![cover](/assets/images/blog/2023/12/13/00-cover.png)

# AGILE IS AN OXYMORON

### It's been a while since I've posted a good tech related rant but this one I cannot keep inside any longer.

I've worked in tech for decades now. I went from a self-taught teenager hacking away in my bedroom on my own apps, to junior software developer, then senior software engineer, all the way up to principal software engineer and eventually founder and consultant of my own tech company. During that time I've seen everything and I've been exposed to every different type of technology and way of working that came and went along the way.

I'm old enough to remember when Agile seemed like a good idea and I'm bitter enough now to tell you that it was the biggest mistake we made as an industry.  As with most good things eventually corporations latch onto it, misunderstand it, abuse it and ruin the lives of every once passionate dev along the way.

I mean it I look at our industry now and I think it's a joke.

Agile when used correctly inside true tech companies that understand the development process can be great. I'm no fan of Scrum as you'll discover reading this post but even Scrum when done properly can be very effective. I've worked in many mature product teams that iterated quickly and delivered working software to end users and stakeholders at the end of every sprint. When it's like this Sprints really do feel like sprinting. Sadly I've not seen this for many years now as I tend to consult in large enterprise organisations where Agile is a buzzword and has absolutely no semblance on the day to day operations. Companies seem to think they need to say they are Agile, they seem to think that Scrum is something they should be doing in perpetuity completely missing the point of Scrum, and are so paralysed by process in the giant enterprise machine that everything either grinds to a complete halt or demoralises and burns people out completely.

It's my honest belief that most big organisations just don't know how to do Agile. They are too big they are usually too set in their ways and the corporate structure and bureaucracy crushes any chance of being truly Agile. These companies instead pick and choose which bits they think applies to them and discard the rest like it somehow wasn't important and doesn't matter. Most of the work planned in by these types of organisations doesn't fit with Agile and usually doesn't have any actual requirement to be Agile. It's usually a lot of design up front and/or hard deadlines, everybody knows it's waterfall but we must all join in and pretend that it's Agile for the sake of being Agile.

They then bring Scrum into the mix to really slow things down. Meeting after meetings because it's "Agile" that do nothing except reduce the time available to actually build working software and cause devs to context switch and come out of their productive focused zone.  Countless wasted hours trying to come up with futile estimations that do not matter in any way AT ALL because the work needs to be done either way and everybody knows it. This leads me to one of my biggest bug bears with out industry today:
### [Poker Planning](https://pokerplanning.org/)

Our industry has so much delusions of grandeur that we invented an entirely new way to estimate work rather than just JFDI. Honestly I think it's bonkers that we've invented an entire game just to estimate a piece of work. The only rule of Poker Planning is that "time is a naughty word" but we have 2 weeks in the sprint to deliver work and we need to fit as much as we think we can do in those 2 weeks, but remember work doesn't take time it takes points. WTAF.

Oh and the best thing about this is we've obediently played the Poker Planning game for years now, we banish the concept of time from our minds and we shout out numbers that represent the complexity of the work at hand. We then fit what we can into a sprint and then we deliver to a deadline that is usually decided by some outside force, either management or an overly aggressive salesman, definitely not by anybody that was close to the work itself or had any understanding about the complexity or risk involved in said work.

Seriously Poker Planning is so retarded it genuinely hurts my soul and if these companies would just get away from the dogma of insisting we do the Scrum dance they'd see how entirely pointless it is.

I've worked in teams where we regularly did [Mob programming](https://www.agilealliance.org/glossary/mob-programming/) where we practised #noestimates.  It's amazing how much more productive a team is that has zero meetings and doesn't needlessly waste time trying to estimate how long the work will take and instead just does the work, but I digress.

Even if we don't go as extreme as this there are better ways of being "Agile". [Kanban](https://www.atlassian.com/agile/kanban) for example is all I believe companies really need. As long as there is a list of work that needs to be done and some general agreement about which work is highest priority then what more do we need? Just work on the thing that is most important and concentrate on delivering it. Once it's delivered work on the next most important thing and deliver that.
Scrum is just Kanban with extra steps and these extra steps suck all the enjoyment and enthusiasm out of a tech team faster than anything else I know.

So why do companies insist on Scrum if there are other alternatives that waste less time, deliver more working software and don't generally make the devs want to hang themselves? CONTROL. Corporations especially big ones cannot let go of that control. They must micro-manage everything that anybody tries to do to the nth degree until they've ensured that every deadline has been missed, every budget has been exceeded and every once happy engineer is questioning the terrible life choices they made to get here. I can prove this is about control and not an attempt to better set peoples expectations too. You only have to refer back to some of the parts of Agile that corporations like to conveniently ignore. Burn down/up charts.

Now we all know they love to track all these numbers that we put on estimates and salivate whenever it's time to review a good burn-down chart but this proves my point. True tech people know that those things aren't meant for middle managers. They are supposed to stay in the team. They are an artefact for the team to learn from so they can adjust their own expectations and try to improve things. THEY ARE NOT MEANT FOR MANAGEMENT TO USE AS A STICK OR AS A TOOL TO CONVINCE MANAGEMENT THAT PEOPLE AREN'T SLACKING OFF.

Here's a little secret: Empowered, autonomous engineers don't slack off. They enjoy the work that they do and feel proud when they deliver.

Want some more proof that control is more important than effective, motivated software teams that deliver value regularly to customers? Just look at the tool that we're all forced to use:

### [JIRA](https://www.atlassian.com/software/jira)

This software is slow, clunky, never flexible enough to feel like it complements your workflow and above all else...[HATED](https://ifuckinghatejira.com/) universally across the board by literally every single software engineer on planet earth.

How in 2023 in the tech industry have we settled on a tool this bad that doesn't even let you assign more than a single dev to a ticket? In all the best organisations and software teams that I've worked in, pair-programming (at the very least) has been a standard way of working if not mob programming.  JIRA is forced on everyone for one reason and one reason only. It lets Scrum Masters and other management types track meaningless numbers and gives them that gooey feeling inside when they can look at a burn down chart. The problem is the "board" isn't the work, the work is the work but the work becomes secondary to the board. Things change during a sprint, you don't always end up working on the things you thought you would a week or so ago but heaven forbid we move things around and bring new tickets into the board to reflect the work we actually had to do, oh no, that would affect the managers view of the success of the sprint.

You see the problem?  What are we doing here?

We've adopted the name Agile and somehow we've ended up implementing the least Agile and most Rigid way of working possible. We have layers and layers of process that we refuse to address. Nothing ever changes once you get into this mess. Instead of just trusting expert engineers to collaborate and work on solutions and ship working software on time we just demoralise everybody instead and make sure that nobody can be effective.

It's not just Jira either. Organisations usually like to top-down dictate the tools and processes that they want to apply across the organisation. This to me again couldn't be further from Agile if it tried. What happened to "people and conversations above processes and tools"? What happened to letting the experts on the ground collaborate and choose the correct tool for the job? Instead one bad choice early on instead leads to many more bad choices and this never ends.

In companies that empower smaller teams to truly innovate on the projects they own I always find that more collaboration happens naturally and better solutions result.

Another big reason why I think Agile has failed as an experiment is Scrum Masters. These people that have usually not come up through the ranks as software engineers just go off and do a course and come back with a head full of theory on how software engineering works. This to me is no different than a diversity hire. I, like all the best software engineers and software engineering manages I know work best when I'm working with fellow experts.

Reminds me of a quote that my good friend Luke used to like back in our MediaCom days...

> **In theory, there is no difference between theory and practice; but in practice, there is.** - Albert Einstein

---

In truth I don't know what the answer even is to this situation that we've all stumbled into. Every big corporation you walk into now has the exact same set up, makes some combination of the same mistakes and sprinkles in a few of their own that come from their unique brand of corporate demons in the cupboard.

I think Steve Jobs had the right idea though.

> **It doesn't make sense to hire smart people and tell them what to do; we hire smart people so they can tell us what to do.** - Steve Jobs

I truly hope that as the next generation of software engineers comes through with all their fancy new AI and tooling that real change can happen in our industry. I don't see youngsters having much patience for the bullshit that we currently have to endure in the industry. They move fast and deliver innovation and I would be willing to bet most of them haven't heard of a JIRA board.

/rant 

OUR INDUSTRY IS DUMB

Did you enjoy this post?️

You Might Also Like

WARP!

THEMES ARE HERE