I recently read some posts on the damaging effects of interruptions and wanted to explore this in the context of my current job – sysadmin across multiple k-12/ks1-5 educational environments – and offer some thoughts on how to change things.
First off, the core educational environment itself is pretty much built around interruptions. You generally have a single teacher in a room of a few dozen kids. I’m simplifying here, but the teachers essentially have particular targets to hit each lesson (“teach this thing”, “make sure the group produces this result”, etc) and these lessons are typically somewhere around an hour long, though from what I’ve seen these lessons are shrinking in length to squeeze more different material into a day/week. Teachers have to hop between different targets every hour (or less) as the group they teach changes (the students have to similarly hop, but across entire subject ranges. Is there any surprise that, combined with the usual suspects (advertising, media, social medias’ infinite scroll, etc), young people have shorter attention spans?) Often these differing targets are in the same subject (Science, or English, or Mathematics, and so on) however different classes are at different points in the curriculum and even if some of those classes are at the same point in the curriculum (the students are of the same age) they’re often grouped by capability – some classes need more time and effort and revision of material than others, whilst the top groups delve into subjects at a deeper level.
A teacher has an hour to achieve a goal before moving on to the next goal with a different group. During this hour, they will typically outline the thing to be done or learned, then work with the class to help them get there. This naturally results in questions (interruptions) continuously. It’s in the teachers best interest to answer these questions well to ensure that individuals maintain pace with the group, but paradoxically the more questions that are asked the slower the group can move forward, as spending time explaining material again for the benefit of one means that those who do understand it already are not really learning for the time period it takes to get the one who asked caught up. Thus in my uneducated opinion, the priority for the teacher is to get the effectiveness of the first explanation of any material perfected (as closely as possible given the ability of the class) ensuring that most of the time most of the class (if not all) comprehends it to a suitable degree on the first attempt. This minimises interruptions and enhances learning for all, as the class learning velocity is kept high.
This problem is outside the scope of what I want to talk about but it is relevant, because one side effect of this we-must-acknowledge-interruptions behaviour is that I feel it becomes habit. Being interrupted a lot becomes the norm, and it is my opinion that this encourages the teaching individual to also be the one interrupting others more.
Let’s go back to the classroom. You have one hour to teach a group of 33 twelve year olds about some algebra. You fire up your laptop and switch the projector on to find that no matter how you try you can’t get an image to appear on the board. This unplanned interruption has immediately taken up brain power and, critically, time, even if you have backup plans for each and every lesson. As you have chosen to use the technology resource you have clearly decided that it is important, so we can assume that without that resource your teaching and the students learning will be less effective. You are after a fix for this problem quickly. The quality of your teaching is likely somewhat diminished and the longer this goes on the worse this gets.
So you do what anyone in your situation does – you try and get it fixed. This is where tech support comes in. You make a phone call or send an email to that tech you like, or if you’re a hero, log a ticket on the ticketing system. The hero did the right thing – logging a ticket. No complaints from me there. (Pro tip for techs reading this: make every avenue of communication a ticket generating event!)
However emailing an individual or phoning interrupts the support techs. This is often warranted and is always understandable, and your job is a constant stream of interruptions so one more won’t hurt, right? This is where, unfortunately, tech support suffer. We operate in the opposite universe. Your tech problem, as a teacher, is probably one of your biggest current problems. And we totally get that. Promise. We really do. But… it probably isn’t our biggest problem and is almost certainly not something we’re going to want to deal with right away.
An interruption to us does not progress anything, it in fact stops everything. If we are in the middle of resolving another problem, being forced to stop for a reason (whether it’s to deal with a major issue or respond to a phone ringing or being pulled out of the moment by a name being called) causes us to disconnect from “the flow” and at best mentally change gears, at worse slow the brain CPU way the heck down. This is jarring and as evidenced by the many articles published about interruptions can in fact be damaging, both mentally and economically.
The more this happens the less effective overall an individual or a team can be. I argue that an ops team lead (Helpdesk Manager, IT Director, etc) should put work into minimising interruptions for the betterment of the team and yourself, regardless of your vocation or the objectives of the team. Here’s what I propose you can do to help within your tech support team regardless of work environment… Though you will likely need managerial sign off on some or all of these. It’s worth noting that I didn’t come up with these, they’re merely an amalgamation of things I’ve learned and tested which have worked for me. I am assuming you’re part of a team instead of solo hero, too, though if you are solo then interruption reduction could potentially save you from burnout. Some of these might help.
Everything should (automatically) log a ticket
Make it easy to log tickets. Saying “no ticket no fix!” feels good but doesn’t actually achieve anything for the organisation, except to piss off someone with an issue (and let’s hope that person isn’t a C-Level.) Let people call, let people email, let people visit, let people knock on the door. You can more than likely automatically create tickets in your ticketing system that come through to email, and if you can’t, get a better ticketing system. Get a generic “techsupport@” inbox set up and configure your helpdesk to add everything sent to it as a ticket. Bonus points of you reply right away (automatically of course) telling the requester that their ticket has been logged. At least they know it’s in the queue instead of sitting unread in some mystery inbox.
Ten CC’s of Triage, STAT!
Yikes, you now have a flood of tickets. This looks bad. And hey, maybe it is, but at least now you know it’s bad instead of it just feeling bad. Maybe management would be surprised to see that you’ve got 143% more tickets than you last reported, because everything is a ticket now. No more invisible work.
Anyway, I digress. Triaging tickets is essentially reviewing a ticket and deciding if it’s a priority or not. There are many ways to judge this, but typically you want to take into account how urgently this needs to be fixed and how much of an impact it would have if it didn’t get fixed.
A PCI Compliance audit deadline of two hours ago is pretty urgent – it’s something that needs to have happened already. But… the banks aren’t going to block all transactions immediately. Business doesn’t stop because you’re a few hours late on submitting a self eval form (this is not an endorsement to delay PCI compliance!)
A projector in a classroom has a pretty high impact – that’s 30+ students (plus one pissed off teacher) per hour per day. That adds up quickly. And yes, it’s pretty urgent too, but worst case the teacher can still teach like they did in 1985. Right?
Triage is important for two main reasons:
- It helps you and your team decide on what to work on without having to decide on what to work on
Generally, you work on the thing with the smallest SLA time left, and if you have negative SLA counting down then you should probably get on that ASAP. Assign the Priority (the urgency and the impact) an SLA (say, high impact high urgency = 4 hour SLA, low impact low urgency = 14 day SLA – whatever you decide will be unique to you and the organisation and likely requires understanding managements expectations as well as your teams abilities) - Reporting, evidence, promotions, wage increase, redundancy protection, efficiency gains, areas of focus and more!
Having all this triage data to look back on will highlight where things are working and where they’re not, but bigger picture style. You can use the data and the knowledge that you’ve gained to more closely align the team with the objectives *and reality* of the organisation. Perhaps when triaging your tickets you could also categorise them – hardware, software, licensing, accounts? Or go one level deeper – projector, laptop, desktop, TV, printer. Office, Web browser, MIS. After time passes you can really easily see where your team spends most of its time and where resources (like money) can be allocatied to reduce load in those areas.
Triaging tickets is generally not a full time job, so you could make this person…
Hero Assignment: Interrupt Shield
Designate an “interupt hero” – this can be one person or more and ideally they would triage tickets in the queue too. If you’re lucky enough to be able to do this, put them in another room and leave everyone else in the relative peace and quiet of the main office. Make everyone else go to this smaller interrupt room for direct in person support.
They deal with all phone calls and in person visits and group emails. They also deal with any quick support calls (when not on a phonecall) that don’t require them to leave their desk and that are interruptable, for example password changes, group membership updates, etc. You should rotate this person very regularly so you’re not moving the “interrupt problem” to one person all the time, as that’s a recipe for disaster. Let them do a day at a time, or two days if you use a hero team, rotating half the team out every day so handover and “current events” knowledge transfer (what were yesterdays big issues? What’s on the radar today?) can occur more naturally.
Redirect everyones incoming calls to the interrupt hero team. This further reduces interruptions to the main body, though you may want to make exceptions for VIPs. Your interrupt hero team should be able to forward calls to you, but generally they will be able to deal with most quick things (and push to the focus team via a ticket if not, of course)
This has two advantages. First off, the toil is reduced for the team at large – your senior sysadmin isn’t doing the third password change this week for little Joe – and secondly (and most relevantly) the rest of your team can focus for long periods of time on their work. Better work is done by most of the team. Better work is better results (whether those results are money/profit or a higher quality of education.)
Don’t overload on active tickets
Don’t assign anyone (including yourself) too many tickets. There is no perfect number – it’s different depending on role, personality, type of work, and many more factors. But try and get a ballpark. I would suggest starting with no more than 10 non-pending tickets (for clarity, I define a pending ticket as one which is waiting for something outside the control of any member of the team) and only one should be worked on actively at a time. You can’t install a printer whilst also diagnosing wireless issues. Pick one. Focus.
Low remaining (or negative/expired) SLA first
Plucking tickets from the queue because they look easy isn’t productive, long term. But it’s fine, short term. Don’t sweat it, manager. But keep an eye on it.
Tickets should be worked based on their SLA, accounting for the available suitable resources (don’t put the guy who sucks at DNS problems as the lead on critical DNS problems – train them? Yes. Rely on them? No.)
Sometimes it’s nice to jump 70% down the queue and grab a few easy tickets. It feels like a break.
Just don’t make it a habit.
Having repeat logs of the same call that has gone unanswered for three months is – you guessed it – yet another interruption. Old (the definition of which is, essentially, defined by your SLA – expired SLA? Ticket is old!) tickets should not exist (with very very few exceptions. Like budget availability.)
Don’t judge people on their call closure rates
How many tickets I’ve closed vs the other techs means absolutely nothing. Get honest feedback from the source – ask the staff and, yes, the students, to rate the support. Then pay it attention. This feedback is so valuable. Whenever a ticket is closed, automatically send a followup email asking for feedback, and make it quick. You want the emotional immediate feedback, not the “dwell on it for three days and [calm down/forget]” feedback.
- Was the support: [] Amazing! [] Ok! [] Shit!
- Optional comments here:
Negative feedback is the most valuable information you can possibly obtain
I argue that getting positive feedback means absolutely nothing, except as a “look how great we are, CEO!” line in the yearly review meeting. The feedback you really want, the feedback with actual damn value, is the negative feedback. And it’s all well and good getting it. But make sure to pay attention to it, don’t let it sit in the database doing nothing. Analyse it. Understand it. This is the closest you’ll get to dipping your hand in the metaphorical technological currents of your organisation. You’ll be able to feel the effect of the teams work, and when something goes against the current, you can fix it immediately.
This makes better, happier staff and students. Better happier staff and students equals better results. Which equals… you guessed it, fewer interruptions.
Interruptions aren’t going away, but you can reduce them. And when you do, that firefighting feeling will decrease, work quality and rate will improve and as a direct result the stability of the environment should improve, too. And when the stability of the environment improves, the interruptions decrease.
At least, that’s my experience.
There’s plenty more about this on the internet at large, much better explained and with additional steps than what I have listed here, however I feel that these steps are the big hitters that might just give you enough time to step back and take a real close look at the bigger picture without worrying about where the next fire will start. And it’ll take time. But it will work.
