket, or different in a bar compared to a restaurant - even varying by a factor of 4 or 5 times the lowest available price. Some components of price are therefore impossible to test except in the real world (the value of convenience in this case). So in general, where ever possible we recommend some form of real price testing. Surveys and what people say they will do are one thing, but what they actually do in practice when there is real money involved is another. Companies often have clues as to price sensitivity through the offers and promotions that they run, or by observing local price behaviour - for instance where different retailers set prices at different levels. Pricing can also be testing via couponing, or via online adaptive pricing - though this needs to be fair and honest, otherwise an individual customer might see several different prices for the same product.
For help and advice on carrying out pricing research and setting pricing strategies contact firstname.lastname@example.org