Certification of software professionals has been a hot topic for quite a while. At least 15 years. Maybe longer.
I keep hoping that the whole thing will blow over.
But it hasn’t. And it’s not going to. Too many people have too much of a financial stake in the success of certifications. Certification customers, including individuals and their employers, want certifications to have value. Certification providers want to continue making money.
But while I’ve been able to ignore most development and test certification initiatives up until now, I don’t think I am going to be able to ignore Agile certifications for much longer.
So I guess it’s time for me to talk about this publicly. I’ll start with tester certifications.
General certification programs, like the ISTQB tester certification, focus on knowledge of “best practices” and definitions. I have nothing against learning the material in the ISTQB Syllabus. There’s good stuff in there (even if the most recent books in the Foundations bibliography were published in 2004). However, I do have a problem with charging a whopping huge amount of money for test preparation classes, and testing people on their ability to memorize the contents on a Body of Knowledge, then slapping a Certified Tester stamp on their forehead.
The classes surely have some value. The trainers I know who teach certification prep classes certainly have much to offer. I see no harm in learning what these people have to teach.
But the cost of these classes is high. Certification preparation class can cost hundreds of dollars more than comparable non-certification classes, and the ISTQB test fee is another couple of hundred dollars.
And for what?
It is not clear to me that there is any evidence demonstrating a positive correlation between competence at software testing and possession of an ISTQB certification. (Some wags have argued that there is a negative correlation. I’m not going there.)
Rather, I suspect there is no correlation. I do not believe that certified software testers are any better at testing, on average, than uncertified testers.
And because I do not think there is a correlation between tester certification and competence, I see no value in software testing certifications. I think they’re a marketing scheme concocted to increase training revenues.
But people buy into this stuff, and classes leading to certification outsell classes that don’t lead to certification.
It’s important for me to note that I don’t have any problem with certifications in a specific technology. When Microsoft certifies someone as an MCSE, it means that Microsoft, the creator of a technology, is certifying that the candidate has met minimum competency requirements in that technology. Microsoft is not pretending to certify someone as a developer; they’re certifying that the candidate knows some specific fiddly details about a specific technology related to development.
It may seem like I’m splitting hairs. But it’s an important difference. There is a right and a wrong answer for specific technical questions like how to change a Windows Registry setting (hint: it’s not the form you submit to Microsoft to register your copy of Windows). General topics, like Software Testing, are not so clear cut. What’s right in one context could be dead wrong in another.
So, enough on tester certifications. I’ve successfully ignored them up until now, and plan to continue ignoring them in the future.
What I’m really concerned about are general Agile certifications.
I started hearing the rumblings around Agile certification some years back. In response, the Agile Alliance published a statement about certifications. It’s a good statement. I’m delighted the AA published it. I was in the room when Brian Marick and a small group decided to write the statement and I think they did a fabulous job on it. That can’t have been an easy thing to write.
And I was delighted when Laurent Bossavit and Brian Marick started WeVouchFor, a different kind of certification involving endorsements of competence rather than tests of knowledge.
Sure, there’s always a risk with endorsements that it becomes a kind of you-scratch-my-back-I-scratch-yours mutual love-fest. But I think that even such reciprocal endorsement arrangements say far more than commercial exam-based certifications that have a pass rate in the high 90%s. I’m not alone in that perspective, but endorsements alone are not enough for a lot of people. They want the certification.
I do understand the desire to have Agile certifications.
Agile is relatively new. There are a lot of people, and companies, sporting big ol’ “Now with Agile!” stickers slapped on top of their old RUP/CMM/CMMi/current-hot-thing stickers. So it can be difficult to tell those with deep Agile understanding from those who think they’ll make more money by adopting the hot buzzword-of-the-month.
And so employers look for objective evidence, like certifications, that someone who claims to know Agile actually knows what they’re talking about. And individuals want those certifications as a form of evidence.
The only real Agile certifications that I am aware of right now are the various Scrum certifications. Since Scrum is now the most popular Agile process, it’s no surprise that the Certified Scrum Master (CSM) is the most commonly-held Agile-related certification available today.
As an aside, it’s my understanding that the CSM designation started as a kind of an in-joke. I got my CSM by taking the CSM class. In it, Ken Schwaber said that the certification meant that we probably knew a little more at the end of class than we did at the beginning. But he wasn’t guaranteeing it. And then he taught us all the “secret handshake” (woof) so we could prove to other CSMs that we were in the club. (For the record, I took the CSM class so I could meet Ken Schwaber and learn about Scrum from one of the originators. The certification was a side effect of taking the class. The AHA! moments and resulting deep learnings are far more valuable to me than the certification.)
Then the Scrum Alliance decided to take the CSM, and other Scrum certifications, seriously. They put teeth in the certification.
And that’s fine. It seems to me that the Scrum certifications are like technology certifications. The Scrum Alliance is certifying knowledge of Scrum, a specific process. They’re not trying to certify general knowledge across all things Agile. They’re not saying that being a CSM means you’re generally competent. They’re just saying that being a CSM means you know what a Scrum Master does within the Scrum process. You know the mechanics.
Further, the CSM class is fabulous with or without certification. It’s experience-based, participative, interactive and leads to deep learning. I recommend it.
But this week, I am again hearing the rumblings for general Agile certifications, not just Scrum certifications.
People are asking me how to become Certified Agile Testers. The very thought makes me queasy. Agile Testing isn’t a process or a technology. It’s testing in an Agile context. And that’s not something I know how to certify someone in.
And just today I ran across a site run by the self-proclaimed World Agile Qualifications Board. And that made me angry. Really angry. Angry enough to write this post and not to link to their site. You can search them out if you want to, but I won’t drive traffic their way.
However, even if it turns out to be a practical joke, the strength of my anger surprised me.
On reflection, and being brutally honest, I realized that it’s an anger borne of fear.
I fear that the quest for certification and availability of general Agile certifications, no matter how dodgy, will lead people away from the non-certification classes and services that I, and others like me, offer.
In this economy, that could be disastrous for my business.
I would like to believe that people in this industry are sufficiently discerning that they will come to my Agile Testing classes because my classes are valuable. The kinds of things I teach are not the kinds of things that are certifiable.
How do you certify someone on the realization that they’ve been playing the hero all too often? Or on the bone-deep visceral understanding around the effects of changes in feedback latency. These are the kinds of lessons that I believe participants in my Agile Transformation simulation learn. And they are not things that translate to a certification exam.
And so I am afraid, even though I know that fear is a lousy compass.
Being afraid makes me feel pressure to offer something certification related. But doing so goes against my principles for all the reasons I’ve already explained.
Maybe the best thing I could do would be to offer a self-certification. Want a certification? Declare yourself certifiably Test Obsessed. Here’s the certificate: