For most of last year I used Otter.ai for meeting notes. It worked. The transcripts were accurate enough, the summaries were decent, and I’d set it up so long ago that I’d stopped thinking about it.

Then in January I sat in a client call where the other person said, mid-conversation: “Is that Otter recording this?” They weren’t angry. But there was a pause. A noticeable one.

I switched tools the following week. And switching forced me to actually evaluate what I needed from a meeting note tool — which turned out to be quite different from what I thought I needed.

What I Actually Needed (Versus What I Had)

When Otter joined a meeting, it appeared as a participant. There was a bot in the call with a name. On internal team calls, nobody cared. On client calls, some people cared a lot. Even when they didn’t say anything, the dynamic shifted slightly — there was now a visible record being made, and everyone in the call knew it.

What I actually needed was simpler: accurate notes from the conversation, organised enough that I could write a follow-up email or a task list immediately after the call ended. I didn’t need a bot. I didn’t need a participant list or sentiment analysis or a searchable meeting library. I needed clean notes, fast.

That gap between what the tool offered and what I actually used is what I’d been ignoring for over a year.

What I Switched To — and Why It Works Better for Me

I moved to Fathom. It records locally from your device — no bot joins the call, no participant appears in the meeting. The person on the other side of the call sees only you. Fathom runs quietly in the background, captures the audio, transcribes it, and produces a summary automatically once the call ends.

The summary quality is better than what I was getting from Otter. Not marginally — noticeably. Fathom’s summaries identify action items correctly, attribute them to the right speaker, and produce a follow-up email draft that I can usually send with minor edits rather than major rewrites.

The free tier covers unlimited calls, which is the thing that surprised me most. I’d been paying for Otter and the replacement was free.

The Broader Pattern This Pointed To

The Otter-to-Fathom switch made me think about how I evaluate tools generally. I’d been using Otter for so long that I’d stopped asking whether it was still the right tool. The evaluation had happened once, years ago, and then the tool became invisible infrastructure — something I ran without reconsidering.

Most tools end up in that category eventually. They become habits. And habits don’t get evaluated on a schedule.

What I do now — imperfectly but intentionally — is ask two questions about each recurring tool every quarter: what am I actually using this for, and is there something that does just that thing better? Not “does this tool have more features.” The question is whether it does the specific job I use it for better than the alternatives.

Otter had more features than Fathom. It also had a bot in my client calls that I never needed and eventually cost me a moment of discomfort I could have avoided.

The Tools I’m Evaluating Right Now

Applying the same logic to the rest of my stack:

Granola — another no-bot meeting tool that works well for Mac users who want the lightest possible footprint. I’ve tested it alongside Fathom and both are solid; Granola is slightly simpler, Fathom produces better structured summaries. Either is a better default than a bot-based transcription service for client-facing work.

tl;dv — if you need video recording alongside the transcript (for internal reviews, training recordings, or async team communication), tl;dv does this well and also has a no-bot option. The free tier includes unlimited recordings.

Fireflies — still bot-based, but the post-call organisation features are strong if you need a searchable archive of all your calls. For teams doing sales or customer success where call history matters, Fireflies makes more sense than it did for my use case.

The right tool depends on what you actually do with meeting notes after the call ends. If the answer is “write a follow-up email and a task list,” almost any transcription tool with a decent summary feature will work. If you need a searchable library or video review, the criteria change.

What the Switch Actually Changed

The practical difference: my client calls feel cleaner. There’s no moment of explaining what the bot is, no visible recording participant, no slight shift in the room when someone notices it. The notes I get are good enough to act on immediately.

The less obvious difference: switching made me realise that I’d been optimising for the wrong thing. I was paying for features I never used and ignoring a friction point I encountered every week. That’s a common pattern with tools you’ve used long enough to stop seeing clearly.

The meeting note tool I use now is free, less intrusive, and produces better summaries for the tasks I actually use them for. The only thing I lost was a bot with a name and a sentiment analysis dashboard I opened twice in 14 months.

Sometimes the upgrade is a downgrade in features and an upgrade in fit.

Related reading

• Three AI Tools That Turn Long Meetings Into 5-Minute Written Summaries
• How I Turn a Customer Sales Call Into a Case Study in Under an Hour Using Granola, Claude, and Notion
• Whisper vs Otter for Transcribing a 1-Hour Podcast Interview