When it comes to services, pricing is a very sensitive issue. While having a good pricing strategy may not be a competitive advantage, having the wrong price structure can mean disaster. The wrong price, at the very least, means reduced profits, but it may also lead a company into cash flow problems and, eventually, bankruptcy. On the cloud, everything - platform, infrastructure, and software - is sold "as-a-service" -- that is, based on some sort of usage metric. This allows for a much broader range of pricing strategies, letting companies experiment with different models as they go along. At the same time, this broader range of options holds increased risks for mistakes and that's what we all want to avoid.
When pricing goes bad
In his book "Behind the Cloud", Marc Benioff, the CEO of Salesforce.com, discussed how the company had trouble with its pricing model early on. They charged customers on a monthly basis -- exactly what would be expected from a cloud service -- but, at the same time, salespeople received commissions based on the yearly value of the contract. When sales rapidly increased, the commissions to be paid quickly overtook the income, leading to cash flow troubles. The way out was both to change the way commissions were paid and to promote long-term plans where customers would receive a discount if they paid a whole year in advance.
The recent woes of Netflix offer another interesting example of how the wrong pricing structure can hurt a company. Netflix's DVD-by-mail service closely resembles a cloud service, differing only on the delivery mechanism: customers pay a monthly fee that gives them access to movies delivered straight to them, similar to what would happen with a cloud-based application suite. When the company misjudged the perceived values of its services and tried to raise prices, customers defected in droves and the stock price fell dramatically, wiping out shareholder value.
These are cases of pricing going really bad, endangering the company, and they are pretty easy to recognize. The most dangerous case, however, is when pricing seems to be working fine. You may work complex calculations and come up with a price that seems to result in profits, but hides some costs that can ultimately result in losses. If a service is priced based on an expected consumption of resources (CPU, for instance), but the real average is much higher, it can quickly lead to disaster.
What is left for me then?
Not everything is a bad, however. By keeping some tips in mind and taking previous experience into consideration, you can pave your way into profits! Research shows, for instance, that when people pay for services for a long time beforehand (such as paying for a yearly plan), they tend to use the service heavily on the first months, but they tend to gradually abandon that service. When they make frequent payments, however, their usage is more even over time. This means that companies can make use of discount packages not only to promote longer term contracts, but to adjust usage levels as well.
Another thing to keep in mind is that recent research has shown that certain pricing models can achieve much higher profitability than others. When infrastructure is purchased on the cloud, for instance, the way that we are charged is very similar to the way we are charged for mobile phone services: every person pays a fee based on the usage time - unlike with software, where we usually pay a flat monthly fee. On this situation, however, the phone companies are more efficient than cloud providers. They use a package-based model, where the users pay a flat fee for a package that gives them the right to a certain usage level, and have to pay a much larger fee for excess usage.
This model allows for the optimization of both usage and profits. Light users pay a higher usage fee, but can receive some discount since they will probably not exceed their limit. Heavy users, on the other hand, can receive an attractive discount. In both cases, the excess usage fees can easily cover any expenses generated by some client going over the limit. At the same time, the package model is also interesting for clients. It allows for much better planning of expenses over time, which is very important for companies of all sizes.
The cloud offers a lot of room for experimentation with prices, both to providers and consumers, and my recommendation is to try things out and experiment as much as possible with different pricing models to see what fits best. As the examples show us, not thinking about pricing or making incorrect assumptions can result in reduced profits or even major disaster for companies who sell services. And, for everyone who is purchasing a service on the cloud, understanding the ideas behind the pricing model adopted by providers can allow for better usage of the services, or even to negotiating better discounts.
After working for a database company for 8 years, Thoran Rodrigues took the opportunity to open a cloud services company. For two years his company has been providing services for several of the largest e-commerce companies in Brazil, and over this time he had the opportunity to work on large scale projects ranging from data retrieval to high-availability critical services.