Pricing Theory - What works, what doesn't, and how to get the most profit?

I've been doing some pricing research and came across this long, but very funny article about pricing. One of the closing quotes was this:

"The more you learn about pricing, the less you seem to know." And true it is! :)

Here's the full article if you're interested in a laugh and some really good economic / marketing education & theory.

http://www.joelonsoftware.com/articles/CamelsandRubberDuckies.html

I've practised in pricing since the 1980s and consulted on pricing for over a decade. So I'm pretty well-qualified to tell you that:

1. Most of the pricing theory you'll read is of little practical value to you in typical web-based services.

2. Lots of the practical advice you'll get about selecting the right price level will be worth considering, but dangerous to depend on, especially if you buy into the underlying theory.

3. Almost all of consumer pricing theory in terms of pricing structures and tactics (as opposed to overall pricing levels) is of direct relevance, but there's no text I know that gathers these up and is geared to a tech audience.

Let's look at some key concepts.

The demand curve and price elasticity of demand

If you want to read just one sentence, here it is: Probably save your time: use the simple and actionable insight that people reach for pricing comparisons first.

Here's the optional, longer bit.

These concepts, applied effectively, remain pivotal in goods and human services industries. But the body of observation and the build-up of theory has virtually no application in typical web services.

The internet is the nearest thing we've ever had to a perfect market - a place where buyers and sellers can gather and review one another's offers to trade (this is an informal definition!).

Over time, a market like this tends to create two types of seller in particular.

First is the commodity. In commodities (or, which tends to be most relevant, for the commodity portion of a total purchase), pricing over time tends to long run incremental cost of the service delivery. Which for many internet propositions is the sum of a small number of vanishingly small dollar sums.

The cost of promotion can be near zero (for passive promotion) or vast (for buying eyeballs to your tiny island in the vast ocean). This encourages waves of innovation and consolidation, in which availability of capital, celebrity and serendipity all play their part.

If you're in a commodity market, your pricing will probably be 95% driven by comparable propositions - that is, alternatives that are visible, credible and desirable to your audience. That also means that selection of and route to audience also virtually dictates your pricing.

If you're in this situation, ideas of elasticity - maybe a higher price will make you more money, maybe a lower price will - are interesting but in terms of formulae have virtually zero predictive force, and typically given the price 'band' you're in they often also have virtually zero applicability - it's actually pretty difficult and possibly dumb to take your $10 monthly fee 2% higher, for instance.

Second is the 'monopoly' (actually I'm including situations of stable competition, and the terminology isn't ideal, but let's not get too technical). I'll say 'monopoly' to remind you and me that the examples I talk about are almost certainly neither legal or technical monopolies, though they generally enjoy some kind of local dominance.

Offerings with genuine and irreplaceable value tend in general to sustainably high prices that maximise short-term profit. But in specific cases the culture of an organisation, external intervention and routine play-out of business strategy make other pricings entirely possible. So for instance a 'monopoly' may use sustainably low prices.

On the Internet, some 'monopolies' look stable (which means no more than there's no reason to expect a change in status over the next three years).

I'd cite Google and facebook as pretty obvious examples of companies that almost certainly would fail legal tests for monopoly but fit this broad category.

Google is a typical high price 'monopoly-like' - it's built a commercial advertising platform that appears to successfully maximise both price (pricing is locally high) and profit (pricing is almost fully variable in respect of submarkets - for instance, advertising customer segments; it's essentially unrestricted; the price-proportional element of cost is in most cases trivial).

If you become a 'monopoly,' you can pretty much choose your pricing strategy. Or you can bounce around on a whim.

But the predominant experience in cloud services is that 'monopoly' is unstable: companies enjoy the pricing advantage for a limited period but innovation erodes the core value proposition.

And because companies are different and access is easy, if you can describe a strategy for companies who are enjoying or used to enjoy 'monopoly,' someone's playing that out. So I think I'd argue that Salesforce is a former 'monopoly' with a present apparent strategy of ignoring the word 'former.' If so, that's sustained by both actual customer lock-in through process dependence, general inertia in terms of the established base and the power and persistence of reputation - along with some level of ongoing innovation.

So this is terribly interesting. But chances are, if you're reading this you're not a monopoly and in most cases if your business plan depends on you becoming one then the key issue for you isn't pricing but pitching.

Wrapping up on the demand curve et al

OK, so I've simplified. There's only one material place that's obviously misleading. And it is quite an important one. So let's look at the question that reveals it.

Who's my audience?

Techies are techies. They often create stuff because it seems cool, maybe with an idea of who'll actually buy it, but often with no more than an idea. So they put out it into the wild. Some people connect, some people don't. So various things get changed - how to promote it, how to present it, how to price it, what it does, and so on.

Sometimes the audience is the constant. But very often, discovering the audience is the big step forward.

Now, different audiences expect / will pay / may require different prices. So what's going on is a kind of random walk.

Two bits of terminology come in now. There's the pivot, and there's split testing / A-B testing.

A pivot in my view is a substantial change in any one (or more) of the dimensions of the proposition. It's part of the random walk. If the direction you're going in isn't working out, you change direction and see if the experience gets better.

Split testing is a methodology for trying multiple changes simultaneously. Most times you hear the term, the context is a website, where different copy or process flow or styling or pricing is being checked out. And usually the assumption is that you're going to look for the best outcome and go with that.

But split testing doesn't have to be just about those details; nor is it necessary to look only at a couple of outcomes; nor is it necessary in all cases to commit to one path and abandon the other.

What's this got to do with pricing?

Well, back to my minimalist point - your proposition's pricing level is going to be highly constrained by what your audience view as comparable substitutes. But different audiences will view different groups of propositions as substitutes, so there almost certainly are multiple answers to the question, "where (roughly) should I pitch my pricing?"

So, if you've followed carefully, you're going to notice one thing. That is, your proposition at a different price level may become better tuned to another audience.

Now, sometimes that's as simple as saying, you need a couple of pricing choices, tuned to those different audiences. And if you have scale, that may be your best answer.

But if you're looking towards scale, the best answer may well be that you need to align to the the two audiences with propositions that differ by more than just price - and that may carry different brands.

And now you're in an area where your best education is to read up on the last thirty years of consumer marketing. Because pretty much no-one in the web startup domain is saying this, even though it's not just true, but trivially true.

For your startup, you just know that the most insane thing you could do is to try and run two, five, twenty propositions at once. How could that work? Maybe for someone, but no way for you.

I mean, that's about as crazy as, oh, I don't know, StackOverflow taking its core tech and folding it into a bunch of other markets.

It's a big field. There are people that make a career out of studying pricing.

I wrote a rather long and detailed blog post explaining the basics of pricing which may help get you started.

Why not cut to the chase? Add up cost to produce X thingies and then add 1 year of holding and packaging and administration etc., then add your target price needed for selling all out? So, $20 for product $1 to hold it, $7 to package and send it and 3 for admin = $31. Add your computation from dividing the 1,000 thingies into what you want for profit with the $31 and walla - the price per unit. If this is a really expensive item to make then you may need to go gaga-college level to attract capital. If not then read on....

Building 5,000 thingies for sale would drop the overall costs but your close. Try the sale in a small region for the first 1,000 and then expand as you qualify your expectations. I would temper your price in anticipation of doing 10 times as much to get a true gauge of volume (the demand curve). Lots of buz words out there to erode your self-confidence to hire the sheep skins, but rest assured, if its a good deal it will sell. Just worry about how to get the idea across clearly to the target demographic that you too should have a good feel for.

Good Luck, Ebey

Check out everything Keith mentioned first. Then look up A/B testing. You can run an A/B test to see which product price generates the most profit, but note the process takes a fair amount of time to complete and setting up a fair test can be quite complicated.

I recommend RONALD J. BAKER's book PRICING ON PURPOSE - CREATING AND CAPTURING VALUE. Fantastic reading.

You might have read this already, but if you haven't, check out Neil Davidson's free eBook on software pricing. It's a great resource. Neil briefly discusses the demand curve in the book.