Who’s to blame for bad code? Coders.

by Jason Swett,

When time pressure from managers drives coders to write poor-quality code, some people think the responsible party is the managers. I say it’s the coders. Here’s why I think so.

The ubiquity of time pressure in software development work

I’ve worked on a lot of development projects with a lot of development teams. In the vast majority of cases, if not all cases, the development team is perceived as being “behind”. Sometimes the development team’s progress is slower than the development team themselves set. Sometimes development progress is seen as behind relative to some arbitrary, unrealistic deadline set by someone outside the development team. Either way, no one ever comes by and says to the development team, “Wow, you guys are way ahead of where we expected!”

Often leadership will beg, plead, cajole, flatter, guilt-trip or bribe in order to increase the chances that their timeline will be met. They’ll ask the developers things like, “How can we ‘get creative’ to meet this timeline?” They’ll offer to have lunch, dinner, soda and alcohol brought in for a stretch of time so the developers can focus on nothing but the coding task at hand. Some leaders, lacking a sense of short-term/long-term thinking, will ask the developers to cut corners, to consciously incur technical debt in order to work faster. They might say things like “users don’t care about code” or “perfect is the enemy of the good“.

Caving to the pressure

Unfortunately, in my experience, most developers, most of the time, cave to the pressure to cut corners and deliver poor quality work. If you think I’m mistaken, I would ask you to show me a high-quality production codebase. There are very few in existence.

I myself am by no means innocent. I’ve caved to the pressure and delivered poor-quality work. But that doesn’t mean I think it’s excusable to do poor work or even (as some developers seem to believe) heroic.

Why bad code is the fault of coders, not managers

Why are coders, not managers, the ones responsible for bad code? Because coders are the ones who wrote the code. It was their fingers on the keyboard. The managers weren’t holding a gun to the developers’ heads and demanding they do a bad job.

“But,” a developer might say, “I had to cut corners in order to meet the deadline, and if I didn’t meet the deadline, I might get fired.” I don’t believe this reasoning stands up very well to scrutiny. I’d like to examine this invalid excuse for doing poor work as well as some others.

Invalid excuses for shipping poor work

Fallacy: cutting corners is faster

Why exactly is “bad code” bad? Bad code is bad precisely because it’s slow and expensive to work with. Saying, “I know this isn’t the right way to do it but it was faster” is a self-contradiction. The exact reason “the right way” is the right way is because it’s faster. Faster in the long run, of course. Sloppy, rushed work can be faster in the very short-term, of course, but any speed gain is usually negated the next time someone has to work with that area of code, which could be as soon as later that same day.

Fallacy: “they didn’t give me enough time to do a good job”

If I’m not given enough time to do a good job, the answer is not to do a bad job. The answer is that that thing can’t be done in that amount of time.

If my boss comes to me and says, “Can we do X in 6 weeks” and my response is “It would take 12 weeks to do X, but if we cut a lot of corners we could do it in 6 weeks” then all my boss hears is “WE CAN DO IT IN 6 WEEKS!”

Given the option between hitting a deadline with “bad code” and missing a deadline, managers will roughly always choose hitting the deadline with bad code. In this way managers are like children who ask over and over to eat candy for dinner. They can’t be trusted to take the pros and cons and make the right decision.

Fallacy: cutting corners is a business decision

It’s my boss’s role to tell me what to do and when to do it. It’s my job to, for the most part, comply. As a programmer I typically won’t have visibility into business decisions the way my boss does. As a programmer I will often not have the right skillset or level of information necessary to have a valid opinion on most business matters.

This idea that my boss can tell me what to do only applies down to a certain level of granularity though. I should let my boss tell me what to do but I shouldn’t let my boss tell me how to do it. I care about the professional reputation I have with my colleagues and with myself. I think there’s a certain minimum level of quality below which professionals should not be willing to go, even though it’s always an option to go lower.

Fallacy: “we can’t afford perfection”

I wish I had a dollar for every time I heard somebody say something like, “Guys, it doesn’t need to be perfect, it just needs to be good enough.”

The fallacy here is that the choice is between “perfect” and “good”. That’s a mistaken way of looking at it. The standard mode of operation during a rush period for developers is to pile sloppy Band-Aid fix on top of sloppy Band-Aid fix as if they desire nothing more strongly in life than to paint themselves into a corner from which they will never be able to escape. The alternative to this horror show isn’t “perfection”. The alternative is just what I might call a “minimally good job”.

Fallacy: “if I don’t comply with instructions to cut corners, I’ll get fired”

This might be true a very very very small percentage of the time. Most of the time it’s not, though. It’s pretty hard to get fired from a job. Firing someone is painful and expensive. Usually someone has to rack up many offenses before they finally get booted.

Fallacy: “we’ll clean it up after the deadline”

I find it funny that some people can make this claim with a straight face. Fixing sloppy work after the deadline “when we have more time” or “when things slow down” never happens. Deadlines, such as a go-live date, often translate to increased pressure and time scarcity because now there’s more traffic and visibility on the system. Deadlines are usually followed by more deadlines. You never get “more time”. Pressure rarely goes down over time. It usually goes up.

Fallacy: “I would have done a better job, but it wasn’t up to me”

Nope. I alone am responsible for the work I do. If someone pressured me to do the wrong thing and in response I chose to do the wrong thing, then I’m responsible for choosing to cave to the pressure.

Fallacy: “I haven’t been empowered to make things better”

It’s rare that a superior will come along and tell me, “Jason, you explicitly have permission to make things better around here.” The reality is that it’s up to me to observe what needs to be done better and help make it happen. I’m a believer in the saying that “it’s easier to get forgiveness than permission”. I tend to do what my professional judgment tells me needs to be done without necessarily asking my boss if it’s okay. Does this get me in trouble sometimes? Yes. I look at those hand-slaps as the cost of self-empowerment.

Plus, all employees have power over their superiors in that they have the power to leave. I don’t mean this is a “let me have my way or I quit!” way. I mean that if I work somewhere that continually tries to force me to do a bad job, I don’t have to just accept it as my fate in life. I can go work somewhere else. That doesn’t fix anything for the organization I’m leaving (bad organizations and bad people can’t be fixed anyway) but it does fix the problem for me.

My plea to my colleagues: don’t make excuses for poor work

I think developers make too many excuses and apologies for bad code. (And as a reminder, “bad code” means code that’s risky, time-consuming and expensive to work with.)

It’s not realistic to do a great job all the time, but I think we can at least do a little better. Right now I think we as a profession err too much on the side of justifying poor-quality work. I think we could stand to err more on the side of taking responsibility.

Leave a Reply

Your email address will not be published. Required fields are marked *