In the 90s, I worked for Amdahl, a now-defunct organization that built huge IBM-compatible mainframes and software systems to go along with them. Within fourteen days of my hire, one of our biggest customers came to us with capacity issues. They’d already dedicated the best of their own senior systems engineers to the issue with no success, and everything pointed to the machine having reached its maximum capacity level.
Though they concluded they would have to invest in a very expensive upgrade, they came to us to see what solutions Amdahl could offer.
Instead of a team of engineers or a comprehensive and expensive service solution, our Support Manager sent one man, Peter, and me as the new guy to watch and gain experience.
Peter spent the entire weekend camped out near the central control unit. He worked on those huge mainframes while I watched in awe and fetched coffee. In one weekend he changed the system’s parameters and freed up 15% of capacity!
The customer could postpone an expensive upgrade.
That was the moment that I realized that a top performer can have a tremendous impact, and that some people simply perform much better than others at certain tasks.
Throughout my career this pivotal discovery was reinforced several times.
One time the CIO of a big Scandinavian bank retired and joined 7N as a management consultant. I had a welcome-aboard coffee with him and asked if he felt he’d transitioned his 2,500-person IT organization to his successor well. “Yes,” he said. “I gave him a list of the 100 names he couldn’t afford to lose, and he understood the value of that list.”
My immediate reaction was to ask, “What about the other 2,400 employees?” His reply was brisk and businesslike: “They can be replaced relatively easily. But the 100 named individuals are the ones who drive the development of the bank.”
Another time I was part of a business delegation on an international excursion and met with the founder of a large music streaming service. I asked him: “You have disrupted an entire industry with your software and solution platform. How many of your developers would you say have made the major difference on your journey?” He told me that of the two thousand developers he employed, there were about fifty for whom he always made time and who he believed were truly outstanding and the real heroes driving progress and technology.
Fifty out of two thousand?
The insight I have gained from working with Peter, the Scandinavian bank’s retired CIO and the founder of the music streaming provider has helped me understand how much an organization’s total productivity is driven by its top-tier performers.
I have learned that there’s a difference between being equal in worth as humans and being particularly good in a specific area of competence.
When I set out to determine what drives performance, I recalled those past anecdotes. I knew intuitively that the difference in performance was likely a matter of the ratio of super achievers on each team, and wanted to know how much better these super achievers perform and how much more they produce when compared to others on the team. I immersed myself in research.
Though I had anticipated that there would be a significant difference, the actual numbers were remarkable. At 7N we have always assessed people by assigning a numerical category from 1 to 10: the lowest-level performers are a 1 and the top performers are 10s. My research in performance revealed, conservatively that 10s, who are the top 1% of all IT professionals in a normal statistical distribution, perform at a level that is almost two times greater than that of the next level’s performers who we call 9s and who typically represent the top 3% of a population.
Going further down the line, 10s perform at a level that is twenty times greater than that of an average IT specialist level of performance. 10s are at a level that is one hundred times greater than that of individuals classified as 1s. The performance curve was exponential! Allow me to substantiate further:
- The former head of development of a large global ERP-systems provider told me that the top 1% of his developers performed 80 times better than average.
- In an interview Twitter’s former CEO Dick Costolo said that he believed the top 1% performs 50-100 times better than the average developer.
- TopGrading, an organization that specializes in human performance and measurement, states that the top 1% perform 28 times better than average.
- Apple believes the top 1% perform 25 times above the average specialist.
- Google beats them all by claiming that the top 1% performs at a level 300 times above the average developer.
There are a lot of opinions about this. Though the numbers may differ, the bottom line remains the same: whether you believe that the multiplier is 28, 50, 80 or 300, having top performers on a team makes a remarkable difference in an IT department’s performance.
In the Nucleon formula, I have chosen to go with our own more conservative numbers because they confirm our observations and suggest that a 10 performs 20 times better than an average employee.
When it comes to IT development and engineering teams, 10s make a world of difference. Here’s a mind-boggling example: iOS X was developed in two years by just 600 Apple engineers, while it took 10,000 Microsoft engineers five years to develop Windows Vista. An influential Harvard Business Review article outlined the context of this comparison and the meaning of star performers.
Your company (including the CEO, HR, Procurement, etc.) needs to understand that by changing the focus from input cost (salaries or hourly rates) to output cost (total cost of ownership/delivery) and striving to have as many 10s on your development teams as possible, you will make a huge difference in helping the IT organization succeed.
This reality is often difficult to communicate, as it represents a shift in philosophy and focus. Many HR and procurement managers have KPIs that motivate them to pay as little per hour as possible for labor, while managers in the development organization intuitively want the best resources on their team so they can meet their goals faster. The tension of their contradictory agendas can be a source of frustration that won’t be eliminated until top management understands that it is expensive to buy cheaply.
For example, the chart below reflects actual rates for freelance project managers in the Danish market, combining their fee with how much they delivered using the 20x performance factor that our studies confirm as a conservative reality.
The most highly paid project managers (earning $1300 per day) actually cost around $13 per delivery unit, while an average project manager whose rate would be around $800 per day in the Danish market would cost $160 per delivery unit. A project manager who is truly a poor fit but who represents himself as an IT professional would never charge less than $700 per day – and in this example, would cost $700 per delivery unit.
As shown by the chart, it is 54 times more expensive per output unit (Total Cost of Ownership) to pay the lowest salary than to pay top dollar for a top performer. Despite the natural preference for paying as little as possible in order to get value for our money, we need to pay for excellence. Management must strive to acquire as many 10s as possible in an IT department, and this holds true for employees and consultants as well. Doing so will change your focus from cost to value, and will not only lower your total cost of ownership but also increase your total performance with respect to smart content and speed of delivery.
The need to attract and retain top talent has gotten a great deal of attention lately, and is both legitimate and important. You should definitely design and operate your IT department as an ideal place for IT specialists to work in order to make sure you attract the best of what is attainable.
But before you focus your policies on providing competitive perks, you need to make sure that you know how to identify the 10s when they come your way. This should be your top priority long before providing breakfasts, espresso machines, or on-site nap rooms. You need to know how to assess people to make sure that you are getting the best possible quality on your development teams.
Though making a practice of assigning numbers on a gut-feeling level is better than not reflecting on individual performance levels at all, I highly recommend a more objective way of assessing people — one that uses standardized measures, processes and tests to provide a better understanding of what an individual can deliver.
Every organization has its own culture and priorities, but generally speaking if you were to do a gut-feeling assessment of your employees with a goal of assigning them a number based on their performance, your scoring would probably follow these summaries for people evaluated as 10s at the top of the scale to 1s at the bottom:
10: Extraordinary performers who consistently achieve above expectations. Key characteristics include being independent, proactive, showing leadership skills, and engaging in mentoring. Extremely high drive.
9: Excellent performers whose achievements are usually above expectations. Key characteristics include being independent, proactive, professional and having a strong work ethic.
8: Strong performers who deliver results on time, within scope and budget. Key characteristics include being independent and proactive.
7: Above-average performers who deliver results on time and within scope and budget, but who have a need for frequent coordination.
6: Average performers whose results are generally on time and within budget, but who have a need for constant follow-up.
5: Below-average performers who barely reach objectives. In frequent need of coaching.
4: Unsteady performers who frequently miss delivery criteria and who need supervision.
3: Unsatisfactory performers who miss several project criteria and who are unable to work independently.
2: Unacceptable performers who have a clearly inadequate skill set.
1: Non-performers who represent a mismatch with the organization.
The numbers we assign do not indicate a person’s potential, and they are not an indication of an individual’s overall value. You should never lose sight of the fact that those numbers are a snapshot of current IT performance, where an employee might perform much better in a different function or environment.
We are assessing human beings – the most complicated and remarkable creation on earth, and our assessment work and ethics should always reflect that.