You need to focus on the real problem at hand, and devise/adapt a correct solution in context. Systems and frameworks provide helpful tools and templates, but they all require contextual adaptation.
FWIW this is what I like about the foundations of Agile for project management. The manifesto points all direct you away from your tools and processes and towards the actual problem. eg "People over process," "working code over documentation," etc. Basically the people on the team should focus on working software, communicating with the customer, and adapting to change. Their processes, documentation, contracts, and plans should all be aligned to that solution, not the other way around.
And out of that... we got millions or scrum certificate weilding alcolytes who will tell you that if your project failed, it's because you didn't do their system religiously enough. God certainly has a sense of irony.
It's more like ticking boxes and pretending to know what they're doing than actual agile project management.
So now we're stuck with a lot of overhead crap and even more workload. And the system actually slows us down when we try to adapt to a changing environment.
Problem is mainly that this happened because the director of project management really liked the reports from Jira and thus the main reason for it all is to get as much data as possible in there (notice how 'accurate data' does not seem to be a focus :) )
When you’re doing something for the sole purpose of keeping someone off your back about it then you inevitably half-arse it, so in reality we’re paying Atlassian a small fortune for nothing more than a wall of fairly bullshit estimates. Don’t even get me started on how slow and flakey the UI is on the cloud version, it’s such an intolerably poor quality bit of software to actually use as a dev. I’d be very surprised if Atlassian actually dogfooded the cloud version because it really is an abysmal effort.
And yeah it's slow but not super slow. Pretty sure we have onprem though. Because I can only access it on VPN. Unless they fiddled with some port forwarder which I wouldn't put past them.
So as long as you define what "leadership" or "customer satisfaction" (and the rest) mean to you, and you have a vaguely plausible but not necessarily effective process that ticks each of the boxes, you have a quality management process. And you can apply for - and will probably get - formal quality management certification.
In reality you can have zero actual leadership or customer satisfaction. But you have a process - so OK.
This isn't to say that we should just do away with all process. There are good, even life-saving reasons why processes exist, but a process alone is no cure-all, and an overdose will paralyze us.
- Abstractions/models tend to completely replace reality. We, software engineers, are very prone to that sort of thinking. Another prominent category are accountants: leaders drive their businesses only looking at the "map" of accounts; governments do the same with their constituents.
- "Paralyze" is exactly the right term. Excessive bureaucracy actively discourages any creative action, for the sake of illusion-of-control.
Cannot upvote this post high enough, you made my day. Maybe humanity is not completely hopeless yet.
It was the first time I'd read ISO 9001 documentation end to end, and it was eye opening.
Total bullshit.
It ensured there was a paper trail documenting the total bullshit, of course, but it nevertheless was total bullshit that said a lot about describing what was done, but nothing about ensuring it actually worked as it should.
This same place had locked doors to the interior stairwell so consultants like me could not even get into areas we were not assigned to because of work done for clients like the police, but had confidential documents about police infrastructure sitting unprotected on the machine I was assigned.
To me that was a good indication they ran everything including security how they handled quality management: it was just a box ticking exercise. Document things and track deviations rather than care about actually getting things right.
I recently read this about it:
https://essenceofbuddhism.wordpress.com/2016/04/19/what-the-...
I've taken to using tools that most closely mimic notepad+pen or whiteboard+marker. Increasingly I am finding that a physical whiteboard, markers, and post-its are good enough. And if you need to create backups, the image quality on any smartphone is very, very good these days.
The more time spent working on the thing that manages your work, the less you are working!
Often questionably necessary, middle management has a hard time doing something when engineers understand the product and upper management has the right balance of setting direction and entrusting decisions to the makers. So they start asking the engineers to do things and get upper management on the train of wanting reporting. Doing that with pen-and-paper engineering and people with context is HARD... so all of these tools exist to make this middle layer of doing work better.
And arguably unnecessary people earn a lot of money by convincing themselves and others that they are necessary, so companies get to be organized to have these high overhead tools to support this high overhead style of organization.
Now to be honest some level of organization is necessary, and becomes necessary as the enterprise and product grows in complexity, but very very few people work long enough or with a diverse enough set of organizations to be able to see what is necessary and when and what is not.
If you hire people and put them in a position and tell them they have to work a certain number of hours they will find or create work to do. Oftentimes this work is not better than doing nothing but the WASP work-ethic and general cultural norms really can't handle the idea that smart, important, useful people can provide the most value sometimes by not doing anything until they are needed.
Thank you for this comment. I have had this is mind for quite some time but I never could phrase it in such a clear, articulate, phrasing.
Hardware on the other hand, needs an organization layer on top of the production layer. Removing barriers, creating traceability, improving processes, and trying to get multiple different teams to work together efficiently.
But that would require a lot less gatekeepng and other artificial moats around them, and possibly a rather more open organisation model which wouldn't fly with middle management (even if your specific adjacent middle managers would like to, they don't operate in vacuum). And that's before you hit cost reasons. Or cost of training, especially since companies really, really don't like training people.
Take away is that just because they're not using an off-the-shelf project management / WIP management tool, doesn't mean they're not organized.
Of course, that’s not feasible but hey, a guy can dream.
The whiteboard was good to move forward a discussion, explain concept and other niceties. To organize work ? Hell no.
> Leaders operate at all levels, stay connected to the details, audit frequently, and are skeptical when metrics and anecdote differ. No task is beneath them.
IMO this particular trait correlates well with a need for less process / project management software. I really believe if more people did this companies would be better off.
Note: this isn’t the same thing as micromanaging but I understand why some people think it is.
Also the dev stack was horrifyingly opaque. Every time an Amazon recruiter contacts me I have to push down the urge to explain how awful it is to work there and no thank you.
It’s something that has frustrated me, as a manager, when fellow managers said it to me. I feel like it gives permission to managers to detach themselves from the reality of their organization.
The trick is to know WHEN to dive deep, because you are right that managers also have to maintain a broad perspective.
But being able to sniff out problems and then dive into them has always been the hallmark of a great manager/executive, at least in my own experiences.
You seem to imply that UML or other project management software are considered best practices. I have never heard that. Or rather the opposite. Not using UML is currently considered as best practice.
But maybe it depends who you ask.
However it's way too laborious using it for everything and it becomes a burden when things get complicated. For simple things it just doesn't really help because they're simple anyway.
As a learning tool I think it was great. In production? Well I use it every day. The book props up my keyboard on my desk which happens to be 2cm too low :)
Having a shared vocabulary is great.
Modelling everything out with it, on the other hand, is a nightmare.
When such a rant actually feels sensible and pragmatic, there must be something seriously wrong.
I don't mean actual engineers wringing bug free code for a space telescope attitude control system. Seemingly they have it figured out - their code does not seem to fall down at a higher rate than bridges or office buildings. They have a method.
But this method, whatever they implement, is surely too costly for us regular chumps, at least provided your employer, as you put it, has competitors.
Also, most best practices peddled are supported only be flimsy data and anecdotes - only very few areas that are different. I have seen some examples in medicine that had a lot data behind them (not all of medicine by a long shot, though).
Agile was invented by consultants who's agenda was to maximize their profits first, not just to improve SDLC.
Also, there is no silver bullet, and everything is a trade-off, so each best practice has to be adapted to a particular project.
Last thing: sometimes any process, even as bad as "Agile", is better than no process at all. Especially for the dead-on-arrival cases like government-related projects.
--
1. SDLC - https://en.wikipedia.org/wiki/Systems_development_life_cycle
And I've seen it done well.
> This brings back memories from the days of my early career (an ex-PM survivor here). I would be curious to see some data, even anecdotal, on the success of this approach. Here's some interesting statistics from the industry (not specific to software, but you can extrapolate): http://apepm.co.uk/project-management-statistics/
> In my view, traditional software project management is ineffective. I would put it somewhere between the Myers-Briggs personality test and modern day astrology.
Commendable amount of spaghetti's mentioned though.
There's a high correlation for me between bad managers and managers that think they can get you to use a tracking tool that provides zero benefit to yourself and gets used to do their job for them and also occasionally to pick favorites, lay people off, or deny raises/promotions. You're dealing with smart creatives, nobody is volunteering for that shitshow.
You got a tool that helps me keep track of my day or be prepared for meetings? Maybe. But if I have to stop what I'm doing to use it, then no fucking way.
I have a somewhat similar stance on software architecture. Sure, I’m TOGAF certified but if you actually expect me to work with notation correct Archimare/UML/whatever and not just draw boxes on a napkin you’re either insane or sell a product with no competition.
This is anecdotal of course, but after 30 years of working with hundreds of companies to supply our municipality with software, we have yet to see a correlation between quality and “best practices”. It doesn’t matter if companies do this or that testing and use the full confluence suite or if they just pull spaghetti out of their asses and support is a phone call… the quality is the same over time, hell, often we get more from the spaghetti companies than the “best practice” ones. You might think that it’s a short term thing and that the spaghetti and no testing catches up, but it doesn’t. Maybe because we don’t have to fund those 8 people, that you never really learn the role off but it sure isn’t technical, that sit in on every meeting as opposed to talking directly with the spaghetti slinger and a sales person? I’m not sure, and I’d love to tell people that following this or that “best practice” is the way to go, but that’s just not what our data shows.