Quality Assurance (QA for short) is a must in today’s digital revolution. But have you thought about why are you doing it? Even if you are doing it right, it may be for the wrong reasons.
QA is generally defined as a way of preventing mistakes or defects in manufactured products and avoiding problems when delivering solutions or services to customers… but even those of us working in the digital world is probably more familiar to QA than in any other industry.
I will not be addressing QA for industrial processes but rather focusing on online services such as user-generated content moderation, community management or social listening. These activities rely heavily on individual efforts rather than on systems accuracy; deal with people, not machines, and are far more relevant in today’s world.
Our customers want it all: good, fast and cheap. Unfortunately, as we all know, you can only choose two options.
Since the competition forces prices down, everyone goes with cheap. And since the market changes so rapidly, everyone chooses fast. So why do they complain when things don’t turn out that “good”?
When managing large and complex projects, customers set the bar on the error acceptance levels, the expected turnaround of responses, the acceptable time gaps, and other quantitative measures. Oddly, they call these metrics “quality”, although they primarily focus on quantity.
The main problem with using metrics like these is that the origins of quality control were primarily designed for machinery and industrial outputs. Social media, eCommerce, and other tools work primarily with people and for people. We are talking about highly complex projects and not a simple sweatshop where you can measure output simply by counting a number of items done “right”.
Some history
Quality assurance started as such in the turn of the 19th century, during the industrial revolution. At that time, workers were nothing more than a “flesh appendix of a large machine”. Working conditions were, to say the least, unacceptable by current standards. But labor was cheap, demand was high and there were no laws to protect the workforce.
During the 20th century, this situation changed drastically when the workforce was needed for more complex and specialized tasks. In this scenario, quality started to be shaped in forms that went beyond the simple output numbers. Being effective was no longer sufficient, and efficiency took precedence above everything else. Salaries and working conditions were enhanced to the few that worked using their intellect above other abilities. Training and specialization were at an all-time high and salaries and working conditions changed for those skilled workers due to a new type of demand.
At the beginning of the 21st century, however, it seems that we have turned the clocks backward. Efficiency, output, and “quality” numbers started to be required for new types of industries, including technology companies and social media. New increased numbers, high demand, and low sales changed the market, and “human talent” was downgraded again to “human resources”.
Why do we need Q&A?
As a manager, I like to think once in a while “why are we doing things like this?”. In the case of QA, this question is particularly relevant:
- We want QA to have comparable metrics and measure our operational success.
- The project needs “hard data” to show our clients (even when not two projects can be compared using numbers alone).
- We have to keep our upper management happy by delivering nice graphs and percentages.
- We need some justification to remove resources (personnel) that may not be giving the expected output.
If you answered any of the above questions with a yes, you are doing it wrong!
The big lie used by everyone in operations
“You cannot manage what you cannot control. You cannot control something that you cannot measure”
I am a big fan of attaching numbers to almost everything. Percentages, data, and graphs are the easiest way to make your point. If you go deeper and start analyzing trends and finding patterns, you may even come with timely solutions by identifying gaps or deviations from the planned paths. But anyone with a little statistical training knows that you can present the same numbers in a different way to say something completely different. No need to hide data or to change numbers, you just need to make the numbers “say” what you think that your audience wants to hear.
For example, what does a 68% customer satisfaction actually means? If we compare a previous result, maybe it will show some increase (or decrease). But online projects change so drastically in a short period of time, that chances are you are not measuring the same things or using the same criteria. Either using math or logic, a comparison then makes absolutely no sense… so the numbers, in this case, mean nothing!
Competition vs. cooperation
While in theory “everyone is involved in QA”, in reality, there are a few individuals selected for these particular tasks. The main problem starts when these people start focusing on the numbers and forget the fact that they are dealing with people. Have you started having this “gotcha” attitude lately? Have you considered any errors in the appropriate context? Were systems there to help or hinder? Is technology (or the lack of appropriate tools) affecting negatively? What about motivation? What is the staff morale after applying QA policies?
More often than none, QA processes start a competitive attitude among the team. Competition does not allow room for cooperation. Teams start to break-apart and problems arise everywhere. Do we blame QA then? No, more likely we thank QA for finding it!
Labor is now cheap in comparison to other resources. Thanks to a vast number of freelancers, the legal and geographical borders begin to blur, so this “resource” is more abundant now than ever before. If QA is used to replace personnel based on a percentage, we tend to believe that we are being efficient when in reality, we are just aiming to be effective. Have you thought about the hidden cost of recruiting, training and management that is wasted every time you replace someone? By no means, this is an efficient move.
Training, management, and accountability
QA finding should show primarily errors in the selection, recruiting, training and management of personnel. While QA numbers usually affect the workers that made an error, in reality, the process should aim to identify the gaps in the supervisory positions that failed in the way. The final error caught by QA is just the effect, when the lack of proper training, for example, may be the real cause.
Why do we punish the ones making the final error, and we celebrate and promote those who find it? Why we even reward those managers that find more QA as being effective, when they should be accountable of those findings as well? This “punishment” approach, even when disguised as modern management, is the main cause of job dissatisfaction, lack of motivation, high staff turnover and a general morale problem.
While no one should consider these points as an argument for not having quality control, I hope they are understood as a need to revisit the reasons behind QA, the approach taken towards the findings, and the need to analyze more than just simple numbers.
“You must never feel badly about making mistakes… as long as you take the trouble to learn from them. For you often learn more by being wrong for the right reasons than you do by being right for the wrong reasons.”
(Norton Juster, The Phantom Tollbooth)