MeetDone
Meeting transcripts → follow-up emails in 30 seconds
Hey Indie Hackers,
Day 8 of building MeetDone.
Not a feature update today.
One thing became clearer this week:
the hard part of follow-up emails is not writing.
It's deciding what actually matters from messy meeting notes.
I used to think the prompt was the main problem.
It isn't.
Structure matters more than prompt.
Most notes are a mix of context, side comments, and half-decisions.
The useful part is usually just 4 things:
- what was decided
- what needs to happen
- who owns it
- when it should happen
So I started using the same simple format every time:
1. One-line recap
"We aligned on X and the next step is Y."
2. Decisions
- Decision 1
- Decision 2
3. Action items
- [Owner] -> [task] by [date]
- [Owner] -> [task] by [date]
4. Next checkpoint
"We'll review progress on [date]."
A few things this fixed immediately:
- less blank-page staring
- less tone tweaking
- less rereading notes
- much easier for the other side to skim
The biggest mistake I used to make was turning follow-ups into mini essays.
Most people do not need an essay.
They need clarity.
And one rule now feels obvious:
if there is no owner and no date, it's not a next step.
It's just a hope.
You don't need my product to test this.
Try the structure manually after your next 5 meetings and see if it cuts your follow-up time.
If you send a lot of follow-ups, what slows you down most right now?
- extracting the decisions
- writing the email
- making action items clear
P.S. I'm building MeetDone around exactly this format: notes/transcript in -> follow-up email out in about 30 seconds.
Quick reality check from building MeetDone.
Day 5:
5 sessions, 0 signups.
That told me almost nothing.
Day 7:
19 sessions, 0 signups.
That told me something.
At very low traffic, zero conversions can just be noise.
Once traffic starts moving and conversions stay at zero, the diagnosis changes.
It stops being:
"I need more traffic."
It becomes:
"Something on the page is making people hesitate."
I had the landing page audited and found 5 trust leaks:
1. The free plan was not described consistently.
2. One claim sounded stronger than I could prove.
3. The privacy wording did not feel fully aligned.
4. Scarcity was there, but not believable enough.
5. There was no clear before/after proof above the fold.
None of these issues is huge on its own.
Together, they quietly tell a visitor:
"Be careful."
That is enough to kill conversion.
My takeaway:
Traffic is only useful when the page deserves it.
So if you are getting visits but no signups, I would check these 5 things before spending more on acquisition:
1. Is the offer consistent everywhere?
2. Are you making any claim that sounds too polished to trust?
3. Is your privacy story simple and believable?
4. Is your urgency real, or just decoration?
5. Can someone see proof in 5 seconds?
That is what I am fixing next for MeetDone.
Not new features.
Not more channels.
Just trust.
Because 10x more traffic to a page people do not trust is still 0 signups.
Curious how other founders think about this:
what is the smallest trust fix that improved conversion for you?
I think a lot of meeting recaps fail for one boring reason:
they hide the unresolved part inside the summary.
The email looks clear.
The client says "sounds good."
But the real blocker is still sitting in paragraph 3, mixed in with everything else.
That creates fake clarity.
I see this a lot in client work.
The recap covers:
- what was discussed
- what sounds decided
- a few next steps
But it does not clearly say:
what still needs an answer before the work can move.
That is why I think open questions need their own line.
Not because uncertainty is bad.
Because hidden uncertainty is expensive.
If the unresolved part is explicit:
- everyone sees the real blocker
- someone can own resolving it
- the next step becomes honest instead of vague
The follow-up structure I trust most right now is:
- decisions
- action items
- open questions
- next step
For freelancers and consultants, I think this matters more than writing a polished recap.
A short follow-up that makes the blocker obvious is usually more useful than a longer one that sounds complete.
If you do client work, what gets buried more often after meetings: the blocker, the owner, or the next step?
One thing I keep noticing in bad follow-ups is this:
they recap the meeting well, but nobody can tell who owns what next.
That kind of email feels organized.
But it does not move anything.
I think a useful action item needs at least 3 things:
- owner
- task
- date
If one of those is missing, execution gets fuzzy.
- no owner: everyone assumes someone else has it
- no task: people agree in principle and do nothing
- no date: it quietly becomes "sometime this week"
That has become a core rule for me while building MeetDone.
I care less about making the recap sound polished and more about making the next action impossible to miss.
So my default checklist after a call is simple:
- what we decided
- who owns each action item
- what date or trigger moves it forward
A shorter follow-up with clear ownership is usually more useful than a longer recap with vague language.
If you do client work, what gets missed more often in your follow-ups: the owner, the deadline, or the next step?
The moment after a client call is not when people want a new workflow.
It is when they want the shortest path to done.
That is the product lesson I keep coming back to while building MeetDone.
After a meeting, most people are already switching to the next thing. They do not want to:
- invite a bot
- clean up their notes
- rename everything
- learn a new ritual before they send the follow-up
They want to get the email out while the conversation is still fresh.
So I have a simple rule now:
build for the tired version of the user.
If the product can work with the messy notes or transcript they already have, it has a chance.
If it needs a brand-new workflow, usage drops fast.
That is why I keep MeetDone simple:
- paste notes or the transcript you already have
- generate a clear follow-up
- send it before context disappears
The lesson feels bigger than this product.
A lot of software loses people not because the core idea is bad, but because the workflow asks for too much at the wrong moment.
The best tool is often the one that asks for the fewest extra steps.
What is one step you removed from your product that made people actually use it more?
Yesterday I wrote about structure.
Today I think timing matters just as much.
A lot of follow-ups do not stay unsent because they are hard.
They stay unsent because they become a "later today" task.
And once that happens, three things usually go wrong:
- the exact decision gets fuzzy
- action items become softer
- the energy from the meeting disappears
So I have a simple rule now:
if the call just ended, send the useful version first.
Not the perfect version tonight.
Not the polished version tomorrow.
The useful version now.
For me, that usually means just 3 things:
- what we agreed
- who owns what
- what happens next
If I want to improve tone later, fine.
But the job is already done.
A short email sent while everyone's context is warm is usually better than a beautiful email sent after the moment is gone.
This is also how I think about MeetDone.
I am not trying to help people write longer follow-ups.
I want to help them ship clarity faster.
If you do client work, what gets you more often after a meeting: overthinking the email, or telling yourself you will send it later?
Yesterday I wrote about the real problem happening after the meeting.
Today I want to make it practical.
I think most follow-up emails are too long for one reason: we treat them like writing tasks instead of clarity tasks.
After a call, people try to sound polished.
They try to include every detail.
They rewrite sentences nobody asked for.
But most clients usually need 3 things:
- what we decided
- who does what
- what happens next
That is it.
If those 3 things are clear, the email did its job.
The mistake is thinking a follow-up needs to sound impressive.
It does not.
It needs to remove confusion.
The fastest follow-ups I have seen are usually short.
Not lazy. Just clear.
This is the structure I keep coming back to while building MeetDone:
- summary: 2-3 lines on what was agreed
- action items: owner + task
- next step: deadline, next meeting, or waiting point
I am not trying to replace thinking.
I am trying to remove the blank-page friction after a call.
If you do client work, what do you usually overdo in follow-ups: too much context, too much polish, or too many open loops?
Restarting daily posting from day 1.
Not because nothing happened. Because I want to talk about the real problem, not just my metrics.
A client meeting can go well and still create zero momentum. The meeting ends, the next task starts, and the follow-up email becomes one more job nobody wants to do.
You stare at a blank message. You try to remember what was agreed. You rewrite the same polite opening line. You tell yourself you will send it later. Later becomes tonight. Or tomorrow.
That delay looks small, but it costs a lot:
- It makes you look slower than you are.
- It blurs action items.
- It kills the energy of a good conversation.
That is the reason I built MeetDone. Not to put another bot into meetings. Not to create more AI noise. Just to help turn notes or a transcript into a clear client follow-up in about 30 seconds.
What I care about is simple:
- recap what matters
- list who does what
- keep the momentum while the call is still fresh
If I can help freelancers and consultants do that faster, the product is worth building.
So this is Day 1 again.
If you do client calls, what part of follow-ups annoys you most: writing the recap, capturing action items, finding the right tone, or sending it fast enough?
Today’s snapshot:
5 total users
3 Pro users
21 emails generated all-time
0 signups in the last 7 days
0 generated emails in the last 7 days
1 active user in the last 30 days
What this really means:
We have a willingness-to-pay signal, but repeat usage is weak.
The bottleneck is activation, not feature count.
Shipping more is not the same as helping users reach first value faster.
Practical playbook I’m using this week (copy this if your usage is flat):
Track one metric daily: % of new users who generate their first usable follow-up email in session one.
Instrument only the critical path: visit -> signup -> first generate click -> first usable output.
Run founder-led onboarding live and log exact hesitation points.
Ship only fixes that reduce time-to-first-usable output.
What I’m doing next:
10 founder-led onboarding sessions
24-hour turnaround on activation fixes
Direct reactivation outreach to inactive users with one concrete use case
If your dashboard is quiet, narrow scope hard and optimize for first value first.
Today’s snapshot:
5 users
3 Pro
21 emails generated all-time
0 signups in the last 7 days
0 generated emails in the last 7 days
1 active user in the last 30 days
The data is quiet, but the signal is clear.
What I learned today:
Revenue signal can exist while activation is still weak.
Shipping volume is not the same as usage growth.
The bottleneck is still first-session value, not feature breadth.
What I’m doing next:
Personally onboard users and watch first sessions live.
Track exact hesitation/drop points.
Ship only fixes that improve first-session output.
Measure one daily metric: first usable follow-up email generated.
Working rule this week:
Dashboard shows where users drop.
Conversations reveal what they were trying to do.
Product changes should connect those two directly.
Comment