How can AI in Slack help you run better 1:1s?

On Monday, you notice the PR has been sitting in review for four days. You think: I'll bring that up in our 1:1.

The 1:1 is Thursday at 2pm. By Thursday the PR has merged, a client escalation has eaten Wednesday, and the thing you noticed feels too small to raise and too old to be fair. So you talk about the sprint. You both leave feeling fine. Nothing changed, and the next PR sits for five days.

That's the most common way a 1:1 fails. Nobody did anything wrong. The moment just went cold. This guide covers what a 1:1 is for, why it drifts, where AI in Slack can actually help, and what to look out for before you let a tool anywhere near the conversation.

What a 1:1 is actually for

A 1:1 is the one meeting built for things that can't happen anywhere else. Status can go in the project tracker. A quick question can go in a thread. What belongs in a 1:1:

• The pattern you've noticed. Not the single late update, the third one.
• Context the other person doesn't have. Why the roadmap moved, what the VP actually said.
• What they're stuck on. Often something they haven't said in a channel because it feels like admitting a problem.
• Specific praise. "The way you handled the vendor thread when the deadline moved saved us a week" lands differently in person than as an emoji.
• What's hard. For them, and sometimes for you.

It also runs in both directions. The person on your team should walk in with something too. If you're an IC, your 1:1 with your manager is the best chance you get each week to say the thing that's been bugging you, and it deserves the same prep.

Why 1:1s go stale

Status crowds out everything

Status is safe. It has an obvious agenda and nobody's feelings get hurt. If you don't decide in advance what the real conversation is, status will fill all thirty minutes.

The real issue cools off

Something happens on Monday. The 1:1 is Thursday. In between, the details blur, you second-guess whether it was a big deal, and by the time you're in the room it feels petty. Most small issues should be named the same day, lightly, in one sentence. In the Accountability Dial™ that's a Mention, and it often doesn't need the 1:1 at all. The 1:1 is for the second or third time, when there's a pattern worth looking at together.

Prep happens from memory

Ten minutes before the meeting, you try to remember the week. What you remember is whatever happened yesterday. The quiet win from Tuesday, the standup where someone went silent, the thread that got tense: gone.

Agreements evaporate

Last week you agreed the release checklist would have owners by Friday. Nobody wrote it down. This week neither of you mentions it. That's how the same conversation happens three times.

What useful AI help looks like

A good AI in Slack doesn't run your 1:1. It helps with the parts around it that tend to slip.

Prep from real context

Instead of prepping from memory, the tool works from the shared channels you've invited it to. It can point out things like: the release checklist still has two items without owners, the design review got pushed twice, and someone picked up a messy customer thread without being asked. You still decide what matters. You just aren't deciding from the last 24 hours.

Noticing the conversation that's due

This is the part most tools skip. Useful help doesn't just collect topics for Thursday. It tells you when something shouldn't wait for Thursday, and when something you've mentioned twice now needs a real conversation. It also helps you judge how serious a situation really is, so a first-time slip doesn't get treated like a crisis and a fourth-time slip doesn't get treated like a passing comment.

Drafting the opener

The first sentence is where most hard 1:1 topics die. You know what you want to say. You don't know how to start without it sounding like a verdict. A good tool helps you find a first line that sounds like you:

"Before we get into the sprint, I want to ask about the review queue. The last couple of PRs sat for a few days. What's going on there?"

Or, for praise, which gets skipped even more often than criticism:

"I wanted to say this out loud: the way you handled the vendor thread on Tuesday saved us a week. That's exactly what I hope you'll keep doing."

It should also ask why the conversation is hard for you. Sometimes the honest answer ("I hired them and I don't want to be wrong") is the thing that's been holding you back.

Recapping what was agreed

After the meeting, a two-line summary: what you agreed, and when you'll check. That's it.

"Agreed: PRs get a first review within one working day, and you'll flag in #eng-review if you're blocked. Let's check how it's going on the 15th."

Next week, the tool can remind you to look at it before you sit down.

What to watch out for

Some AI help makes 1:1s worse. Three patterns to look for.

AI that reads DMs

Direct messages are where people vent, ask the dumb question, and say the thing they'd never put in a channel. A tool that reads them, even "just for context," changes how everyone writes. Once people suspect it, the honest messages move to text or stop entirely. If a tool can read DMs between people, that's a dealbreaker on its own.

AI that agrees with everything

General-purpose chatbots are built to be helpful, which often means agreeable. Ask one whether you should raise the late PRs and it will say yes, and also validate whatever framing you brought. That feels good and doesn't help. The useful response is sometimes "you've mentioned this three times and haven't said it once. What's stopping you?" A tool that never pushes back is a mirror, and you already have one.

Tools that report on people

Some tools turn Slack activity into dashboards: response times by person, sentiment by person, lists of who seems disengaged. The moment a 1:1 tool is also a reporting tool, the person across from you is being measured, and they'll act like it. Help for a conversation and a report about a person cannot be the same product.

A checklist for evaluating any tool

Use this with any AI tool you're considering for 1:1s, in Slack or anywhere else.

Question	What good looks like	What should worry you
What can it read?	Only shared channels you've invited it to	Direct messages, or "everything in the workspace"
Who sees what you type to it?	Only you	Your manager, HR, or "admins"
What does leadership see?	Aggregate themes across a team, never traceable to a person	Per-person dashboards, rankings, or ratings
Does it push back?	Asks hard questions, names what you're avoiding	Agrees with whatever you bring
Does it help you speak?	Drafts a first sentence in your voice	Writes generic scripts you'd never say
Does it notice timing?	Flags what shouldn't wait for the 1:1	Only builds agendas
Does it close the loop?	Short recap of what was agreed and when	Long transcripts nobody reads
How is your data handled?	SOC 2 Type II, clear hosting, no retention by the model provider	Vague answers or "we'll get back to you"

For a deeper version of the privacy questions, see how to evaluate whether an AI coaching tool is privacy-safe.

Where Ren fits

Ren is one example of this kind of tool. It's an AI app that lives in Slack and Microsoft Teams (plus a web app) and helps with the conversations between people at work. It notices conversations that aren't happening, helps you judge how serious a situation is using the Accountability Dial, drafts the first sentence in your voice, and asks why it's hard for you. It cannot read direct messages between people. What you say to Ren stays private, never shared with your manager or HR, and any team-wide view is aggregate and never traceable to a person.

On pushing back: in an independent blind study (The Pushback Report), Ren scored 4.22 out of 5 on pushback and ownership, against 2.84 for ChatGPT, 2.68 for Copilot, and 2.48 for Gemini. It's built for every seat, so an IC can use it to prep for their 1:1 with their manager just as easily as the other way around. You can try it at tryren.com.

---

Related: How to prepare for a 1:1 meeting · Preparing for a tough 1:1 with an underperformer · What is the Accountability Dial? · Ren in Slack