A client emails: “Quick one – when will this be ready?” You glance at your board, feel a little jolt of pressure, and type back the first date that feels brave but not reckless. “Friday, probably.” You hit send, and immediately start hoping. If that little hope-and-pray ritual sounds familiar, you’re not bad at planning – you’re just answering a data question with a gut feeling.
There’s a calmer way, and it doesn’t require a project manager or a spreadsheet full of estimates. It’s called a service level expectation, and it’s simply a promise your board can actually keep – something like “most tasks like this are done within nine days.” Here’s how to build one from work you’ve already finished, and how to use it so “when will it be done?” stops being a question you dread.
What a Service Level Expectation Actually Is
A service level expectation – SLE for short – is a plain-language statement of how long your work usually takes, paired with how confident you are. The honest version has two parts: an amount of time and a probability. “85% of our cards finish within nine days” is a complete SLE. “About a week” is just a vibe.
That second number is the part most people skip, and it’s the part that makes the promise trustworthy. Saying “nine days” alone pretends every task behaves identically. Saying “nine days, 85% of the time” admits the truth – that some work surprises you – while still giving the other person something solid to plan around. It’s the difference between a forecast and a fortune.
Why “It’ll Be Done Friday” Keeps Burning You
The trouble with a gut-feel date is that it’s built from optimism, not evidence. You picture the work going smoothly, you forget the last three times something got snagged in review, and you quote the best-case day as if it were the likely one. Then real life shows up – a sick day, a dependency, a round of revisions – and the date you promised turns into the date you apologize for.
An SLE fixes this by refusing to guess. Instead of imagining how long a task should take, it looks at how long your tasks actually take, again and again, and lets that pattern do the talking. Your board has been quietly recording this evidence the whole time. You just have to read it. The raw ingredient is cycle time – the real number of days each card spends in motion – and once you have a pile of those, the SLE almost writes itself.
How to Build Your SLE (No Math Degree Needed)
You don’t need statistics. You need a list. Look back at the last 20 to 30 cards your team finished and, for each one, count how many days it took from “we started this” to “this is done.” Write those numbers down in a row. That’s your entire dataset.
Now sort them from fastest to slowest and find the point where most of them land before. If you have 20 finished cards and you want to be right about 85% of the time, look at the 17th number in the sorted list (that’s 85% of 20). Say it’s nine days. Congratulations – your SLE is “85% of cards are done within nine days.” The handful of cards slower than that are exactly the risk you’re openly accounting for with that “85%.”
Pick your confidence level to match the stakes. For low-risk internal work, 70% might be plenty. For a promise to a paying client, reach for 85% or even 95% – which will give you a longer, safer number. The beauty is that it’s all drawn from reality, so it stays honest even when you’d rather it didn’t.
Turn the Number Into a Promise You Can Keep
Once you have an SLE, answering “when will it be done?” stops being an act of courage. A new request comes in that looks like your normal work, and you can reply: “Based on how we’ve been running, this should be finished within about nine days – we hit that around 85% of the time, and I’ll flag you early if it looks like one of the exceptions.” That sentence does something a bare date never can: it sets a clear expectation and quietly promises honesty if things slip.
This is where an SLE earns its keep with clients. A shared, realistic timeframe is the backbone of keeping people in the loop without endless status meetings – they stop asking for updates because they already know roughly when to expect one. You’ve replaced anxious check-ins with a number everyone agreed to up front.
Put the Expectation on the Board, Not in Your Head
An SLE that lives only in your memory will quietly rot. Write it down where the work happens – pinned to the top of the relevant column, or noted in your board’s policies – so it actually shapes decisions. “Standard tasks: 9 days at 85%” sitting in plain view turns the expectation from a thing you remember into a rule the whole team can see and defend.
Made visible, the number starts pulling its weight on its own. When someone’s about to pile a fourth urgent request onto an already-full board, the SLE is right there reminding everyone what that overload will do to delivery times. It pairs naturally with sensible limits on work in progress, because the fastest way to blow your SLE is to start everything at once and finish nothing.
When a Card Blows Past the Expectation, Act Early
The real magic of an SLE isn’t the promise you make at the start – it’s the early-warning system it gives you in the middle. Once you know most cards finish within nine days, any card still sitting in progress on day eight is waving a small flag. It hasn’t broken your promise yet, but it’s about to, and now you know while you can still do something about it.
This is the same discipline behind watching your aging cards – the ones quietly getting older than they should. A card approaching its SLE is your cue to step in: unblock it, get it help, or pick up the phone and give the client the heads-up a day early instead of an apology a day late. Given warning, most people adjust gracefully. Ambushed on the deadline, the same slip feels like a broken commitment. And when you want to go beyond a single card and forecast a whole batch of work, the very same cycle-time data lets you forecast a delivery date without story points or guesswork.
The Takeaway
A service level expectation is the quiet confidence of knowing your own numbers. Count how long your recent cards really took, pick the point most of them land before, attach a confidence level honest enough to trust, and write it on the board. Then use it twice: once to make a promise you can keep, and again to catch the cards that are about to break it.
Do that, and “when will it be done?” turns from a stressful guess into a two-second answer backed by evidence. Keep your work flowing in priority order on a simple kanban board, let it quietly record how long things take, and you’ll always have an honest number ready the next time someone asks.