Now with full and corrected speaker's notes to every slide.
Bad IA in the workplace, the causes of stress, and the rise of bureaucracy 2.0.
Talk given at EuroIA 2012, in Rome, by Jonas Söderström
systems since the mid-nineties, and I’ve seen a lot of things like the one’s mentioned. However, this is not a talk about absurd interface bloopers. That’s not why I wanted to come here and not why I decided to write a book. This is a call to action – a call to arms even. And what motivated me was the trend shown in the next graph:
12 6 0 15 9 3 % In Sweden, from the mid nineties to around 2005, the share of the total workforce that experienced severe stress at work more than doubled. This graph showed a new trend. Formerly, bad times, recession, and thus risk of unemployment, were the main predictors behind stress at work. But the bad times in Sweden were the first half of the nineties. From 1995 onwards, the economy was booming and unemployment soon reached a historical low point. So everyone expected these numbers to go down. Instead, the curve went up steeply. The exception this time, was a radical shift in the workplace, a massive new use of technology, digitising and computerizing a lot of businesses and sectors in a short time. We often call it the dotcom-bubble; but it affected more than e-commerce and public web sites. In workplaces, all kinds of new systems were introduced at very high speed. Systems had low usability, and were not well-adapted to the actual work. Often the developers were boys in their late teens or at best early twenties, straight out of college or even high school, with no own experience of the workplace at all.
Sweden is an advanced country, digitised to a high degree. Out of a workforce of four million, one million white-collar workers spend eight hours a day in front of the computer. Even sectors like health-care and education, are heavily digitised; even as a teacher or a nurse, you’d spend a lot of time with digital devices of all kinds.
week +20000 2012 In a study released this summer, around 20,000 people were surveyed: 60% said they had problems with IT every week: a shocking 20 % reported problems EVERY DAY.
of several surveys of the time lost because of IT problems, estimates it at just under thirty minutes a day. These are the estimates of the users themselves. As UX/IA experts, I and my colleagues often find that people underestimate the time wasted. They often can’t see that a better IA/IxD solution would solve the task more quickly.
2006 Another way to present the same findings, from a similar survey: 75 % report losing up to two hours every week These are the statistics. How does it look in detail?
manuals for trucks A truck manufacturer had a system for creating user’s’ manuals for trucks. In this system, changing a number – for example from ”7,5 litres” to ”9,5 litres” – required eighteen different procedures. Each procedure consisted of several steps. There were many possibilities for mistakes – feedback from the system was often lacking, so you were often not sure if a procedure had been successful. The system was also sluggish and often crashed in mid- procedure.
You feel a complete failure, like your personal competence just blew out the window Being good with words and pictures has no value; it’s all about taming the system I don’t want to work with X-system any more, I hope I’ll nd something else as soon as possible This is what people who worked with the systems told us when my colleagues interviewed them.
the Swedish passport offices started to use a new system to produce passports in 2005. The system simply didn’t work, for a long time. This meant, that the staff had to draw straws in the afternoon, to determine who would have to go out to the waiting room, and tell perhaps 250 people, who had been waiting for up to eight hours, that they wouldn’t get their passports today either. ”Come back tomorrow, perhaps you’ll get your passports then, but we’re not sure.” As a temporary solution to the problem, the offices had to create special ”break-down rooms” for the staff, where they could go and cry after having faced countless angry customers.
the work for us, it seems that we just get more and more stressed out. I argue that this is the total effect of a lot of small changes in how we work. The workplace has been transformed, in many small steps. Each and every one of them might have looked perfectly OK. Every one probably seemed well-intentioned. Many might even have been tied to a business case (though not nearly as often as you might imagine). In the book, I examine eight different factors or scenarios that put a lot of burden on the worker. Not every workplace exhibits all of them. But I haven’t yet encountered one single organisation that hasn’t had at least half of them. If all the factors are present, the risk for heavy stress and adverse health effects is very high. We shan’t have time to go through them all, so I’ll just pick a few.
applications (systems, sites, software) that we have to use in the workplace, has exploded. In the supermarket that I studied, they had to use 20–25 different systems - one for ordering meat, one for ordering tobacco, one for ordering dairy products, one for handling coupons, one for handling loyalty cards, etc, etc.
UU+ (budget), Edgar (recruiting), W3D3 (documents), KDB (contracts), UpDok (tracking student’s performances), Time Edit (managing premises), AKKA (catalogue administration), PingPong (course administration), Opus (references), Selma (class web), The student portal and so on … This is a small sample of the administrative systems that professors and teachers at a university have to handle – at the same time as they are supposed to teach and do research.
system) Giraff (internal invoicing) Horisonten (accounting) Prognosprogram (economy) Winst (procurement portal) Rappet (client reports) Personec (HR reports) Time Care (work schedule) Winlas Webb (temps worked hours) Time Care pool (assigning temps) Lisa (accident reporting) Adato (rehabilitation process management) Offentliga jobb (recruiting) Telephone self-service system Lotus Notes (mail) Webbmail (mail at home) Here is about half the list of the c. 35 administrative systems that are used by social workers in Sweden. Note that they are simultaneously using a new and an old system for social benefits. That is often the case; a new system often does not replace an old one completely. It’s often possible to find – or invent – some reason for keeping the old system as well.
one shop - actually a pharmacist’s – two systems where used simultaneously in the computerised cash register. One was to calculate the amount of the prescribed drug; the other was to print the labels for the boxes. In the first one, a certain shortcut did just the thing you wanted: calculate the amount. In the next step, however, using the printing function, the same shortcut was assigned to “close and do not save”.
the other comma 1.5 hours 1,5 hours Having to use parallell systems is a reality for many people. Liza is a consultant; she reports time both to her employer and to her client.
to learn • ”...how did you handle this system, then?” • Even if each system is used quite seldom … • … some system is used each month or week. Used seldom, but ... Many systems may be used infrequently -- which makes learning harder. “How on earth are you supposed to handle this system, then?” But since there are so many systems, you encounter this situation every month or every week.
last ten years is the speed with which new systems are implemented in the workplace. This is thanks to the web architecture, whereby a whole business can be upgraded overnight.
at 8. They then have just sixty minutes to order the goods for the day after tomorrow - or they will have an empty store with no milk and no bread. They queue up, because they have only one computer. On the computer screen, a clock is ticking down: 54 minutes to last order, 53 minutes to last order. And having placed their orders, they have a lot of other things to deal with: making the shop tidy and neat, arranging products and signs, etc. The morning I was observing them, all staff got an e-mail in their inboxes, saying “Hi. Your system has been upgraded.” Attachment: a twenty-page Word document with screen dumps. No-one had the time to read that, not that day, nor the next, nor the next.
and time the user has to spend on things like virus protection, upgrades, passwords, etc. These problems accelerate when you connect many systems to each other - all become potentially vulnerable, and need upgrades and protection.
the workplace is the fact that so little training, so little introduction, so little support is given in handling all these systems. True, systems should be intuitive - but it’s an illusion to suppose that they can all be mastered without any introduction.
system) Giraff (internal invoicing) Horisonten (accounting) Prognosprogram (economy) Winst (procurment portal) Rappet (client reports) Personec (HR reports) Time Care (work schedule) Winlas Webb (temps worked hours) Time Care pool (assign temps) Lisa (accident reporting) Adato (rehabilitation process managment) Offentliga jobb (recruiting) Telephone self-service system Lotus Notes (mail) Webbmail (mail at home) ”It’s just a small system” The excuse for not giving training is often “It’s only a small system”. Yes - another small system added to the thirty we already have!
system she uses has well over 120 screens, which is by no means unique; many other systems are even bigger. Every screen can have over a hundred controls (fields, buttons, menus, etc). But we see only one screen. If we were to visualise the total size of the system she has to master, we could for example print out screen-dumps and paste them to the wall in front of her. But that would not be enough. We would have to use the wall to her right, the table, even the ceiling to try to visualise it ... ... and when it adds up like this, to me it starts to look ...
find out: “Is there a certain method that always gives good results?” His answer was NO. There is no method. But there is a number of traits that you find behind successful products and projects. Photo: Martin Kliehm/flickr under cc-license
six weeks, observed real users using the product or service for at least two hours? This is the most interesting, and perhaps the most important. But when you design systems to be used in the workplace, I don’t think that is enough. You can’t look at just the specific product. You have to look at the total situation at work.
few friends wanted to dine in the Swedish city of Norrköping a few weeks ago, they booked a table at a downtown Italian restaurant that seemed nice. When they arrived, they were greeted by the headwaiter, who asked if they had a reservation. Richard confirmed, and the headwaiter looked at his computer screen. ”Gatarski? Hm, let’s see .. yes, there’s your reservation. Welcome!” The headwaiter then picked up what Richard first thought must be some kind of new, electronic touch- pen, and moved it toward the screen. Richard is a tech savvy Internet entrepreneur, and therefore quite curious about what kind of new gadget they used at this restaurant. So he leaned in and looked a little closer … Photo: Richard Gatarski
whiteboard felt-tip pen. The head waiter just draw an “X” over their booking, directly on to the computer screen! “That’s very interesting,” Richard said to the head waiter. “How come you do that?” “Well, you know,” the head waiter answered with a great sigh. “The guys that create these kinds of systems … they have …. Well, you can’t do things the way you wanna do them. You can check off a reservation in the system, with the mouse, but hey, it’s at least four clicks away from this screen. And you can’t tell if the guests have been shown to their table or are still waiting in the bar. So it’s much easier just to draw on the screen (and when the evening is over I just wipe the screen with a cloth). We’re very busy here, and this works just fine.” The point is that the waiter at this restaurant wanted to give the customers the best possible impression, focusing on them from the very beginning. Remember: first impressions last. He did not want to tell them “wait a minute” and then focus on the machine.
six weeks, observed real users do everything they do at work for at least two days? On the other hand, in a conventional usability-testing situation, for instance in a lab where the customers probably exist only as instructions, this system might have erformed quite well. So this is my version of Jared’s statement. When you’re making systems or products that are to be used in the workplace, you have to spend much more time, to be able to capture the full experience of your users. As a rule of thumb: at least two days.
Sometimes, certainly, when interfaces are clumsy or too technical. But actually, he’s the smallest part of the problem, and often quite eager to do better. (Photo: esc.ape(d)/flickr under a cc-license)
friend tries to order a “black and tan”–half-Guinness, half-lager–in a pub. Although the bartender has both Guinness and lager in the pub, he tells him “Sorry, I can’t do that ...”
in the computerised cash-register” This is perhaps a banal and mundane example, but the same principle shows up in many places and many contexts: systems that limit the way you can conduct your work.
2012 Indeed, in a survey among eight thousand white-collar workers, 50% agreed that “IT systems control my work in an annoying and unreasonable way”. Which is a bit strange - since we invented machines to do the work for us. Are we working for them?
a public home care program Lena is a licensed practical nurse (LPN - in some countries, equivalent to “enrolled nurse” or “Division 2 nurse”). She works in a Public Home Care programme, providing for the elderly. She carries a digital device – a smartphone, or a bar-code reading pen - that registers her every task and every move during the day.
minutes • Bedclothes: 4 minutes • To elderly B: 13 minutes • Food: 7 minutes • Sweeping: 12 minutes • To elderly C: 19 minutes She’s connected to a planning system that breaks down her working day into a single-minute schedule.
minutes • Bedclothes: 4 minutes • To elderly B: 13 minutes • Food: 7 minutes • Sweeping: 12 minutes • To elderly C: 19 minutes ✓ ✓ ✓ ✓ ✓ ✓ ✓ She must check in with her device for every single task done She has to register every single task in her device, connected to a central management database, that creates wonderful reports. But what happens to a person’s motivation for her job, when she’s controlled in every single detail? Where is trust? And where is the compassion?
the things now demanded of us at work are things that we really don’t feel are “our REAL work”. Not really helping the elderly. Not really engaging with pupils in the classroom. Not the things that made us want to be nurses, doctors, teachers, etc. Social workers, doctors, police, teachers are now spending more time on documenting and reporting than on actually meeting the clients, the patients, the pupils.
Uppsala University This is an alarming article, published last year by a group of professors and researchers at Uppsala University. “We’re drowning in administrative systems that take more and more time away from teaching and doing research”, they wrote. (We saw part of their list of systems in an earlier slide.)
previously handled by a central administration, are now pushed out to the periphery - the departments. They now have to do wages, planning, budget ... etc, etc. But, in spite of this, central administration has not shrunk. On the contrary, its budget has grown by 6% (which is a lot, in an organisation where you often get MINUS 1 or 2 % annually). The university administration replied that 6 % was not exceptional; indeed, it was the average for Swedish universities. The professors’ final reply was “then it’s even worse than we thought”. But the number “6 percent” stuck in my head. Where had I heard “6 percent” before? As it turned out, I had to go back in time.
this. As IAs or UXers, we probably take for granted that the reasons for projects, for services that we are about to develop, must be essentially good, rational, benevolent. Of course there will be disturbances or interference from egos, from politics, or for practical reasons. But the reasons behind the project should be sound, about values and the demands of customers. Given what we’ve learned from the history of bureaucracy, I propose a more radical theory of IT systems in the workplace.
to hire new people easily, and increase your power that way. That is virtually impossible in today’s organisations. What is possible, however, is to implement a new IT system. Being in charge of an IT system means having power, to influence decisions, to choose directions ... So IT is an ideal vehicle for the purposes of the middle management class. It’s also sexy, glorified; almost by definition, an IT system is a good thing - it’s the icon of our times, of progress and the future.
will create work for two others. It’s rather easy to see that one IT system creates the need for approximately two other systems. If you start with a simple system for money in and out, it’s easy to put in the argument for a budget projection system, a time reporting system, then you’ll need a system for aggregating data from other sources, etc, etc.
2012 This might sound crazy. But it fits with a number of empirical observations. For example, in one of the Swedish surveys of white-collar workers, a majority - 60 % - said that there was “no clear or firmly-established reason given for the new system.”
85 % ComputerSweden 2010 Or that just 15% of companies, which invested in new business systems, had set any concrete effect goals for the new system ...
years of studies (for example the Standish reports), that around 40 % of IT projects are abandoned in mid-construction. It’s hard to understand why this happens so often, if the projects would really have had good value as a clear goal.
to build products or services that provides good user experiences, but too often we are ignored still - or perhaps politely listened to, yet ignored in the end. Why? Well, if the real - the deeper, unacknowledged - reasons aren’t service to the customer, but something else ... it would make a bit more sense.
responsibilities turned us into part- or full-time administrators” I’ve given you mainly examples from Sweden, but the signs and traces of bad systems in the workplace seem common in many countries. The following slide is a quotation from David Graeber, writer and historian:
IT departments / support ✓ Solution: more money to IT companies And they are typically the problems that first spring to mind, when workers are asked about their problems. Unfortunately they are often the ONLY problems that are considered by IT departments and management. That means that other, bigger problems are overlooked.
by users ✓ Not well understood by support / IT departments ✓ Blame the humans: ”Not technology savvy”, ”Whiners” Bad IA/UX is much harder for users to spot. They often don’t have the training to see that a system could be designed in a better way. So they sometimes don’t recognise these inefficiencies. And of course, in contrast to the former category, these are problems that arise when systems do work - or at least seem to work, from a purely technological point of view.
is carried out Ever-increasing documentation and reporting ”Self-service” solutions Command and control Prescribing process/ workflow management Geo-tracking And then we have a third category. What we see now is the rise of what some scholars call– instead of “the information age” - “the administration age”. Others speak of “the control society”. Computers and IT go extremely well with increasing demand for command-and-control in the workplace. This is often given names like “process management”, “workflow systems”; and it comes with detailed tracking, documentation and reporting.
is carried out ✓ Loss of flexibility, routinisation, digital Taylorism ✓ De-professionalisation ✓ Transfer of power to central management ✓ IT at the core of this development ✓ Middle management + IT companies = ♥ It is often claimed that these IT solution leads to higher efficiency. However, there are several studies that actually show the opposite. And they also have a number of negative consequences for the individual. The most important thing to understand is that IT is actually the driving force here. Nobody would dream of implementing these policies using paper forms; that would seem ridiculous. And in contrast to the guild of form-makers, which catered for the needs of the bureaucrats of Parkinson’s first discovery, the IT industry today is an aggressive, multi-billion industry. And yet many workplace systems are indeed just plain forms - on screen instead of on paper.
/ frequency of introduction of new systems • The burden of vigilance If you design systems that are to be used in the workplace, please consider these points.
a large extent grey, ugly, poorly-adapted to human tasks and needs. There is no reason why the systems we (have to) use at work should not be as pleasant, easy to use and well-integrated as the systems we choose to download to our smartphones. Workers need “consumer- grade usability”.
that the web and digital age have given us as consumers, in our private life, give us freedom. Enable empowerment. Build upon trust. Make self-organising possible. The trend in the workplace is the opposite: More control. More distrust.
conferences like this, e-commerce or consumer products and services are often in focus. I invite you, I encourage you, I beg you - come and work in corporate systems. Your help is badly needed. People are hurt. IA and UX can have a profound impact.
& Health agencies I can see no other group being in a better position to do good here. And it’s a huge business opportunity. Because the occupational health & safety people don’t have the qualifications to do it. They’re - sort of - still “stuck with the chairs”.