Episode 436: Paralyzed by checkboxes and I'm on a "must keep happy" list
33:34 Share

IT takes more than spamming the word or so gono to be a great software engineer. This is episode four hundred and thirty six of the soft skills engineering podcast. I am your host, James is and dance. I'm your host dave smith. Soft skills engineering is a weekly advice about all of the non technical things that go into the technical field of software development, like using the right shift .

lets a ship weight. I know that word .

is the right to bullet. Is that nonsensical? Is IT just chip ltd?

Is that like an ATM machine, it's word used by a sect or belief regarded by others as empty. Something that only means something to an inside group. Is that what that is? It's like a signal that you're part of the group. Ah right. So what sagan al means, i'm an engineer.

Yeah you're trying to say it's different or unrelated, but you want to sound's smarter.

Every time I use the water dog ano, I can't help but point my hands up at ninety degrees from each other.

Now, do you make a tea, or do you make like a little cute little v with your hand?

Uh, I think maybe the easiest way to visualize IT is like the song Y M C A.

but not the make a big way of your head.

except rotate that ninety degrees in the plane of my eyes and mouth, and then you point IT forward at that, and then also rotted at ninety degrees on the plane of my nose to my spinal court.

I don't know. I lost. I don't know. How do you describe .

back yn in your own body?

I tried to do that, and I have a back in try now. Boy, I feel like maybe it's crapped across more of the culture now, but I feel like especially functional programmer really liked to that word.

Used to like rich key thing. Oh yeah, Ricky.

I like d coupling, d comp .

completing that were didn't really catch on the complexing word no.

But I do think simple made easy achieved the the biggest marker of success that any idea in software can, which is that IT became deluded into meaninglessness and now you say no, but this is simple and and IT, doesn't you you saying shibuya status .

is the ultimate science success when you go in a term phrase.

yeah, I think so. I mean, Angels been there forever.

Oh yeah, it's like a twenty years ship. Lif, yeah.

All right, let's get out here. That's not what the shows about, dave. Do you want to think our patrons?

Yes, we want to recognize or say the names or cheese. I don't even know how to describe this segment anymore. Basically, we say what you type if you give us enough money.

So here goes an example. Cousin cv ten, prince lucas morita is cool. Go to ten jokes on new lucas. I am not a basic interpreter. Nick molino A A attribute error, none type object, no attribute to string, have your consolers tue timbo. I found some cash and I back on the list, baby, become a senan engineer dot com until the different tries are morally objectionable. Then from drone deploy, chase dw norton, level up your type scope, type you out.

Dev, never is not just to create on mars from moi, I like chicken, I like liver mimics, mimics, please deliver trash panda kill boss kin c dods nevis, not displaying the vocant system Jenny kim, or in shorter these stocks tic parrot, hello icon out A I best observably tool for A I red panda is best panda type script is a microsoft conspiracy JoNathan kings and I beautiful functional user documentation IT takes more than setting a funny name on patron to make David jami's and laugh to be a great engineer, the creator of William Angel dot net. We can't actually say this name out loud. But I just did travel's britain gains john grant to joe grosberg, if you would like to join this illustrious crew.

Dave, don't close up for this bit. Maybe it's different. Go to some skills on audio. And cody sae, ah my favorite part of the show, this is A A delightful thing.

We never could have interest lately.

I honestly can't imagine how much free advertising we've given to me.

Mix time to go collect.

Wouldn't be funny if the person who wrote that patrol on name is actually on the marketing team. Over at mimics.

you've discovered a grow tech already knows soft engineers just shout down on cable to stay productive. And we love cats. We're famous for the cable and energy drinks. That's the fuel that makes good code cabled. His dog food, right?

I don't know. I actually don't have any pets.

I don't know. You know. Okay, well, thank you. Okay, thank you. Thank you. We appreciated you.

Thank you so much. The laughs. I just love IT.

okay. All right. Should I read our first question?

Yes, I am hoping you would. All right.

This is from a list or named Marcus zao berg. I don't think that's the real thing, but since when did that stop us from reading a name? They say, I work in A A mega court whose recent focus has been on reliability.

The company R T has mature. S soo, coverage low, stands for service level objective outage response standards. But my org has taken IT to the extreme this year. For example, there is now a dashboard of service health that is reviewed by engineering. Leadership in its services are marked unhealthy permanently upon fAiling a check.

Think like an H G T P slash health wrote, to achieve a quote, healthy state, one must manually explain the failure with an entry in a which must be reviewed and sign out. Increasingly, I feel this has the opposite effect, discouraging nuance to work to improve reliability and instead becoming checkbox driven development as well as impacting our ability to ship on our existing roadmap items. Additionally, how actually is fairly junior and frequently fails to communicate the organs expectations to the team leading on us to be under the gun of the reliability dashboard. Often any advice on how to make the best of the situation?

Well, this is so beautiful.

There are so many ways to screw up. Yeah, there are so many .

ways to get IT around. I just, I want to relish the image of having your service marked unhealthy by engineering leadership and then having to face the tribunal and tell them, yeah, here is what we have done, oh, great engineering leadership to restore our service to health.

This reminds me of when I worked at a mega corp, we had postmortems. They were not blame us. They were explicit blameful.

Because one of the main come of the post mortem was to assign the dollar cost of the incident to a team. Oh, so IT was like hot potato. Everybody in the room. Whoever yet allow us is like, you can like rally the troops against the team. And if you don't stick up to say no, exactly this teams s fault and IT like goes on your permanent record.

So you saying, like the best information you could come with prepared to that meeting is a sweet narrative that makes another .

team look that exactly here's why is this person's fault?

Oh my goodness.

Yeah, IT was like .

actual names or teams.

like names of people. I was teams. okay. yeah. I think they they sort of like vaguely shrugged IT blamelessness by trying to not say, like what this engineer press the wrong button.

IT did happen sometimes, but they they you could tell the kinds new like, hey, this is bad. But IT was all crushed under the the incentive of, like, do not end up with negative a million dollars on here on your team's head. Did you have like some kind of leger that tracked all the money lost by each team? I assume. So I never actually saw that part because I was too slowly in the ogg IT was usually a like high mucking mux battling IT out with each other while we would like run around between their legs, not get .

stoned on trying to get stumped. I love that the giants are fighting in the answer, running around, yeah.

yeah. But I guess the point is postmortems are good, right? It's good to talk about what happened.

And well, they can be good. I get apparently not .

all reliability is good. It's good to it's good to try make reliable services like there there yeah, good goals that just get twisted into things that make your life horrible. And this certainly feels like that.

This feels like the ultimate expression of good heart' law. yeah. And a measure becomes a target IT cees to be a measure, especially when that target is actually a target painted on your back .

where people shoot at IT. Yeah, I have a great way to make sure you never fail a health check, make a specific health checks that never goes down and doesn't actually check the health of anything. IT just returns.

Returns true. yeah. returns. Yes, this service is called the two hundred service.

Oh, why is the number two hundred? Yeah, that's the status code. Returns, no matter what.

Yeah, it's actually a razberries e under bob's desk with triple redundant, the U. P. S. Power back up.

i'm gna say about this stuff that you are no question ask her. But the ideal number of fraud incidents is not zero for most businesses. If you are an airplane, then probably zero or NASA.

Yeah, even then, I mean, like idea number of airplane crashes is definitely zero, but how about the ideal number of space utt explosions? Yeah, that's probably zero also. But if it's software, there are greater and greater costs to achieve higher and higher levels of reliability.

Assuming setting a the whole like people are just gonna lie or avoid the pain, but saying, assuming this is like you trying to influence behavior and you actually do, you don't incentivize anything bad yeah IT turns out IT. To achieve ten minds of reliability, you must never deploy any software ever again. And you must go into the the cave. And like, I don't know you have to like account for cosmic bit flips and infinity redundancy .

yeah and we only hire s aries and we don't change the .

code yeah yeah the the thing that is most commonly correlated with unhealthy services is change, right? And if you're not change in stuff, then you're not chipping anything.

And so well, I mean, if you wanted listen, if you if this leadership team wanted the rest of the all the engineering teams to obsess on server health and they're gonna get that right and and there will be trade offs, it'll come at the expensive progress. And IT kind sounds like this organization has made that choice. And I guess that's a choice that's the right to make.

It's also tRicky to have the reliability of your service. I I know they said think flash h tp health. I'm assuming it's more complex than that. But yeah, reliability is tRicky to define and measure well and is a lot of work to define and measure well.

And you can always kind of it's easy to count the number of five hundred that is codes you return yeah, but it's harder to say like for body sizes a below this thing, we expect our p ninety five to be within this range and like it's it's hard to capture the nuance of what healthy means yeah. So IT does tend the more you sume out, IT does tend to get reduce to like number of bad things, which doesn't I assume that's what they mean by discourse ing. Nuances work to improve reliability like yeah just go go every time there's an error, make IT error. That special era not happen yeah there .

is good news here. This process has no chance of sustaining itself in the long run. And the reason I say that is because the people who have to sign off, I I assume are the same people reviewing the service health, which which is just listed here as engineering leadership, which I have to assume means some kind of like VP or director level at this mega, or which is these are people with full schedules and giant paychecks.

And once someone hire up in the organization realizes that these people are spending all of their time reviewing the notes written by engineering teams to convince them that their service has returned to a healthy state, some is going to freak out and they're going to say, stop that. You, I don't pay you a million dollars a year to review the notes written by engineers to then certify their service is healthy. I pay you to strategic lead this organization. And so I I think the beauty is you just wait, this process will definitely pizzle out.

Yeah, it's interesting because I again, I can squint and see. I think the google S R. A book mentions that services can be kind of marked as as like too unhealthy and they and they don't get S R support at that point.

Like you can you can imagine a big enough org, an unhealthy enough service. You sort of want to say we're not we've seen enough. We know no matter what your dashmore says right now, like you are in a bad state and we need to change something. But again, IT feels like this has been Carried way too far and lost all, all nuance into a single bad thing happened yeah my guess .

is that at some point, the the velocity of new unhealthy services showing up on the sheet every week is going to come down. Not the least of which reason is because engineers are going to eventually gain this system, but also just because things are onna get Better over time. Like when when an organization focuses this obsessively on one thing, that thing tends to improve and in this case, service health.

Okay, fine. So it's going to improve. And then what's going to happen is one week you're going to have a dramatically smaller number of service health incidents that need to be reviewed by leadership or they're onna.

Look at the the velocity over time and see that it's come down to an acceptable level. And again, they're just gonna stop. So I I apologized in advance for repeating myself, but this process will die. You don't have .

to worry about IT yeah I think it's fair to give feedback upward to say, hey, this is making us go a lot slower but what what's an .

engineering leader going to do with that? They would be like, yeah, I know I told you to make the services as healthy its reliability to as our focus. I I knew they're betrayed us.

I think it's going to hopefully reach that point, but it's possible that someone has said we must focus on reliability. And by the way, everything does get done in the same one of time that, that could cause pain if you feel like what we have to do this and it's pushing a roman back and like nobody understands that IT will take more time to deliver on a road that .

they understand.

Hopefully that's not the case, but I don't know. I think it's fair to communicate here's how much time we are spending on this yeah and and make sure that pushed up. Because maybe maybe they expect this to be like, I don't know, a ten percent time thing, right? You just I know just sprinkling a little reliability work and and you are actually spending fifty percent .

year time on IT. I just James is new your youthful faith in this organization? It's just so refreshing.

Yeah, I have worked a company with, like, thirty people for a time. I I remember doing this, so I remember feeling like these things came down from above. This is so stupid.

This is making our life's worse. This can't possibly give us the outcome we want and complained to my boss. My boss was a good sport and they said, yeah, I i'll deliver that feedback, but I don't think it's going to change anything to see. No, they didn't shut me down. They also didn't give me false hope and IT didn't add know stuff kept changing.

I I think part of the problem is when you are leading a giant mega corp, you have like one big lever yeah and you pull the level that says like reliability and you you can go to every team and explain the nuance and make sure makes good tradeoffs. You have to pull the big lever. And then this army of people below you in the org take all these actions yeah ends. Some of those are going to be negative and some of the teams are going to be grumpy about IT, like you're just going to get ground up in the gears of this giant exactly?

Or you are .

the gears, yeah, in this case, you are being ground up. But I think the gears are the people reviewing the spread.

eh? I see, I see. yeah. Now that's more like a James on that cy cynical.

negative view. But I screen 女生， i don't know. I still do think you you can carve out some room as long as you play the game well and if the game involves making sure you don't show up on that spread s sheet and doing that as fast as possible so that you can use the extra time to do Better reliability work, maybe that's what you do, but you have to think about your bosses incentives here. And I assume given the spread sheet thing reviewed and signed off like IT looks bad to have your thing on the spreadsheet if you're the tech lead of the team lead. So you you might want to strategize with them about this.

I have an example working at a mega corp. Where we turned a bad process like this into an opportunity for career growth. And maybe if we all gather around, i'll tell the story.

Now, would you like that? Oh, sure. yeah. This why you invite the bear? okay. So we had A Q four code, like a deployment freeze that was enacted in the weeks leading up to Christmas, which this particular company was our biggest day of the year for traffic to our services.

And there was like a VP level exemption process by which you could request an opportunity to do as of the deployment. So familiar this time. Yes, I know. Are you like transplant ted back .

to yeah ah I had not thought about this in years and bringing back all kinds of things.

all good, i'm sure, wonderful warm holiday feelings and yeah yeah so we had a desire on my team to do is often deployment to change some parameters on some of our I think there's dyna db tables or something to more resilient, to Better withstand the traffic. And we have tested that we knew as good, and we had good reasons. And so we wrote those reasons in a one or two page document.

IT was a very well written document, and we took like several days as a team to help one of our t members write the dock and then iterate and provide a ton of feedback and review to make IT Better. And I still remember the team members named to this day. This was like seven years ago, but I remember his name was mike.

Anyway, mike roth, the document we reviewed the heck out of IT really made a crisp into the point. And then we put IT on the list of exemption reasons for an upcoming change, you know, committee review and the V, P, S. Read the dog very quickly, because I was short, succeed to the point, had some beautiful graphics and nice graphics.

And they just immediately approved the request, and we made sure that might get credit for writing the dog. And I remember the VP was like, great job team. Like very clear, totally understandable you approved.

And like, there you go, mike. Like hat tip to you. You just got a little VP visibility. You wrote a really clear engineering doc.

And so we turned an ugly process into an opportunity for an engineer to get little career growth in. So maybe that's what you could do here. You've got your spread sheet. You have a place where you can fill out why your service has been unhealthy. You can reflect your engineering muscles and show how amazing you are at resiliency and reliability. And really just a fantastic engineer by writing up a sweet post modem about what happened, very descriptive of very clear for an executive level audience and then get some nice reputation and .

promotion points. You saying don't write the jay song was haunted .

in the exactly? I don't know.

IT was a one time just a random coincidence, I guess I don't know.

We we never figured that out.

Yeah, I just didn't happen. And so I think, good, good. Yeah okay, that's fair.

I noticed that we took a week to work on the document. Certainly isn't ship and stuff. Yeah so that's part of the cost here.

Yes, but IT but who cares because of tons of money, that's what they want. yes. I mean, that's that's the process they have design.

yeah. IT was only good. I'm telling you IT was only good for this team. The the engineer got a bunch of experience.

The VP had an elevated opinion of our team after the fact that was all good. Now did the company actually benefit that? Who knows? That's not something we can measure.

That's why why we're here. yeah. Oh, man, I am just reminisce about the fun days of the I started in like october um and first there was a soft freeze and where you had to get, I think was like that to know who super director approval or senior director approval and then IT went to a hard freeze.

Then you needed VP or C, T. O. approval. And the assumption was if you were on the list, you had screwed up real bed to not have gone to done early. IT was not, but we did not try and write a beautiful document.

Yeah, you took lemons and turn them into portrait lemon. Just I took .

lemons and just squeeze them right .

into my mouth. okay. Well, have we answer this question? Surely we have OK.

I'll read next one. Yeah, this comes from an anonymous and says, hi guys. I'm a senior engineer at a midsize software company. The company has had a couple of high level departures recently. And during the process, I have come into the knowledge that my name is one of a handful on a list of, quote, engineers to keep happy.

I feel like this information should be of used to me, but i'm unsure on how I should lever IT on the one end is a nice to know i'm valued, but I think i'd rather be explicit told that or Better yet, received dollars in lu of prays. I'm also at the point of my career when i'm looking for staff roles in the topic of promotion has come up several times with my manager. He supports me and I believe, but we agree that, that would be difficult to make the case to the business. What do I do with this new knowledge? And is there a way to benefit from IT without accidentally triggering presentive search for my own replacement?

Good news. I think you'll be very hard to trigger preemptive search for your own reply. Yes, if you're on the list of engineers, keep happy.

You are unlikely to. If you say, hey, I need more money. They're going to give you more money. I don't know that the point of the list.

I need a fake promotion. okay. Yeah you got a engineer. yeah. I I don't .

think um there's probably something you could do, but I I think it's very unlikely that you will push for more and the result will be while you're fired. Despite our high profile losses and the fact that you're apparently very important to the study.

I mean, IT really is a great place to be. And now you have some leverage to use the word leverage. I'm unsure hand how I should leverage IT, but that is exactly the right word because most people only have leverage when they are joining a company, when the company is actively negotiating to encourage them to join.

And you have now just discovered a piece of leverage which should work in your favor. And if you truly feel like you are underpaid, this is a great time to do IT. I mean, you really can just ask for dollars and your boss has already acknowledged that you are ready for a promotion, but the business doesn't have.

I'm interpreting difficult to make the business case. I am interpreting that to me. There's just not a spot for you to do staff level things and that's a great chance for you to say, well, say, what if you can't quite get the staff title, at least we can get the staff .

dollars yeah I think about so if you're on the list of engineers to keep happy, one of the things that you can offer to them is assurances that you like, here's what will make me happy because part of what they are doing is trying to guard against risk, right? The risk of you leaving has become more expensive and more potentially expensive given the other departures.

Most importantly, they believe there's a .

risk yeah and you could even use the departures to bring up the conversation and say, hey, i've noticed all these high profile departures and I I want to stay here. I want to be effective. I feel like to cope with the added workload in my my new important role in the covering for all this stuff.

I need this to be really happy here. And and this would keep me happy working here. You can even like salo's. And I think if one of the outcomes of this is they give you a raise and they get to say your name is off the list, like not because they don't care about keeping happy, but like check kept happy, that is good for them.

Yeah, now you don't have to worry about anymore. They would be Better. Yeah yeah one letter that it's not that you want them to remove your name, right? Like you're say, James, you just want to check mark .

next to your name yeah and IT is a little tRicky because you could come off as threatening the right that doesn't feel awesome all the time, but I think you can say we can both benefit from this. You are concerned about my happiness and my willingness to stay. I will give you more assurance in return for more money.

Yes, this is A A business transaction.

If H I don't i'm try to think of the way you could praise IT. If I were to receive a raise for this amount, I would feel very comfortable saying i'm happy working here and plan to stay for a long time now .

I I don't know you agree with this, Jason, but I probably would not disclose that I have seen the engineers to keep happy list.

Yeah, it's probably gone to get your boss in trouble if they know that, assume that who leaked IT to you maybe not but whoever whoever leaked IT to you like, yes, you have put the company in a weaker negotiating position because of this. So yeah, yes, that's fair. Yeah, that's why I think you need to use the like why I was just looking around and I kind of figured based on all these high profile departures, that you can use some assurances that I wanted say.

is that really what you would .

say he was I don't know what I would say to other people who say, yeah, yeah, you just did IT feels a little. I've never been in this situation that I know. I mean.

I IT seems to me like it's so hard not to tip your hand and reveal that you know that you're on this magical list. And so I would probably try to approach this out of left field. That was not the right way to say that that feels is not the right way to say what I would say.

I would try to approach IT, look over there, give me rays, try to distract. Little slight of hand misdirection. yeah. no. What i'm trying to say is I think that you've already got a good conversation going and your boss is supportive of you getting a promotion.

Use that angle and say, you know, your boss is obviously willing to work with you here, even though the business doesn't itin have a case for exactly what you want, which is the promotion of staff engineer or to do staff level work. But hey, a race could very well be in the work, and that might just work. And that's what you said you wanted, right? I want dollars in lieu of praise.

I like to tell people, look, business has only one language when IT comes to its team members who work for the business and that money. And so when a manager says good job like that feels nice, but the business didn't really recognize you fully, the only way a business recognizes you fully is with money. That's how a business tells you what your value is to them.

And it's kind of cold and maybe little sad to kind of reduce your entire life's work down to the dollar amount, but that's business. And so if you can get the business to recognize you in the form of money, then I think that's great. That's true recognition.

This is interesting. Sounds so hartless. There are one one part with money as you can. You often don't then turn around and state everyone, hey, the business recognized me with more money yeah. So we keep that part secret.

There's yeah there's there's something like I agree, it's it's a clear marker of how much they recognize you, but I do think there's something about kind of prestige and place in the organ influence and stuff fit the business can put their thumb on the scale for. And you mentioned praise and money being nicer than praise, but IT still feels good to feel like I work at a place that people think i'm smart and they know I do good work. And that's something that you can that, that the business can influence.

And like all things that can be abused and there's I can tell you about cases where someone widely known as super important to the ogg and very impactful also is is not paid very well because they just feel really great about how important they are yeah but I don't know. I guess i'm pushing back a little. Did I don't think it's the only way and it's also very there's a word I can't think of. It's it's not it's not very continuous and you it's not easy to just like nudged IT up regularly as they recognize your value exactly. It's very discrete like every year to like chunk here is your recognition.

So or every month you well actually wait in IT. Typically, that recognition does come through on a very consistent and repeatable basis in the form of picture pager.

Yeah, I guess i'm talking about changes to your yes.

which which is true since we're all he onic tremail Walkers, right? We only detect .

I don't paycheck about .

how to be right?

Yeah, I do care about my paycheck. If you are listening and have control, stop IT.

Please keep this love because you don't want to to change in the wrong direction. It's only changes unerotic though.

Yeah, that's that's fair. Um what would I do? What I do yeah trigger a preventive search.

I don't know. I think you've give good advice. Talk your manager, push on the stafford thing, say OK. Well, what about rays? I would be happy with that if there's not a business case for, uh, promotion right now.

And I agree that you have to be delicate, but I do think you can dangle the I would be really happy staying here despite the departure if if we got this through you enough to say that exact maybe that the phrases are the word happy would kind of trigger people in b two on the nose. But I think you can it's always kind of implicit in the discussions around raises and promotions of like and this will make me help help retain me. But if you are more explicit about IT, I think that might make IT easier given your your spot on the list.

Yeah speaking of the list, IT IT is interesting that you stumble upon the engineers to keep happy list. Now I think you need to go out searching for the engineers to make sad list.

Let's called stack .

ranking at the bottom. The bottom bucket.

you have the bottom ten percent. Yeah, those are the engineers to make sad. You make them d by saying you have been, have been to measure and found wantons.

We're not going to fire you, but you're also not going to kidding any raises. So leave. Here's the door if you like. But to your own choice, pretty phony. We just think you're the worst engineering.

A painful stack rinking yeah.

And here to makes that list. I like that. I didn't even .

think of stack ranking when I said IT. Unlike all crap, that list does exist. I thought that I think I was make a joke, but it's .

real IT IT is actually real. But it's also probably populated with people that like lively, snarky reviews on poor request for something. Maybe you just got to craft the perfect code change that you know we'll just send them wild like, I don't know, add extra semi, tweak the White space a little bit in, don't do that be ever? Yes.

I think so too. Good luck. Think you're going to be great. Are in the best possible position.

yep. What can people do if they want their own questions answered?

They can go to saw skills that audio and click the ask a question, but that's the button and .

we will ask I heard in my head you are to say support I had I had to find yeah but .

here's the thing. As of this week, if you click to ask a question button and fill out our form, we will add you to the engineers to keep happy list, so we will come into your company, find the list and put your name .

on IT just to preserve our cover, though we also need to strike a deal with the company that they cannot admit that we put you on the list so you'll look for evidence of IT and you won't find IT and that's that's how you know it's working well. Let you think about that gay. Thank you for listening. We will catch you next week.