Do Productivity Increases Generate Economic Gains?

by Jakob Nielsen on March 17, 2003

Summary: Usability improvements can save time-on-task, but critics argue that this is not the same as saving money. Others worry that productivity gains cause unemployment. Neither is correct: usable design saves money and saves jobs.


For many kinds of design projects, the main argument for a user-centered process is that usability improvements lead to increased productivity. This is true for intranet designs, enterprise software, call centers, transaction processing systems, factory-floor equipment, and any other design where people are paid to perform tasks with a user interface.

(For other types of design, productivity is less important, and benefits such as increased sales, improved customer satisfaction, or fewer support calls are the main arguments for usability. My report on the return on investment from usability has a detailed discussion of these issues and estimates of the expected improvements; this column focuses on productivity-driven design.)

Assessing Gains

There is no doubt that paying more attention to usability in user interface design can make a system easier to learn and faster to use. Still, some people doubt whether such improvements lead to actual monetary gains for the company.

In some cases, the connection is clear. For example, some systems are so complex that users need extensive training courses before they can become qualified operators. Industrial process control rooms are the classic example, but many administrative systems also require training. Typically, design processes that emphasize ease of learning can cut days or even weeks off the training classes. I doubt anyone would dispute that the resulting training budget savings are a direct economic gain from the usability improvements.

Productivity improvements are more difficult to assess. Consider, for example, the design of an intranet's employee directory . In our tests of employees performing tasks on a wide range of intranets in different companies, we found that the time required to find employee contact information differed dramatically:

  • 58 seconds on average for intranets with good usability
  • 2 minutes and 38 seconds on average for intranets with poor usability

In other words, improving the employee directory's usability from poor to good saved employees an average of 1 minute and 40 seconds each time they looked up a phone number or other contact information.

Arguments Against Productivity

Even people who agree that a redesign might save 1 minute and 40 seconds sometimes argue that this gain isn't worth the required investment in usability.

Their main argument goes as follows: Yes, employees might save a minute or two, but they'll just spend more time at the water cooler. Just because it's faster to look up phone numbers in the employee directory, doesn't mean that people will get any more real work done.

A second counter-argument is that even if employees did do useful work during the time saved by improved usability, you cannot calculate the value of that extra work by simply multiplying the amount of time by the employee's loaded salary. Instead, you'd have to look at the specific accomplishments and value them individually.

Time Saved is Money Earned

If, over the entire history of the universe, we only succeeded in creating a single instance of somebody saving two minutes, it would indeed be hard to argue that much additional work was accomplished during that two minutes. But that's not the case.

In reality, we save a minute here, two minutes there, another minute here, and we do this over and over again. Even small time savings add up.

Consider this: You arrive at the office at 9:00 a.m. and start your first task. Let's say it takes an hour, and you start your second task at 10:00 a.m. This task also takes an hour, and at 11:00 you start the third task, which takes an hour, too. You complete it at noon, and then break for lunch.

Now, let’s say that you get a new intranet with sufficient usability improvements to let you perform each of these hour-long tasks two minutes faster. You thus complete task 1 at 9:58, and immediately start task 2, which you complete at 10:56. Finally, you complete task 3 at 11:54. You now have six minutes until your scheduled lunch break. Yes, you might decide to leave early and extend your lunch, but would you do this every day? One day, you might look at your watch, see that it's 11:54, and decide that you just have time to call an important client and resolve an open question. Or complete an expense report so that it's not looming over your head when you return from lunch. Or maybe you even decide to start on task 4, and go to lunch late that day.

Averaged over a lot of people over a lot of days, one or two minutes saved per task adds up to a substantial number of additional tasks being performed.

Companies can realize an economic value from having more tasks completed per unit of staff-time in two possible ways:

  • Keep the same number of employees and have them perform more tasks, thus increasing the enterprise's output and collecting the added output's value.
  • Reduce the number of employees, and have the same work performed with a lower payroll cost. On average, across a big company, a smaller staff also lets the company close buildings or lease smaller offices, thus reducing overhead costs.

Marginal vs. Average Added Value

As for the second argument against the value of productivity gains: Yes, it is true that the most precise estimate of created value requires that we measure the new work product's monetary value. A well-managed company is already devoting all of its efforts to the most valuable projects, so if time becomes available to do another project, that project must by definition be lower in priority and thus have slightly lower value than the existing projects.

In practice, very few companies see a sharp drop in value when they move from existing high-value activities to those that are awaiting action. In reality, the value declines slowly and steadily along a curve.

Also, for a company to stay in business, the average value of an employee's work must be more than that employee's total cost. That is to say, people must produce value above their loaded salary (which includes overhead, not just take-home salary) or the company will run out of money. Thus, even though the company's next-possible activity would have slightly lower value than its current activities, its value is typically very close to that of the employee's time, since the current activities are valued higher than the employee's time.

To simplify matters, we can reasonably estimate the value of time saved based on the loaded cost of the employee's time. This might be a few percentage points off, but since it's an estimate , it's close enough for all practical purposes.

Does Increased Productivity Create Unemployment?

One of the ways in which improved usability leads to economic gains is that it increases productivity, which lets companies downsize their staff and still get the same amount of work done. This doesn't sound very happy in a bigger context, however, and some people worry that productivity creates unemployment.

In the 1980s, usability professionals were often faced with protests from unions and other labor representatives who did not relish the prospect of participating in user tests that would result in "better" designs that would in turn get them laid off. In recent years, this has been less of a problem, initially because of the boom years, and currently because there are other sources of unemployment that worry people much more than usability. Still, in some people's minds, concerns remain.

In individual cases, some people might indeed lose their jobs because improved usability led to such high productivity that fewer employees were needed. This is unfortunate, but it is not the representative case.

Basic economics says that the lower the cost of a good, the more that good will be consumed. In the case of usability, higher productivity means lower cost per unit of work produced, because the same person can do more. This again means that more work will be required, not less.

The idea that higher productivity leads to fewer employees comes from the fixed-work hypothesis, which holds that there is a given amount of work that needs to be done in the world. This may be true in a few cases, such as directory assistance: the faster each operator handles each call, the fewer operators will be needed, because the workload is fixed by the number of people who call up to find a phone number. On the other hand, if labor costs can be reduced, the phone company might lower the price of directory assistance, and thus more people might use the service. So, even directory assistance may not be a fixed-work scenario.

In any case, most areas of the economy are certainly not fixed. If things can be done cheaper, there will be more demand. Higher productivity need not lead to unemployment.

On the contrary, unemployment is much more likely to result if we don't improve productivity substantially. Many professional jobs are currently at risk of being relocated from high-salary countries, such as the United States, Europe, Japan, and Australia, to cheaper areas. The Internet creates the death of distance and the easy availability of people who are just as smart as people in expensive countries, but make 10% of their salary.

Increased productivity is the only way for high-salary countries to keep their jobs. Thus, usability professionals should not worry that they create unemployment when they improve a design and increase users' productivity. On the contrary, such efforts are exactly what's needed to keep us all in business.


Share this article: Twitter | LinkedIn | Google+ | Email