An Extreme Reaction

 

Being one of the thousands of people who use the British public transport system to get to work I like to think I have a pretty high pain threshold when it comes to problems and disruptions to my daily routine. Late trains and disruptions are a common part of my commute, however a recent event elicited an emotional response from me far greater than the situation in isolation merited. The event, and my reaction, provided me with an interesting example of the mindset of technology users in the face of enforced changes, and how our subjective perception of importance can shift dramatically given a context in which user options have been restricted. I also gained some compelling evidence of the importance of considering your environmental variables when testing.

We hate change

In November last year ticket barriers were installed at each entrance to my local station which required a valid train ticket to enter the station. Typically prior to this I had purchased my tickets from a machine in the station, however if queues were bad or machines faulty I could also go to the manned kiosk or buy a ticket on the train, risking only a mildly grumpy train manager. After the barriers were installed my only option to access the station by the street entrance that I used was a machine which was placed outside the automatic barriers. Going to the manned kiosk was still possible but required a long walk around the road to the main entrance on the other side of the station, and buying on the train was no longer an option. 

My natural inclination was to be wary of the changes, however after a couple of weeks of painless commuting I settled into things and reluctantly accepted that the changes weren't as bad as I expected.

Environment Variables

One cold day soon after, running a little late, I was pleased to find the queue for the ticket machine empty. I selected my season ticket and tried to enter my code on the touch screen. I quickly realised that many of the characters, particularly those around the edge of the screen, were unresponsive. I was having to press really hard on the screen to get anything to register, and then it could be the character below or above the one I wanted requiring me to use the equally non-functional backspace and try again. Three characters in and I had already taken far longer than I should need and it slowly dawned on me that it was taking so long to enter my number that I was at serious risk of missing the train ...

A little perspective

Now is probably a good time to add a little perspective. The train I catch is at 8.15am, the next train I can catch is at 8.31am. If I miss my train I'll catch the next one and be in 16 minutes later to work, which has little impact on my day. Occasionally I am delayed at home and decide myself to take the later train to avoid rushing. Sometimes I take it to catch up with a friend who travels on the same one, the day after writing this I took the later train to wait and travel with a colleague who was running late. Taking the later train, whilst not normal, is a perfectly acceptable alternative for me.

An irrational panic

...back at the ticket machine the availability of a train 16 minutes later was the last thing on my mind. As I struggled to mash the digits of my code into the screen I started to enter a state of frustrated panic. This infernal machine was preventing me from catching my train. Prior to the barriers machine failure occurred quite often, and would have registered as nothing more than an inconvenience in my journey. The absence of any other options now meant that my only path to catching my train was this infernal machine. I started to shout at it,  I find this always helps with errant technology, particularly if your main goal is accumulating 'funny looks' from passers by. With only a couple of minutes to go I was on the last character, a P which was located right in the corner of the screen. It wouldn't work. No matter of pressing, even with the pressure of both hands, would get the letter P to register on the screen. I started to panic. I could hear the train pulling into the station. If a genie had popped up at that moment offering 3 wishes my first and only one would have been for that @¥&£#% P to work so I could get my ticket. Realising that the cause of the malfunction was probably the temperature I leaned over the screen and began to frantically breathe on it to try to get it to warm up ( I couldn't risk using my hands for fear of accidentally firing another button). To passers by my alternate breathing into the screen and pumping my fists on it must have looked like I was trying to bring it back to life through some kind of technological CPR. Just like in the movies, at the last moment my efforts finally paid off, the P registered and I got my ticket. Running through the barriers I shouted at the staff as I raced past "THAT MACHINE DOESN'T WORK WHEN IT GETS COLD". Racing recklessly down the platform steps I managed to get on the train with seconds to spare and collapse into my seat in a flurry of anger, and adrenaline fuelled elation.

Rationality in reflection

On first reflection my actions were totally irrational and quite foolish. The sensible course of action would have been to calmly take the few minute walk around to the manned desk, buy a ticket and take the later train. I probably lost more productive time in arriving to work flustered than I gained in getting that train. So why was I so determined to get that one on this occasion? I think that the reasons become more apparent if we examine the emotional context that was in place coming into this incident:-

• An unwanted change had been imposed

I'd recently had some changes imposed on me which affected an established pattern of operation. Whilst as an individual I am quite receptive to changes and am often an instigator of change in my organisation, when a change impacts an established routine then it is rarely welcome. I'm not alone in this, many people have a natural tendency to favour the current situation and established behaviours over change. When the change was announced I had started to mentally accumulate reasons why it would not be welcome. Even though, for example, I hadn't done it for months I was still annoyed that the presence of the barriers would prevent me from taking my children onto the platform to see the trains. This is referred to as a "status quo bias" and is associated with a tendency to place a higher emphasis on the potential losses of a change than any corresponding potential benefits. As soon as it had occurred phrases such as 'I knew it' and 'typical' started echoing around my internal monologue. If I'm honest I was probably intentionally magnifying the severity of the problem to justify my previous objection to the situation. 

• I was not the beneficiary of the change

All of the benefits of the installation of the barriers were going to the train companies. These things were not there to make my life easier. My emotional position on the change was that it was going to be for the worse and there were few obvious reasons for me to change this.

• My options had been restricted

Whilst I occasionally had problems with the previous system I usually had the flexibility to work around them. As I mentioned in addition to the machine I used to have the options of buying at the kiosk or on the train. These were not readily available to me any more which placed a much greater emphasis on the machine working. The tendency to oppose any perceived restriction in one's personal autonomy is an extremely powerful bias known as "Reactance". Reactance applies when we feel that something is removing our personal freedom to choose.

• The fault could have been detected

This was probably the final straw in my over-reaction. The fact is that this was a machine that had been designed to operate outside in Britain yet didn't work in freezing temperatures. For most people this would be annoying. For a software tester it was exasperating. My whole frustrating encounter was the result of a singular failure to consider the possible environmental variables when testing the machine.

So the event occurred in an emotional context of: a strong set of biases against the recent change; a removal of alternatives; a piece of technology which hadn't been tested for its target environment; oh, and a disgruntled software tester.  

A Troublesome Business

The situation above may seem far fetched, however this is exactly the kind of emotional context into which my software, and many other business technologies, are implemented. With regards to the product I work on, our customers are often looking to replace an established product or process. The primary value in these implementations is usually to achieve the equivalent functionality as an existing system at a much lower cost. In this context the people involved in implementing, administering and using the system are rarely the beneficiaries of the value in the change. They are also likely to have an emotional investment in the existing process given their experience the technologies and associated tools involved.

Business software implementations are often in the context of:-.   

• replacing an existing process - thereby exposing the risk of status quo bias toward the existing system
• being imposed on the users rather than being their choice - engendering reactance based emotions against the software before it is even switched on 
• restricting the flexibility of operation - again giving rise to reactance based frustration at not having the flexibility previously enjoyed. Users can't, for example, stick a post it note on a computer form or write in the margin of an html page

The result is many low level stakeholders can harbour strong levels of bias against the implementation of the exact nature shown by my own experience. Whilst as a software vendor this is sometimes frustrating it is also totally understandable and something that software developers should anticipate. Business based software testers in particular need to consider the presence of negative bias in their implementations and examine the system accordingly. 

Occasionally I see comments from testers of business software almost jealously bemoaning the way that Facebook and Twitter thrive with little testing and a volume of functional flaws that would be unacceptable for business use. What we need to remember in business targeted software is that software used in a social context is rarely imposed and is usually done in the presence of a range of options (twitter, Google+, FaceBook, LinkedIn, Xing) out of which the user will have made a personal choice. This is much more likely to result in a positive emotional state when using product, and a more forgiving approach to faults.

To counter potential negative feelings towards our business software we need to focus on making sure the product and associated services are geared towards overcoming such a position. Users experiencing feelings of resentment towards our software are unlikely to enthusiastically digest every word of our documentation, so it is important that the software usability is considered. Is it consistent with the systems that we are replacing, or self-explanatory if not? Is it easy to identify and recover if mistakes are made? As well as the functionality at hand do the testers understand the greater goals of the customers. Documentation still needs to be accessible and searchable and written from the perspective of not only achieving key tasks but making necessary decisions, rather than a flat reference structure. Technical support need to be helpful and willing to step the users through their problems in the early stages to develop positive customer relationships and help to nurture a culture of positive and skilled use on the part of the customer.  

Of all the software implementations I have experienced, the ones I'm most proud of are when we've started out meeting some resistance to the product and turned this around into a situation where those users actively recommend it to others. Situations where initial resistance is overcome through our diligent work to understand the customer needs, test that the product allows them to meet these, and support them in doing so, creates some of the strongest advocates of our software and company.  The results are definitely worth it, as investment in startup companies can come on the back of references from existing customers. So the future of an organisation can depend heavily on the quality of product delivered in the present and the ability to generate raving fans. Failing to understand the target environment and failing to consider the emotional context that you are delivering into, and you may just end up with raving mad users.