A dystopian testing future

April 05, 2019

So I finally watched the original Blade Runner (I was one of those people who watched the “2049” one first), and I was overwhelmed. I think it’s a beautiful science fiction piece of work, and definitely one of my favorites movies from here on out.

The movie explores some narratives in a wonderful way and definitely leaves one wondering and reflecting on them. One of the things it left me thinking was about the steps that would lead to a society where corporate moguls reign, in particular, one of them, like the Tyrell Corporation, which gets to build an empire on genetically manipulated humanoid slaves (aka. replicants). Some part of it hit close to home to me for reasons I hope will be clear throughout this post.

First, let’s start with a question:

Have you ever wondered what dystopian reality the future holds for the software testing craft?

I’ve been thinking and wondering recently about multiple half-realistic and half-dystopian future states that the testing craft and industry as a whole might reach. Right of the start there’s one thing that becomes apparent when thinking about this: the “testing-as-a-service”-vendors influence in the industry will probably be a big weight in building that future, the vendors themselves might as well become the main infamous characters of the narrative.

Picture the following scenario:

Huge reliance on “testing-sweatshops” and “testing-as-a-service farmhouses”. Entire “corporations” that, in a “Blade Runner”-like fashion, are ready to satisfy all your software testing needs with disposable testing labor resources (sort of like “replicants”).

Could we ever reach such a state in the near or distant future?

The “industrial age”

I didn’t know that, besides the great illusion of a UN of testers made up by some “distant council of elders”, that there’s also this huge trend where tons of companies are paying (probably tons of) money for “Quality Assurance” vendors, not knowing what they’re actually getting for, money-worth. Most times because they maybe don’t understand testing or it’s importance, and think: “surely outsourcing testing is as easy as ordering a pizza online”.

I had no idea the extent of a phenomena currently underway, there’s a whole market out there of what some may call “Testing sweat-shops” or “Testing-as-a-service farmhouses”.

I call them sweat-shops, jokingly, of course, the more formal description is something in the lines “Testing-as-a-service” vendor. There are other names for it, but the goal is pretty much the same.

Most of the times I’m a bit “disillusioned” with the whole idea. It all starts with the way a portion of these vendors (safe some happy and decent outliers) market themselves, here are a few usual marketing pitches:

• a) “You pay us for testing services, and we’ll provide thousands of test cases and advice!”
• b) “We’ll provide thousands of latest-tech super duper automated tests!”
• c) “We’ll automate all your testing!”
• d) “With the power of BDD you will prevail!”
• e) “We’ll do great testing on holidays like Christmas, during weekends and during the night!”
• f) “We have a wonderful super-duper expert tester ready to allocate with you right now!”
• g) “We have hundreds of wonderful expert automation and manual testers that know 100 programming languages and libraries, with 30 years of experience in Cucumber, waiting in the docks!”
• h) “We guarantee we’ll assure the quality of your product!”
• …

Each of these points is just another different flavor or bottle of snake oil a company is trying to sell, and I doubt the inherent value of each. Anyone who’s been dealing in with the “Dark side of the Force”, what we can also call the “underground”-scene of the testing craft functioning as an open-society, can demystify each of these marketing pitches no matter which spectrum of “The Force” you tend to align yourself with, here’s a few examples, respectively:

• a) Increase in test-cases is not equivalent to better testing efforts and is a horrible means of displaying evidence of testing. We don’t turn to airplane pilots and ask them to show us their Pilot-cases. By the way, how many manager-cases is your manager churning out per week? Is the manager covering the product with a lot of management-cases correctly reported in JIRA? And what about the Scrum Master’s-cases?
• b) Mistaking checks (which are logical in nature) for tests and testing; Plus, the “Latest-tech”-cliché, …
• c) Testing (not checking) is not something that can be automated, it’s about learning, reasoning, experimentation, questioning, and other Humane activities that require Humane agency… Plus, you don’t turn up to a programmer and ask if they’re a manual or automation programmer.
• d) I feel like puking, I’ll stop here, you get the point.

The trend itself is not much different from other industries: it’s a simple cluster of companies that knowingly try to make money selling recent sayings, keywords, and dreams. This is not something new.

There are outliers

Not all is bad. There might be decent chances the actual external vendor-provided tester who’s providing you with testing, might be a skilled and caring tester (for example, up until now I’ve experienced this multiple times, having worked and currently working with awesome non-internal testers). In most good cases you end up not noticing the difference: “here’s a skilled tester, a good colleague, who happens not to work in the same office I work in but is helping the project with valuable and meaningful testing”.

Or maybe the vendor doesn’t blatantly sell you lies and catch-phrases like “We (magically) assure quality!”, and they tell you upfront in their marketing documents: we don’t make miracles, but we’ll try to do a reasonable amount of skilled testing, show evidence of it, and maybe do some resourceful automation in testing with the time that you give us. There are a few decent vendors like that out there, who don’t make a business of selling the good old “Quality Assurance” or “Automated-Testing” snake oil. But again, not all vendors, and not even the good majority, I’ve been observing, operates in this truthful form.

Now, for many minds, what one gets at the end is a labor resource, harshly put, a brain and a pair of hands, a headcount, to work on your project from a testing perspective. That’s it. That’s the “crude reality” for many minds that have the money and power to make decisions.

From a (dumb) business (and testing-agnostic) standpoint, in theory, the end result should not stray very far from the path of sourcing and hiring internally some testers out in the wild. So what’s stopping those nasty vendors fighting as much as they can in the attention market and increasing the chance to get new clients? What’s the problem? Right?

But, as testers, we know that the outcome of this labor resource depends on multiple circumstances. And because it’s vendor provided, as a company you skip the “dreaded” “problems of internal sourcing”, cut some costs (which makes shareholders happy), and everybody’s “happy in the end”, in detriment or benefit of the project of course… the harsh reality of “subcontracting” is you did nothing but trade problems for other problems, a good deal of those problems unknown “unknowns” from start to finish.

Now imagine a “Tyrell corporation” of testing

How would we go from a pool of mostly evil-intended testing-as-a-service vendors (sweat-shops) (with some exceptions/outliers) to a future ruled by Tyrell corporation?

I don’t have a perfect 1-liner answer for this, but, I personally think of a dystopian scenario might be reachable with the following steps, not in any particular order:

• Big testing sweatshops start purchasing small ones, and with time you end up seeing a monopoly of big “testing-as-a-service” vendors, the kind of which provide services to all those companies listed on any stock exchange because “internal sourcing/hiring” “is hard”;

• Smaller companies see the trend of big companies outsourcing testing to evil vendors, and decide to also outsource testing;

• Regular Companies (whose product is not for example selling headcounts), that might influence the testing market (for example, they employ key testers that are influencers in communities) slowly stop completely wanting to have the trouble of sourcing for their own testers, and become easy clients of testing sweatshops, again, because “internal sourcing/hiring” “is hard”;

• Continuous Growth and spread of the phenomena/fetish of certain certifications added with an increased global influence the “UN of Testers council of elders” as the sole “voice of truth” in testing, with complete theory control of the industry and a slow fading of the open-society model in the testing craft;

• Fading and merging of smaller independent and local testing communities under the umbrella of “happy and inclusive and positive” communities, that might turn evil in the future (which in turn would promote bad testing teachings or spread confusion/disinformation) - example, imagine a future where umbrella groups like for example Ministry of Testing, at one point turn completely evil and become basically a big worldwide network of “Jedi-lies Temples”.

• Known testing influencers and individual consultants being financed by testing sweatshops and tools-vendors (this is happening right now in the industry);

• Known testing influencers and individual consultants actually owning or forming their own testing sweatshops, and then using their contact networks and their own channels to further promote their companies;

• Wolves-disguised-as-sheep testing influencers, with voices that end up prevailing since that scream higher and have more “likes”, conquering all attention share of the industry, and then using that attention to promote evil or bland teachings. This one deserves some more explanation. My idea is that in an era currently ruled by social justice warriors fighting for attention time and also “offensiveness is offensive” virtual words-of-order being thrown at every corner, I think the testing craft is also not free from falling into traps that might promote, for example, a narrow “cheerleader” view of what the testing society should look like.

• Big and (supposedly) independent testing communities being financed by testing sweatshops and tools-vendors (this is happening right now in the industry); Imagine testing podcasts that involve communities, and the podcasts are non-innocently financed by third-parties. Here I say podcasts but it can be so much more: websites, tutorials, classes, any kind of content!

• Testing conferences guided by agendas of specific Sweatshop companies or companies with strong links with umbrella testing comunities. Have your doubts? Do a something of a bot-like “attack/test” on registry and voting systems of any of current “Big” and “Major” testing conferences - you’ll soon be getting private messages of CEOs of private companies that you didn’t know they “owned” the conferences… Let’s just say a close friend recently found out that a conference he innocently thought was “neutral” and “made by testers for testers”, is, in reality, actual property of an actual company.

• Bonus “Blade Runner”-like step: one mega testing-sweatshop corporation figures out how to genetically engineer and produce actual replicants with a 4-year lifespan to perform software testing.

• …

Wrapping up

This is mostly plain imagination. But the core ideas are there:

• Some of the points are related to “evil-mode” testing vendors;
• Some of the points are more related to the current observable vanishing of the “open-society” model in the testing craft (bear in mind the core business of the “UN of testers” is not exactly friendly towards this model);
• Others points are related to typical capitalist/business inclinations (“big company buys a small company to maintain bigger market share”);
• A few of the points are also related with power moves inside communities and umbrella communities. This is ages old, and not particular to any craft alone, but the testing craft if not free from it.

There are probably more points. Now take into account all of these and think for a moment. Are we really that far of reaching a state where the future is indeed characterized by many these points? How could we tackle the development of such a future? Again, most of these points are born of my imagination, but I’m sincerely worried that some of them may or are already turning into an actual reality.

Each day it seems to me we’ll start seeing more new vendors, all advertising “the satisfaction of any quality assurance and testing desires” of both big and small companies alike. All of these vendors are also always “vomiting” the same market pitches, and sadly also usually the same kind of “robot”-like resources instead of human and humane testers.

So what’s stopping this and other trends of becoming something worthy of a Blade Runner-like script?

It’s just my fertile imagination, right?