Article

A practical approach to on-call compensation

Picture of incident.ioincident.io

Asking engineers to be on-call is usually a tough sell.

Think about it: if someone asked you to add even more to your already packed workload, that would be a difficult proposition to say yes to. And that’s before you mention that this work typically happens late into the day and even (some) sleepless nights.

Companies need to have an on-call function to keep their services and products running smoothly—it’s practically a non-negotiable at this point. But it’s important to acknowledge that no one really wants to hold a pager and sign up for extra work.

So, how do you turn on-call from the thing nobody wants to do, into the something people see value in? There’s many answers to this, but having a solid approach to compensation, and being able to articulate the value and growth opportunities are two of the most impactful.

Here, we’re going to outline some practical things you should consider when visiting on-call compensation and the incentives you create around it. We’ll also share how we approach this conversation here at incident.io.

💡 One quick note: nothing about on-call is one-size-fits-all, so while we’ll be sharing some approaches that work for us, consider the stage of your company to decide what makes the most sense for you.

Being on-call presents a great opportunity–talk about it!

Before jumping into on-call compensation, it’s worth highlighting that there are a few benefits to share with folks that are holding the pager. A monetary reward can and will play a huge role in getting folks OK with the responsibility, but there’s a huge opportunity here beyond that.

Developing operational skills to become a better engineer

When folks are on call, they’ll run across several issues that they’ll need to fix with stopgaps that, while not ideal, get things back into a working state quickly.

Put another way, engineers dealing with incidents need to learn how to address incidents with fixes that require the lowest effort but have the highest impact. When mitigating an incident, speed is often more important than technical elegance. This is a muscle that’s rarely flexed during regular work hours when they have the benefit of time (and clear eyes) for more permanent fixes.

This situational awareness helps folks become better engineers in the long run since they’ll be much better at figuring out how to debug and fix things quickly and efficiently.

It’s a chance to better understand customer pain points

Another benefit of being on-call? Folks holding the pager will get much closer to customers and better understand their pain points. By dealing with incidents that customers are already experiencing, folks on-call will be better able to speak to the needs of your customers and, in turn, build better and more resilient products.

The single place you can turn to when things go wrong

At companies where engineers have lots of touch points with customers, like incident.io, this can be a major plus and a huge incentive.

On-call compensation

As we’ve said, nothing about on-call is one-size-fits-all, so here we’re going to detail a few different compensation structures you’ll come across. Remember, consider what works for you when deciding which to run with.

No additional pay

This is actually more common than you think. Many organizations, especially smaller ones, don’t offer any additional compensation for on-call. This can be due to a number of factors, but usually it boils down to the fact that smaller companies have less severe incidents, so the burden of being on-call isn’t considered significant.

But it’s also something seen in more mature organizations that never set up a compensation structure and are just sticking with the status quo or just factor the responsibility into salaries (more on this in a bit).

Context means everything and each situation is unique. The size of your business, your product, where you’re located, and so much more will dictate how you pay folks to be on-call.

It goes without saying that, if you choose this route, you should regularly check in on how your team is feeling.

On-call, whether the pager is going off or not, is a significant burden. Life happens. And it’ll get to a point where folks are so overwhelmed by things going on at home that they’ll feel like being on-call should come with some compensation. Down the road this can lead to a lot of spitefulness and frustration.

Part of a base salary

Another option is bundling on-call pay with your employees’ salaries. On the pros side, having on-call compensation be part of someone's base salary removes any complicated hourly calculations. It also makes it easier to set guidelines around how many hours someone is expected to hold the pager for based on this number.

For example, someone may have a base pay of $150K, along with an on-call schedule of no more than ten hours per week.

On the cons side, there’s a few things to think about.

Overrides are a part of on-call life. When someone has something come up, they’ll need to be able to hand the pager to someone else for coverage. With pay that’s part of a base salary, these overrides have a tendency of becoming very transactional. A favor for a favor, in other words.

Also, it’s reasonable for folks to feel like they aren’t actually being compensated at all when on-call pay is part of their base salary. Out of sight, out of mind.

And when they don’t “see” thay they’re being paid for this time, they may approach it a bit begrudgingly.

🇺🇸 A note about on-call compensation in the United States

In the United States, compensation for being on-call is a relatively rare occurrence, especially at bigger companies. Many businesses stipulate that being on-call is simply a part of the job duties and factor it into employees' salaries, thus not giving out any additional compensation.

However, some companies break the mold and do compensate their employees for being on-call, like Google. According to their SRE guide, “...Google offers time-off-in-lieu or straight cash compensation, capped at some proportion of overall salary. The compensation cap represents, in practice, a limit on the amount of on-call work that will be taken on by any individual.”

Ultimately, whether or not you compensate your team for being on-call in the United States is your decision. But as always, consider your situation and the stage you’re currently in.

Compensation per incident

With this approach, folks on-call are paid for every incident they respond to. This approach makes a ton of sense. If someone is being woken up several times a night, it’s fair for them to be paid more than someone who only responded to a handful of overnight incidents throughout the week. This can, in turn, help make a bad week full of incidents feel not so bad once a check comes in.

On-call, whether the pager is going off or not, is a significant burden. Life happens. And it’ll get to a point where folks are so overwhelmed by things going on at home that they’ll feel like being on-call should come with some compensation.

But there are a few instances where this approach can have some drawbacks.

The first, and it’s an edge case, is if folks know that they’re going to be compensated per incident, what’s the incentive to create more resilient systems? Or, put another way, why ensure that the fixes they implement are any good if they’re going to be compensated for them breaking often? Again, this is an extreme example, but the risk is there.

Another thing to consider: no incidents during the week means no additional pay. And even if folks aren’t responding to a single pager alert, the inconvenience of carrying their laptop around everywhere they go suddenly becomes much bigger.

Compensation by time spent on-call

Finally, the most straightforward compensation structure is paying folks a flat rate for the time they’re on call. For example, $300 to hold the pager for eight hours per week.

This compensation is irrespective of whether or not they actually responded to incidents. This structure is usually run as a weekly rotation, or rota, so everything feels balanced and everyone has a fair shake.

This structure is especially helpful with overrides where if someone takes the pager from someone, there’s no pressure to “pay them back” by taking back a shift.

But just like the other compensation structures we’ve covered here, there are some things to consider.

First, a flat rate sounds great until someone has had a particularly bad week and incidents are happening all night long, creating a lot of noise. This is where overrides come in handy yet again. If someone has been up all night responding to a flurry of incidents, offer them some time off to recover.

The number you land on, and the structure you use, will always be a challenging decision to come to. But as long as you’re approaching that decision sensibly, you can land on something everyone is OK with.

Another thing to consider is that certain teams may end up having more alerts than others. When everyone is getting paid a flat rate for being on-call, this can lead to a feeling of imbalance. While there’s no perfect workaround here, consider having a higher rate for teams who typically experience more incidents than others.

What we’d recommend

Context means everything and each situation is unique. The size of your business, your product, where you’re located, and so much more will dictate how you pay folks to be on-call.

At incident.io our approach is this: as long as you compensate people fairly, set clear expectations about responsibilities, and acknowledge the impact being on-call has on lives, you’ll be onto something.

Acknowledge the inconvenience of on-call and compensate accordingly

Holding a pager is a significant burden for your team. It can cause sleepless nights, rescheduled plans, stress and much more. Because of this, it’s imperative that you compensate folks in a way that factors in all of these realities.

The number you land on, and the structure you use, will always be a challenging decision to come to. But as long as you’re approaching that decision sensibly, you can land on something everyone is OK with. At incident.io, we pay a base of $350 per week and expect folks be paged a few times a week outside of working hours. We’ve written about our setup in more depth in On-call by default

Context matters

At larger organizations, a flat rate like the one we described above might not work. This is because different rotations will naturally have different burdens of responsibility:

Team A might get paged once a month, while Team B might be paged ten times per week. Paying both of these folks the same amount might not work in the long-term.

Here’s a few ways to make sure everything stays balanced and responsibilities feel equitable.


Tiered compensation by incident frequency and criticality

It’s important to acknowledge that pager load will be different between teams. To compensate for this, you can calculate pay based on a different set of expectations:

  • Teams who need to be available at very short notice, with a higher pager load, are paid a higher rate
  • Other teams with much lower alert noise, or where the response time isn’t as important, can be paid a lower rate

It boils down to this: if you’re on a rotation that’s responsible for a critical service and need your laptop with you at all times, it’s fair that you’re compensated higher than someone who can leave it at home without much worry.

Compensating folks who are pulled in to support

It can also be useful to compensate those who aren’t on-call but are pulled in to support with an incident. In this case, we’d recommend paying these people for each incident they respond to.

We’ve historically used a manual reporting process to make this work. In this situation, people who aren’t on-call, but joined an incident out-of-hours to support someone can simply fill out a Google (or equivalent) form outlining when and why they joined.

It's up to you!

Unless your local laws stipulate otherwise, what you decide to pay for on-call, and whether you do it at all, is up to you.

Again, there’s no such thing as a one-size-fits-all approach to on-call compensation. What you pay is heavily context-dependent and will vary based on your business, your incident load, the people who’ll be on-call, and a number of other factors.

But in general, as long as you’re keeping things simple, pay folks fairly, and acknowledge the burden out-of-hours work can have, you’ll be headed in the right direction.