How to Do a Mid-Year OKR Review: 5 Moves That Keep the System Alive

It's the first week of July. Someone opens the OKR doc from January for the first time in months, and there it is: twelve key results, four of them hit, three of them dead, and five that nobody has touched since March. The room splits into two instincts. Half want to close the doc and quietly move on, because looking too hard at the dead ones feels like admitting the year went sideways. The other half want to scrap the whole thing and re-plan from scratch, because surely a clean slate beats this graveyard.

Both instincts are wrong, and both come from the same place: a failure of nerve. The mid-year review isn't a confession and it isn't a do-over. It's a re-score and a re-baseline. You grade what happened honestly, you keep what's still load-bearing, you retire what the business outgrew, and you reset the survivors for the second half with the reality you actually have now. Here are the five moves that make it work.

A mid-year OKR review, in one paragraph

A mid-year OKR review is a structured re-score and re-baseline, not a reset. You grade each H1 key result honestly — hit, at-risk, or dead — keep the ones still worth chasing, retire the ones the business has outgrown, and re-baseline the survivors for H2 against current reality rather than January's assumptions. Done well, it takes about an hour per team and protects the system. Done as a full reset, it teaches everyone that OKRs don't survive contact with the year.

Move 1 — Re-score before you re-plan

The biggest mistake teams make at mid-year is jumping straight to "what should H2 look like?" before anyone has graded H1. It feels productive, but it's avoidance — re-planning lets you skip the uncomfortable part where you look at the goal you missed and say so.

So before a single new objective gets written, grade every existing key result against its original target. Three buckets: hit, at-risk, dead. No new goals enter the conversation until the old ones have a grade next to them. This is not busywork — the grades are the raw material for everything that follows. You can't re-baseline a number you haven't honestly assessed.

I keep this to twenty minutes. Pull up the doc, go line by line, assign a grade, move on. The discipline is in refusing to debate H2 priorities until the scoring is done. The moment someone says "well, that one doesn't matter anymore because we're pivoting to —" stop them. Grade it first. The pivot is move three.

Move 2 — Kill the dead ones out loud

There is a specific way OKR systems die, and it isn't dramatic. A key result quietly stops being mentioned. It's still in the doc, technically, but everyone has silently agreed to look past it. Do that two or three quarters running and you've taught the whole team that goals are decorations — there when convenient, ignored when not.

The fix is to retire dead key results explicitly, in the room, on the record. "We set this in January. The market moved. We're not chasing it in H2, and here's why." Naming it clears the doc of clutter that drags down every future review, and it sends the opposite signal from silent abandonment: goals here are real enough that we formally decide when they stop.

This is also where honest scoring pays off. A dead key result that gets named and explained is more useful than a 0.4 that everyone privately knows is fiction. The score isn't a verdict on the team. It's data. Soften it to protect feelings and you've corrupted the input your H2 plan depends on.

The harder problem is catching that silent abandonment before July, not after. This is what we built Rhythms' Radar to do — a key result that hasn't moved in six weeks gets surfaced on its own, so the drift shows up as a flag in real time instead of as a graveyard you discover at mid-year.

Move 3 — Re-baseline the survivors for H2

Now you re-plan — but lighter than you think. For most of the goals that survived the cull, the answer isn't a brand-new objective. It's the same objective with an updated target, reset against where you actually are in July instead of where you guessed you'd be back in January.

Here's the difference in practice. Re-planning says: scrap "grow qualified pipeline to 10,000 leads," invent something new, socialize it across four teams, and spend two weeks getting everyone re-aligned. Re-baselining says: you're at 5,200, the original 10,000 assumed a hire you didn't make and a channel that underperformed, so the load-bearing version for H2 is 8,500 — same goal, honest number. One of these takes an afternoon. The other eats most of a quarter.

This is exactly where Rhythms' Goals & Alignment earns its place. When a target changes mid-year, the cascade underneath it updates on its own — the team goals and the individual key results that rolled up to the old number re-point to the new one without a re-communication tour. The difference between a one-hour re-baseline and a two-week one is almost entirely whether that propagation happens automatically or whether someone has to chase it down through thirty Slack messages.

Move 4 — Cut the count

Mid-year is the best moment all year to admit you set too many goals. Most teams do. Twelve key results in January almost always means three or four genuinely mattered and the rest were hedging.

The re-score makes the cut easy, because the dead and untouched buckets are doing the work for you. If a key result has been ignored since March and nobody fought to keep it in move two, that's your answer. Coverage was never the goal. Focus is.

The test I use: can each person say their team's goals out loud, without opening the doc? If the answer is no, the list is too long, and H2 is your chance to fix it. Walking out of a mid-year review with fewer goals than you walked in with is usually a sign it worked.

Move 5 — Put it on a rhythm so H2 doesn't drift

Here's the part nobody wants to hear: the reason your H1 goals drifted isn't that you wrote them badly in January. It's that you didn't look at them again until July. The mid-year review exists because the system has no heartbeat between planning and panic.

The Scale-Up Institute found in 2026 that 65% of high-growth companies that abandoned OKRs did so because goals "got lost in the daily grind" — and 92% of those failures traced back to having no ongoing rhythm to keep goals in front of people. The failure is rarely the goal-writing. It's the silence afterward.

So before you leave the mid-year review, schedule the next beat. Not "we'll check in around Q3 close" — an actual recurring review where the surviving goals get looked at against live data, the way you'd run any operating review. This is the whole idea behind the Executive Operating Review, and it's what keeps OKRs from becoming shelfware: the goals don't sit in a doc waiting to be rediscovered, they show up in a review you're already running. Put the rhythm back, and next July looks a lot less like a graveyard.

What to do before your next planning session

If you do one thing this week, don't re-plan. Re-score. Open the January doc, grade every key result honestly, and say the dead ones out loud. That single hour will tell you more about your H2 than a day of fresh goal-setting ever could, because it forces you to start from what's real.

The teams that keep OKRs alive aren't more disciplined than everyone else. They've just stopped treating the mid-year review as a verdict and started treating it as maintenance. Goals drift. That's physics, not failure. The only question is whether you have a rhythm that catches the drift in July, or one that lets you find out in December.

Try it free at rhythms.ai.

Frequently Asked Questions

Should you change OKRs in the middle of the year?

Yes, selectively. Retire the key results the business has genuinely outgrown and re-baseline the ones still worth chasing against current reality. What you shouldn't do is swap goals wholesale every time they get hard — that trains the team to treat OKRs as disposable, and the system stops meaning anything by Q4.

How do you re-score an OKR that's already off track?

Grade it honestly against the original target — hit, at-risk, or dead — and resist the urge to soften the score to protect anyone's feelings. The score isn't a judgment on the team; it's the data your re-baseline depends on. A dead key result that gets named is far more useful than a 0.4 nobody actually believes.

What is an operating rhythm and why does it matter for OKRs?

An operating rhythm is the recurring set of reviews that keep goals connected to weekly work. Research consistently finds OKR failure traces less to bad goal-writing than to the absence of this rhythm. The mid-year review is one beat in it — not a one-off event you brace for once and forget.

How long should a mid-year OKR review take?

For a single team, about an hour: roughly 20 minutes to re-score and 40 to re-baseline H2. If it's stretching into a full day, the time is going to data-gathering and deck-building rather than decisions — a sign the prep should be automated, not that the meeting needs to be longer.