We’ve all been in this position as a manager.
You’ve noticed something’s off with one of your team members for a few weeks now.
Maybe it’s the quality of their work slipping slightly. Maybe they’re missing small details they’d normally catch. Maybe their communication has become less proactive.
But you’re not certain that there’s a problem yet. These are all relatively minor things and performance hasn’t actually dropped massively.
So you wait. You gather more evidence. You give them time to course-correct on their own.
And then, three months later, you’re sitting in a performance conversation that feels way harder than it should because the problem is now undeniable – and they’re blindsided by your feedback that their performance has been off for a little while now.
Here’s what I’ve learned after making this mistake multiple times: By the time you’re 100% certain someone has a performance problem, you’ve waited too long.
Your team member has been struggling (or underperforming) for longer than they should have, and you’ve let patterns become entrenched that are now much harder to address.
Today, I want to talk about the 70% rule which basically means having the performance conversation when you’re only 70% sure there’s a problem. And why that earlier conversation is actually kinder, more effective, and prevents the vast majority of performance issues from becoming much more serious than they need to be.
The certainty trap – why we end up in this position in the first place
We want to be fair. Nobody wants to be the manager who jumps to conclusions or has a difficult conversation based on incomplete information. We don’t want to lose trust or credibility with our team by jumping to conclusions.
So we wait. We want concrete examples, clear patterns, and undeniable evidence.
We also fear getting it wrong. What if we’re misreading the situation? What if they’re having a bad week, not a performance problem? What if we damage the relationship by bringing something up prematurely? What if it’s actually you that has the problem?
And if we’re really honest, we hope it’ll self-correct. If we just give it a bit more time, maybe they’ll sort it out themselves. Maybe they’re already aware and working on it. Maybe next week will be better.
I get it. I’ve been there more times than I can count. No one wants a difficult conversation.
I once managed a senior consultant who started delivering client work that was… fine. Not bad, but not at the level I’d come to expect from them over the years.
The first time it happened, I told myself everyone has off weeks. The second time, I thought maybe the brief that I’d set wasn’t clear enough. By the third time, I had a pattern – but by then, the client had noticed too, and what should have been a supportive coaching conversation became a “we need to talk about your performance” conversation.
They were (understandably) frustrated that I hadn’t said something sooner. I was frustrated with myself for the same reason.
The thing is, waiting for certainty doesn’t make the conversation easier. It makes it harder.
The true cost of waiting
When you wait too long to address a performance concern, several things happen – none of them good.
Patterns become habits
When someone’s been doing something suboptimally for months, it’s no longer a one-off behaviour – it’s how they work.
Changing established patterns is significantly harder than course-correcting early behaviours. What could have been a quick adjustment becomes a fundamental shift in how they approach their work.
The blindside effect
From your perspective, you’ve been watching this develop for weeks or months. You’ve been gathering examples, looking for patterns, building your case and becoming more confident in your observations.
From their perspective, this is the first time they’re hearing about it.
The gap between your timelines creates confusion and frustration on both sides. They feel ambushed. You feel like you’ve been patient. Neither of you is wrong, but you’re having completely different conversations.
You lose credibility
Your team member (rightfully) wonders: “If this has been a problem for three months, why am I only hearing about it now?”
This is even more true for senior team members who you’ve worked with for a long time and have come to expect better from you as well.
It damages trust and makes them question whether they can rely on you for timely feedback. If you didn’t tell them about this, what else aren’t you telling them?
The problem spreads
Performance issues don’t exist in isolation.
If someone’s work quality is dropping, their colleagues are often compensating for them. If their communication is becoming unclear, team efficiency suffers. If they’re missing deadlines, other people’s work gets delayed.
The longer you wait, the more people are affected. The more they notice you not acting.
I worked with a manager once who waited six months to address a team member who was consistently late to meetings and missing deadlines by a day or two.
“I wanted to be sure it was a pattern, not a phase,” they told me.
By the time they had the conversation, the rest of the team had noticed, resented it, and started questioning whether they played favourites. What should have been a straightforward conversation about punctuality became a conversation about damaged team trust and fairness.
Having the conversation at 70% certainty
So what does 70% certainty actually mean?
It means you’ve noticed something that concerns you. You have at least 2-3 specific examples. You can’t explain it away as a pure one-off, but you don’t yet have months of data.
It’s enough to raise the topic, but not enough to make definitive statements.
And actually, this timing is better than waiting for 100% certainty. Here’s why.
The conversation is still exploratory
At 70%, you’re genuinely curious about what’s going on. The tone is “I’ve noticed X, and I want to understand what’s happening” rather than “You have a performance problem.”
This makes the conversation fundamentally different. You’re not delivering a verdict. You’re investigating together. It feels very different for both of you.
It creates space for joint problem-solving
Because you’re raising it early, there’s room for the team member to provide context, identify obstacles, or flag things you don’t know about. You can get more information without them feeling like they need to defend themselves.
Maybe there’s a reasonable explanation. Maybe they’re already aware and working on it. Maybe they need support you didn’t realise they needed.
This is all a chance to find out.
It prevents the problem from defining them
When you address something early, it’s a behaviour or a pattern, not an identity.
“I’ve noticed your last few reports needed more revisions than usual” is very different from “Your work quality has become a problem.”
One is fixable. The other feels personal and entrenched.
When I started using the 70% rule, one of the first conversations I had was with a designer whose recent client presentations felt less polished than usual. I had three examples over about three weeks.
Instead of waiting for more data, I simply said: “I’ve noticed your last few presentations haven’t had the usual level of detail in the rationale slides. Is everything okay? Is there something getting in the way?”
Turns out they were dealing with a family health issue that was fragmenting their focus. We worked out a short-term plan to adjust their workload, and within two weeks, they was back to their usual standard.
If I’d waited for “certainty,” I’d have been having a performance conversation about someone who just needed temporary support.
What to actually say
Right, let’s get into the structure of the conversation itself and make this as actionable as possible for you.
The 70% conversation follows a simple framework:
1. Name what you’ve observed (specifically)
“I’ve noticed [specific behaviour/pattern] in [specific situations]. For example, [concrete example].”
The key here is specificity. Not “your work has been off lately” but “the last three client deliverables you’ve submitted have needed more rounds of revision than usual.”
You can’t leave room for them to fill the gaps themselves which will inevitably lead to them catastrophising things.
2. Express genuine curiosity
This may take a bit of practice in advance or some planning if this doesn’t come naturally to you. But we’re talking about using phrases such as:
“I wanted to check in because this isn’t typical for you” or “I’m curious about what’s going on” or “Help me understand what’s happening.”
This is critical. You’re not making an accusation. You’re genuinely trying to understand.
3. Create space for their perspective
This isn’t a one-way conversation. You want them to share their perspective and fill in the gaps so that you can understand more about what’s going on.
Try to use phrases such as:
“Is there something I’m missing?” or “What’s your experience been?” or “Are there obstacles I’m not aware of?”
Then stop talking and actually listen.
4. Make the purpose collaborative
In the same way that this isn’t a one-way conversation in terms of what is said, it’s also a two-way conversation in terms of how you move forward and improve things.
You should resist the urge to say things like “You should fix this by doing XYZ.”
Instead, say something like:
“I wanted to raise this early so we can figure out together whether there’s something we need to address or support you with.”
5. Be clear it’s not a formal performance issue (yet)
This is key and another big reason why catching issues early is so important. But don’t let the conversation sound more formal than it actually is by accident.
“This isn’t a formal performance conversation – I’m raising it now specifically because I don’t want it to become one.”
Here’s roughly what this looks like in practice:
“I’ve noticed the last three client deliverables you’ve submitted have needed more rounds of revision than usual – specifically, the strategy doc for our SaaS client, the audit for the tech startup, and last week’s presentation. I wanted to check in because this isn’t typical for you. What’s been going on?”
Then you listen. Really listen.
Don’t jump in with solutions. Don’t explain what they should be doing differently. Just listen to what they tell you.
Their answer will tell you what kind of problem you’re actually dealing with.
You’ll be surprised how often that early conversation prevents a later, much more difficult one – and how much your team member appreciates you raising it while it’s still small.
And if you’re worried about getting it wrong? Remember: Having a curious, supportive conversation that turns out to be unnecessary is a much smaller problem than staying silent about something that turns into a crisis.
Your team would rather you care enough to check in than wait until things are undeniable.
