Monday, June 4, 2012

Rewarding IT Efficiency


IT Efficiency

Do IT People have incentives to be efficient? I argue not!
Something doesn’t add up. The primary driver for costing IT services is cost per person per hour - a classic approach for pricing services. If your IT Company is contracted to provide a solution, then you have several options for pricing the service. The most popular option by far is to work out how many hours you estimate your IT people will take to deliver the solution as required. Your people will probably include Project Managers, Designers, Implementers, Testers, etc.. The same applies for pricing operations and maintenance services. In this case you just price people with a different capability.
The above approach applies for most types of engagement. Even if a solution’s delivery and deployment is contracted for a fixed fee, the company still price it per resource per hour, and then add contingency reserves for rework and errors in estimation.
However, we all know that pricing and delivering IT services and solutions are not that simple, and while we try our best to estimate the amount of time we will spend on providing a service or solution, we are by and large usually wrong. 
We are mostly wrong for various reasons. Primary of these reasons is that we have great difficulty to repeat consistently the same processes over and over for different services or solutions. We, in the IT services market are not consistent in the way we provide IT services. And, the way we price it does not contribute to us getting more efficient at doing it.
Pricing “Standard” Services
Let me explain: When you go to a doctor for a consultation, he has a set procedure he follows for examining and diagnosing any ailment that plagues you. The same goes for a visit to the hairdresser. A very similar procedure repeats itself customer after customer, depending on the treatment you buy. 
However, when we get to IT and Lawyers, it is a different story. Here you are bargaining based on the Professional’s capability, and set and repeatable processes are very scarce. The lawyer has agreed hourly tariffs depending on industry based pricing, or demand for his services. However, their bill is usually a surprise, and an unwelcome one at that. because with the best will in the world, he can not estimate accurately upfront how much time he may spend on your particular case. 
The Lawyer mostly does not have a set procedure for his services. When they do have one, like the transfer or conveyancing of a property or another simple transaction, then they are quite happy to give you a fixed fee (priced on his cost for an assistant per hour, to do the job on his behalf, and for him to sign the documentation). 
Similarly in IT we have been striving to get some level of process definition and repeatability for simple transactions, but when it gets to the detailed step by step execution, repeatability, optimization, and process maturity we fall very short. The CMM movement is an attempt to rectify this problem. Alas, it has turned to just another tick in the IT box, and very few organizations has actually and materially improved the way they do things.
We mostly rely on the professional capability of the individual IT person to lead us in the cost and effort for the task at hand. And this is exactly where we tend to get it wrong. Most IT Professionals will have a hard time to give you repeatable process steps and exact estimates for their work. I have made the case before that IT People is not given the opportunity to specialize in a role and therefore they know a little about a lot, instead of a lot about something specific.
IT People are not like Lawyers or Doctors who mostly can function on their own (with some lowly paid assistants by their side). IT People are dependent on other IT People, and also on their client’s ability to deliver. Therefore, it is even harder to coordinate or cooperate for standard, repeatable process steps, and not even mentioning the difficulty of pricing accordingly. The success of the outcome is therefore dependent on the success of the team. This creates a much more complex challenge similar to that which sports teams face. The advantage for sports teams however, is that their games are considerably simpler to master, and yet it takes lifetimes of devotion to get it right.
Payment for Inefficiency
In IT we have customers that are expecting a level of efficiency that confirms that they are getting our services at a fair and competitive price. Let’s have a closer look.
Like the Lawyer and Doctor, good IT Professionals costs a lot of money. Even more problematic is the fact that “scarcity” created by the need for specialized tool knowledge, adds to this price pressure in the market. Secondly, unlike Lawyers and Doctors, there is no industry body that establishes an verifies a baseline competence based on apprenticeships and examinations. So it IT you can end up paying big money for someone you think is good at what they do, and somehow not be sure that he or she is actually what they claim to be.
Now, you have your highly paid (and hopefully competent) IT Professional, but remember he is only as good as the rest of the team (and the customer) allows him to be. The next challenge therefore is ensuring that the rest of the team has a highly Professional Competence to support your highly paid Professional. A weak link in the chain can cost you dearly.
Once you have your team of highly paid IT Professionals you are challenged with organizing them. If you ever tried to organize a group of Doctors or Lawyers, you will realize that it is the hardest groups of people to get to cooperate and collaborate in a team. Similar challenges face you in a highly competent IT services team. Each one is an “artist” and they tend to be frustrated by others’ “incompetence” or lack of professional make-up. You therefore have a group of highly paid professionals, who think they are highly competent, and they are difficult to organize for a successful outcome. 
Lastly, these professionals are so competent and smart that they are very clear on why delivery do not take place, why it is complicated, difficult, problematic, a lot of work, instead of finding reasons for actually making it happen. From the customer’s point of view, you have a wall of wisdom that bleeds your budget dry hour after hour while they invent and re-invent the next generation mousetrap, the way they see it.
The alternative is to use lower paid, “less competent” professionals, and force them to comply with processes that are well defined and repeatable (not necessarily optimized and efficient). Furthermore, you can ensure that you have a clear division of labour and that each professional specialize to innovate and optimize their small part they play in delivering a basic service outcome. When you do this, you are in essence creating a factory or production line, and the competence requirement of the individual is gradually overtaken by the efficiency of the process, steps and accompanying tools.
Only then can IT service have any chance of improving our pricing approach and win over customers’ trust that we will actually be able to deliver what we promise - because we’ve done it all before (many times over).
One Last Example
The dilemma we face is not unlike the dilemma faced by the motor engineering and manufacturing industry at the turn of the previous century. 
Before Henry Ford established the production line to mass produce standard automobiles, at a low cost with unsophisticated labour, cars were built differently. 
It was custom built, one by one, by highly competent engineers and craftsmen. These professionals would design, build, and customize vehicles in small teams at an enormously customized fee, working for months on an outcome. And, they could only do a few units at a time at high risk of not meeting budget and quality constraints on each. 
Gradually engineering and manufacturing practices improved, but it was always totally dependent on the capability of the craftsmen and their tools, and the team. This approached prevailed until Henry Ford’s production line took over. Today, there are only a small number of vehicles still produced in this way, and these are some of the most expensive and sophisticated vehicles on the planet, for good reason.
While we keep on providing IT services in the pre-production line approach, our customers can expect highly expensive, custom built, solutions by highly paid IT Professional artisans with suspect capability, almost no repeatable process, and which is difficult to organize for a desired outcome, that is in most cases not that clearly defined. Pricing these services by the time spent on it is the only way to cost it. Needless to say, while the artists are at work your customer wallet will bleed.
What is needed is a Henry Ford style approach to delivering IT services. Then only will the IT industry have a handle on its cost structure, its processes, and its tools and capability required to optimize what we do.
Conclusion
Highly standardized production line delivery may not be the best approach in all situations. Maybe if we continued to build cars the old fashion way, there would have been less on the road, less global warming, and our oil dependency would long time already have been replaced by innovative alternative transport solutions that are simpler, more energy efficient and cheaper to use. 
Production line delivery has a tendency to curb product innovation. Innovation is distracted to optimizing the manufacturing process, instead of devoting it to product improvement. You can have all the software you like for $1,000 as long as it is black with 50 features in Java. Any customization will come at an enormous cost because it will also require production line customization.
Every customer will tell you that their business is different and special, so in IT we are challenged to come up with an approach that use the best of both the production line, and the capability of the artisan. And this is not an easy one.
As long as customers don’t have an alternative (and some would argue that the software factories in India is just that alternative), we will continue to see expensive, custom build, late, complicated solutions build by highly paid “highly” competent individuals that are hard to organize.
Hendrik van Wyk

Google+ Followers