I’m often involved in pricing discussions, which are typically introduced as “what’s the right price for my product?”  Much more important is the strategy that should precede this question, namely “what is the right pricing unit for my product and my market?”

An example: Imagine that we’ve seen the success of various network appliances — firewalls, email servers, VPNs — and want to capture the next appliance market.  Our technical team has built a gadget that corrects grammar in emails and memos as they are sent out, making all of our customers sound well educated.  How will we price it?  More specifically, what’s the right pricing metric?

There are lots of alternatives… we could sell an appliance based on its raw throughput (5M bits/sec of email, or 50M bits/sec).  We could also imagine setting prices by the document, per word, per user, or as a monthly lease.  In a twist on “paying for performance,” we might give away the service and charge only for the corrections that are made — careful, well-schooled users pay nearly nothing.

Which Should I Pick?

A priori, none of these is necessarily better than another.  We have to match them against our target customers.  What do our customers need, and how will they want to buy?

  • Most corporate software buyers are used to per-user pricing for desktop applications like virus protection and MS Office.  Our easiest selling strategy for these mainstream businesses could be a per-seat price, upgrading to whatever size appliance is needed to handle the load.  Hot swap units seem like an obvious upsell to this audience.
  • Large insurance companies will need huge stacks of memos processed, so are likely to buy appliances outright — and will need a throughput pricing model in documents-per-hour.  They will also want to lock down the style options (“long-winded, complex, obscure”) for a consistent tone.
  • An online proofreading service for university students probably wants a transaction model: charging students for each essay turned from straw into gold.  If we want to sell into this ASP market, a per-word or per-page charge would match their revenue with our pricing.  Providing word counts and document tracking would also make their billing processes easier.
  • ISPs and mail hosting providers could offer a “grammar check” option for a few additional dollars per month per subscriber.  Knowing how risk-averse ISPs are, we might offer a revenue-sharing model and take half of the ongoing subscription.
  • Corporate HR departments might want to buy this for hand-selected management trainees who dress well but are functionally illiterate.  The number of people in such a program would be secret, so selling this “per seat” or “per user” won’t work.  Maybe a monthly lease could easily be hidden in an “education for excellence” budget.  An upcharge to turn off logging and IP address capture also comes to mind.

You get the point.  Our pricing unit depends on whom we sell to, how our product will be used, and how it will be paid for.  Unique customer groups may each have their own natural unit of pricing.  Parenthetically, as we think about how our customers will use the new grammar appliance, we see that every target market will also need some unique features.

Let’s Do Them All!

What about offering nine kinds of pricing, each in five sizes, to cover all possible buyers?  Ouch!  You should expect to see at least three kinds of problems:

1.  Most buyers can’t handle pricing complexity.  If you give them too many choices, they will panic.  Some may suffer through your sales process, while others vanish forever.  Remember that your sales people are also human: their skill is selling, not navigating bizarre spreadsheets.

2.  Some buyers are very clever.  Presented with many pricing plans, they can figure out which will save them the most.  Academics call this “adverse selection” but you can call it “lost revenue.”  Smart corporate purchasers love to work each of your price plans until they find the one that’s nearly free.

3.  Targeted discounting doesn’t stay targeted.  I’ve seen companies introduce education-only pricing (or “special one-time VAR discounts” or cheaper consumer-labeled versions) while keeping corporate pricing high.  Your sales channel won’t ask buyers for their student IDs, and the market will quickly figure out how to qualify for your limited offer.

In short, pricing units are one part of fitting a product into a target segment.  Just as with features and support options, a pricing model must be imagined in a specific customer context. Thoughtful pricing isn’t generic, but market-specific.  Innovative pricing models can open up new segments if the accepted approach creates barriers to buying.


Picking the units for your price plan is strategic, and depends heavily on your target customer.  It’s also part of your feature-function mix, since different segments will need unique features from your product.