Dear Tig,

How might I go about testing the relative merits of different prices using my site? Is it just as simple as switching the price a few times and measuring the results?

Sincerely, Nervous at the Switch

Dear Switcher,

In principle, it's just that simple, But in practice, there are lots of devils in the details. Here are some issues to think about before you conduct your testing:

• The ultimate goal for a price test might be to see whether or not you can squeeze out an extra dollar or two from each customer without losing too many of them. But if you conduct your test correctly, with enough price points, you might find yourself unlocking the secret to price determination: the “price elasticity of demand.” This elasticity factor is essentially a graph that shows how many customers give up at which price points, allowing you to determine the perfect price in terms of maximizing revenue.
• Price sometimes is a major brand factor. Few people would purchase perfume for three cents an ounce. The very fact that it's expensive gives the fragrance an allure. This economically irrational factor can cause havoc in attempting to interpret an elasticity curve. If your results seem to be influenced by this, consider a brand extension that would allow you to charge at different price points for a substantially similar product.
• When using your site to test pricing many details must be considered. The text on the site must be “cleansed” of references to previous pricing. Advertising and promotions that make specific price offers must be either dropped temporarily or special accommodations made for people expecting these non-test prices. Site visitors who come one day to purchase and then come another day need to see the same price both times for their behavior to be correctly gauged for the purposes of a pricing test, requiring the use of cookies or a similar mechanism.
• A price test must be run long enough to eliminate concerns of random influences, such as time of day, temporary traffic spikes from marketing programs and seasonality. Sometimes the most efficient way to do this is to repeat a price test several times throughout a year.

---------------------------------------

Dear Tig,

It seems as though whenever there's some sort of disaster or war-related news, advertisers make a big deal about pulling their ads. Is this something we should consider doing?

Thank you very much, Still On Air

Dear Still,

For decades, large brands that advertised in categories that could vaguely be classified as disaster-related have pulled their broadcast spots when calamity strikes. Airlines stop advertising around coverage of plane crashes. Cruise lines pull ads when a ferry sinks. These companies have long put such clauses in their media contracts to cover those contingencies.

The number of advertisers and product categories that have joined this practice seems to be growing. When the September 11 events hit, completely irrelevant companies began to ask if it was too crass to advertise at all on programming that was terrorism-related (and, at the time, everything on TV seemed to be terror-related).

The phenomenon seems to have reached the level where its an “in” thing to do, some people even considering it a strange form of showing solidarity with a patriotic cause. I doubt consumers see advertisers' forbearance as such, though. More likely, the audience – if they notice at all – would consider it an act of conceit.

There are extremes, of course. A frequently seen direct response ad on cable TV shows two U.S. vets from Afghanistan endorsing a piece of exercise equipment, complete with martial music and a huge flag in the background. I'd classify this as over-the-top, but I'm not aware of any “normal” advertiser that has suffered as a result of news events overtaking unrelated ad placements.