Measuring Price Elasticity and More | Hacker News Reader

Measuring Price Elasticity and More

avc.com

21 pointspiyushmakhija10 comments

I worked for a product strategy consultant and one of his tools was a detailed chart of the price elasticity of many everyday common objects with very high market penetration.

Pens, for example, are available at basically every price point, and everyone uses them. With a lot homework he was able to define the price elasticity curve for pens. Same with TVs, refrigerators, pads of paper, cars, computers, washing machines, watches, etc.

The key was to treat these curves as the likely outside edges of the envelope for new, similar products.

For example, comparing a prospective pen-based computing system to the elasticity curves for pens, pads of paper, and computers would give us a hint as to the best possible likely unit sales of a pen computing system, based on the predicted price.

He also built a combined curve of all products, which defined the outside edge of price elasticity for the U.S. market in general. We all know that a car costing $100,000 is not going to sell 50 million units a year in the U.S. But what is the maximum volume that any product costing $100,000 can expect to sell? There is an answer! Or at least there was. Sadly, I don't have those documents handy right now.

Then we could plot individual products into that general elasticity envelope. Products closer to the edge were incredibly successful; products deeper in the envelope were not as successful and might be more ripe for competition.

Super interesting! Can you share the name of the consultant or advise or where to read a similar report? I'd love to create a similar graph.

The name of the company was 4th Wave, it was basically one guy with an EE PhD and a couple of young helpers (that was me for a couple years).

I don't know if he is still in business; this is still basically the same crappy website I built for him in 2002:

http://fourthwave.com/

We are launching a preorder campaing shortly, and I just don't see how I could do this without upsetting people. In our press releases, emails and other places the pricing is clearly laid out. If you are selling a commidity on amazon, this could work, but I think it is easy to ignore the risk to your brand/integrity if you sell something to one person for 100 and the next person gets 60...

would it be better to start off selling at cost or lowest acceptable margin and increasing price? having an extended period of time early on with low sales could be pretty harmful, no?

I don't think the intention is to run this test for "an extended period of time." You want to gather enough data for it to be significant and then move on to the next price you're testing.

But, it's a very good question about starting high or low. The common wisdom is to always price high and then move lower. I've seen this work well for sales by being able to offer a discount to close a deal. There's also the perception that expensive = high quality. The only real way to know for sure, however, would be to duplicate the test and go high-to-low on one and low-to-high on the other.

But then you never discover what the highest price those customers would have paid. One of those $1 sales might have been a $5 sale, but now the customer gets the surplus value.

You need a lot of volume to measure this.

For a new company/product I think it's better to start low and then increase the price if/when you get more business.

Does this work for services too?

Wouldn't you upset long term customers?

You let old customers stay on their current rate. Then they will think they got a good deal. But of course it works best on products with a lot of growth, making buyers think it's good to buy in now and not later when it's more expensive.

Reply on news.ycombinator.com