Robots are all the rage in the media these days.
Almost every story, whether news or fiction, says the same thing: robots are coming for our jobs. A few claim that robots will bring about a new age where humans are elevated into new jobs in existing and as of yet unknown industries.
William Stanley Jevons, on the other hand, doesn't care about your robots.
Robots In the Driver Seat
Who the heck is William Stanley Jevons and why should you care about him? Jevons was a 19th century economist who can end the productivity debate in your enterprise.
In the year 1865, Jevons was researching fuel consumption in Manchester, England and observed something rather peculiar: Technological improvements that increase the efficiency of a resource (the burning of coal in Jevons' case) didn't lead to a decrease in consumption of a resource.
Increases in the efficiency of a resource actually lead to increased consumption of said resource through increased demand. This phenomenon is referred to as the Jevons paradox.
While it may seem unintuitive in the abstract, it's easily understandable when you consider it in terms that relate to an individual: If you bought a car that got more miles per gallon than your current vehicle, wouldn't you drive it a bit more given that the cost of driving had dropped?
While this paradox is the bane of environmental activists (because increases in fuel efficiency very often lead to increased emissions due to usage spikes that exceed the efficiency gains), it may just be the boon of many technology practitioners who fear the coming of the robot horde.
More Robots Will Lead to More Jobs
Software professionals everywhere are debating what the outcome widespread automation in various practitioner segments will have. Amongst the most contentious of these debates is in the QA segment where traditional QA analysts and testers are feeling the effects of the automation wave.
The discussions on this topic tend to be highly political and guarded because many of the stakeholders, including practitioners and management alike, are nervous to talk about the overall effect on labor headcount.
Corporate cultures are always evolving, but one thing has remained constant: enterprise initiatives sold on increased worker productivity are often, if not universally, expected to reduce headcount. The reason this is most often true is because internal practitioner functions have very limited ability to influence external demand. If this wasn’t the case, then the initiative wouldn’t be sold on the basis of “increased worker productivity.”
This conceptual model is, in the experience of many, the predominant factor behind labor's resistance to the automation wave: because they fear for their jobs after the work is done and the robots are fully operational.
Many economists debate whether the robot wave will create enough jobs to replace or supplant the ones that either become obsolete or require less labor to complete. When we only consider the QA segment, however, I believe most experienced professionals would agree that Jevons paradox applies — automation of quality testing will create more demand for QA talent, not less.
Increased QA Efficiency = Increased Demand
Why is this so?
Mid-transformation testing teams don't cover anywhere close to the level of sufficient testing required to consistently release with both speed and confidence.
- Many traditional enterprise QA staffs are staffed based on a minimally responsible model. When the amount of effective tests executed per employee jumps by an order of magnitude, it will only begin to reach the coverage levels necessary for releasing complex software with confidence
- Manual testing teams are overburdened with documenting and executing tests that cover basic functionality and don't have the time or staff to devote to advanced testing concepts like negative testing, invariant testing or functional testing under load
- Even a fully-transformed team in a large enterprise will ultimately be affected by increases in application breadth as well as software complexity because software, like a gas, expands to the size of its container.
All of these factors point to the same thing.
As QA professionals become more adept and efficient at their jobs, overall demand for QA activities will increase. What was formerly looked at as an expensive and nearly unreachable luxury, will now not only be affordable and possible, it will also be required to get the promised speed and quality returns of agile and DevOps.
The Only Way to Lose to the Robot is Not to Play
For the sake of argument, let's assume that Jevons paradox does not apply here, and that a deep commitment to test automation will in fact reduce the demand upon the labor pool.
Even in this model, the smart play appears to be to lean in to the automation wave.
Imagine an enterprise leadership team running through the analysis of how they'll handle the decisions to shrink their labor force. Assuming that some people embraced the paradigm change and others resisted to the point of not mastering the new skill set, who is likely to retain their job? Who is likely to have opportunities for advancement?
After all the analysis, one curious fact remains: If Jevons paradox does indeed apply here, it would indicate that QA automation isn't an efficiency play after all, given that the end result won't be "to do more with less."
Instead it points to the core message of DevOps, agile and automation practitioners across industries — advanced automation efforts increase speed and quality at once, which to many people will seem like the bigger paradox of all.