Pricing, metrics and evolution (#3)

9 07 2009

So far I’ve talked about how pricing is a mix of several elements which involve math, science, intuition and emotional intelligence.  At some point, these things come together and you have to divide up what you’re selling and assign a suggested quantity of money that should be exchanged for each quantity of your product.

For some products, sizing or metrics may be determined by the market standards (e.g. existing products, packaging, shelf space) or by some other decision designed to garner a net impression (e.g. 100 calorie snack packs of food or beverage).  Consumer packaged goods have a whole pricing paradigm unto themselves versus pricing a “virtual” product or service.

I’ve heard that individuals will typically undervalue their own skills or services when they ask for compensation versus letting the recipient pay them.  However, I don’t believe that’s true for business transactions.  Especially when it comes to software, cable TV, and cell phone plans. 

Certainly in the software sector, many customers believe that the pricing is out of line with the value they receive.  Perhaps that’s because the licensing terms don’t match their expectations very well.

Metrics matter

I believe one of the trickier parts of software licensing is actually the decision on what metric to use and the “break points” that reward a consumer for buying in bulk.  There are several accepted metrics out there, and also a few of failed ones to learn from.

Most users are familiar with pricing per user, per seat, or per CPU.  But if you don’t look ahead at the hardware and software market, you may find yourself backed into a corner as evolution takes its inexorable course.  Not only will Moore’s law increase computing power, its implementation will probably break your licensing algorithm.

Recently, licensing per CPU was considered a solid metric.  Adding processors most likely meant a user intended to scale up their software installation. 

In the bad old days where there were more CPU architectures than was safe or sane, Oracle attempted to level the metric playing field by introducing an artificial assessment of a processor’s power called the “Universal Power Unit”.  This was an attempt to account for the fact that the DEC Alpha, Intel x86, IBM PowerPC, and Sun’s SPARC all had different computing power per CPU.

Needless to say, this failed.  In the benchmark wars between AMD and Intel in the 1990’s, it became fairly well established that benchmarks tested inconsistent attributes across hardware implementations of the same architecture let alone across different CPU architectures.  So, while the UPU might have been a scientifically sound metric, it was doomed to failure by reality and the customer perception that they were being taxed for buying quality hardware that would prorogue their upgrade cycle.

And now, per-CPU pricing is ambiguous.  With CPU vendors putting multiple CPU cores on a single die, what constitutes a “CPU” now?  How do you count virtual CPUs?  Mobile phones can even have a dual core chip in them. 

When multi-core CPU’s were introduced, Microsoft stated that a CPU was a per-die decision – a single CPU package, no matter how many cores, is a single CPU (as did IBM).  Oracle disagreed and created a “processor factor” to account for the multi-core capability.

There are few immutable metrics.  Even per-user metrics will get challenged by innovative customers asking for a refund if they go through a big layoff or employ lots of contractors. 

It’s a good policy to ascertain what customers understand about their current purchasing behavior and their business.

Evolving metrics

New technology is blessed and cursed by being able to establish a unique metric for a licensing basis (if they choose to).  The cost of that flexibility is the curse that technology tends to evolve to a point of commoditization in order to make room for the next innovation.  Devices tend toward ubiquity, software tends toward becoming a platform.

For example, text message pricing really felt like the telegram’s pricing model.  Per word pricing might have actually been more fair for text messaging when all you may want to send someone is “k” in reply to their message.  Now you can get infinite messaging plans (which are still too expensive).

Land line phone companies have seen their entire metric strategy devoured by the more rapidly moving mobile phone market (it’s not just pricing at play there).  Voice over IP technologies have set up the destruction of the notion of long distance land lines and have set businesses up for all sorts of new communication options.

Eventually, your own product may evolve (through development or acquisition) in a way that destroys your own metrics.  By keeping your own product’s roadmap, company strategy, and the buyer’s environment in mind you can sometimes foresee a radical shift in the making.

Questions to assess

It feels like there is a lot more to talk about with respect to metrics.  So, let’s think about framing a discussion around establishing metrics.  Good questions to consider are:

  • Is your product dependent on another product that must be licensed?  If so, what is that licensing basis?
  • Is your licensing metric accepted in the industry?  Explain it to a sales person and see if they can explain it back to you.  If not, try again.
  • Understand the customer acquisition chain.  The user, decision maker, and buyer may all have separate understandings of acceptable licensing metrics.  For instance, a user might consider “per device” pricing as anything with a MAC address, where a buyer or lawyer might think “per device” is per physical asset.
  • Speaking of lawyers, you absolutely must sit with your contracts personnel and explicitly define what you mean with your licensing metric.  If it can’t be rigorously stated in a contract, you can’t recognize revenue, you’ll extend your sales cycle and become the “order prevention department”.
  • Do you need to institute a new metric?  Do you have a transition plan for existing users?  Do you know how this impacts your CRM and order processing system?  A new packaging isn’t ready until you can book an order.
  • If you can, look at your direct competitors in the media and from trusted sources to assess their product’s trajectory.  If you see something disruptive like “all you can eat” licensing or a radical shift to a new paradigm (e.g. from on-premise to SaaS), you should perform some analysis on whether or how you need to respond.

We’ll talk about some of the issues surrounding selecting break points in your licensing scheme in the next installment.  Stay tuned!



2 responses

10 07 2009

“Explain it to a sales person and see if they can explain it back to you. If not, try again.”

Does this mean try a new pricing structure, or get a new salesman?

Actually, great stuff, keep it up. Do you remember “Measurement points” pricing for Sigmafine? How about “Equivalent Distillation Capacity” for refineries?

31 07 2009
Pricing and the metrics that matter « Spackle

[…] and the metrics that matter 31 07 2009 Previously, we engaged the topic of the metrics you might choose as the basis of your pricing and how the evolution of technology will disrupt your well-laid plans.  This is especially […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: