Whose decision the date is
At most companies the launch date is a by-product of a sprint finishing. Engineering says the feature lands on the 14th, marketing is told the 14th, and everything downstream is scheduling rather than choosing.
That would be fine if the date were neutral. It is not. The same asset, the same claim and the same budget produce materially different outcomes depending on which day it goes out, what hour it goes live, and what else is happening that week.
Those are distribution facts, and the person who owns distribution should hold the pen. Why that authority is the thing that decides the rest.
The practical version is small: the date is proposed by product and confirmed by whoever owns the launch, with the right to move it a week. Most of the value in this page comes from having that conversation at all.
Which day
For a technical audience, Tuesday or Wednesday. Measured on 1,978 launches, accounts under 25,000 followers did best on those two days, at 98 and 110 times their following, and Saturday did about a third as well as any weekday. The numbers by day.
Tuesday is the most crowded day, with a quarter of all launches, and it still leads for small accounts. Friday had the most launches passing a million views, and the weakest weekday multiple for the smallest accounts.
The gap between a good weekday and a bad one is real and it is smaller than people hope. It will not rescue a launch with nothing to say, and it is free, which makes it worth taking.
Which hour, and why it matters more
The hour matters more than the day, because of how ranking works.
Reach is decided largely by engagement velocity in the first stretch after posting, measured against your own recent baseline. A strong start earns a wider test and a slow start is capped early and rarely recovers.
So the question is when the largest number of people who would actually reply are awake and free. The mechanism.
For a global technical audience that usually means the US morning, roughly 10am to noon Eastern: Europe is still at their desks and the American working day sits ahead of the post. Across 1,978 launches, small accounts posting on a weekday in that window reached a median of 77 times their following, against 54 times at other hours. The full numbers.
Two things follow from that which teams routinely get wrong. Do not schedule the post and go into a meeting, because the replies in the first hour are free and they are the most heavily weighted signal available.
And do not stack the exec amplification at 4pm, since the window it was meant to influence closed hours earlier.
What to check the calendar against
Five things, and the first two are the ones that actually cost you.
- Major industry events you are not part of. A conference keynote week soaks up the entire attention budget of your category. Being adjacent to it without being in it is the worst position available.
- A large competitor's known launch. Not because you will look worse, but because the same creators and the same feeds are occupied. If you know their date, move.
- Public holidays in your buyer's country, which is often not your own. A US-heavy audience is gone for Thanksgiving week and most of late December.
- Your own other announcements. Two of your things in the same fortnight split your audience against itself and halve both.
- Anything genuinely enormous in the news. Unforecastable, and worth having a documented decision about who can postpone and how late.
Investors introducing several portfolio companies in one category hit the second item constantly without noticing. If that is you, sequencing solves it and discovering it in week two does not. The portfolio conflict problem.
Lead time, and what it buys
The date's distance from today is worth as much as the date itself.
| Lead time | What it gets you |
|---|---|
| Under 48 hours | A booking exercise. Whoever happens to be free. |
| 48 hours | Possible when the date cannot move. Costs more for less choice. |
| 3 to 10 days | The widest creator selection, comfortable briefing, and time for creators to actually use the product. |
| 2 to 3 weeks | The above plus a founder account warmed on the problem space beforehand. |
| 4 to 6 weeks | What a launch from zero needs, because a cold account has no baseline to outperform. The launch from zero. |
Beyond six weeks the returns flatten and the risk of the product slipping starts eating the plan. Booking creators against a date engineering has not committed to is how a campaign gets paid for twice.
Building a quarter
One launch does not do very much on its own, because a buyer needs roughly fifteen encounters before acting and a strong launch day supplies two or three of them. So the useful unit of planning is a quarter rather than a day. Why fifteen.
A workable shape for a company shipping regularly:
- One anchor launch a quarter, fully amplified, with the budget behind it.
- Two or three smaller drops spaced three to four weeks apart, run in-house off the founder account.
- Continuous founder posting between them, which is what keeps the baseline from decaying. A quiet stretch lowers the bar everything afterwards is measured against.
Spacing matters more than count. Two launches nine days apart reach mostly the same people twice and register as one event. The same two spread five weeks apart reach a partly overlapping audience at two different moments, and each starts from a warmer floor than the last.
When the date cannot move
Sometimes it genuinely cannot. A funding announcement tied to a wire, a partner's schedule, a conference slot, a regulatory filing.
Three adjustments make a fixed date survivable. Spend the lead time you do have on the claim rather than the assets, since the angle is the highest-return unfunded work and it is the part that does not need a calendar.
Accept a narrower creator list and select for fit over reach, because the wrong large account is worse than the right small one. And move the exec and customer amplification to the front of the day rather than spreading it, so the first-hour window gets everything you have.
What you should not do is compress the lead time and keep the same expectations. A 48-hour turnaround is a different product from a ten-day one, and the report will say so.
Questions we get asked
What is the best day of the week to launch a product?
Tuesday or Wednesday for a technical audience. Across 1,978 launches, accounts under 25,000 followers did best on those two days, at 98 and 110 times their following, and Saturday did about a third as well as any weekday. The day matters less than the hour and both matter less than whether the claim is worth responding to.
What time of day should we post a launch?
Early enough in the US morning that the whole American working day sits ahead of the post, which for a global technical audience usually means somewhere between 10am and noon Eastern. The reason is the first hour: ranking is decided largely on engagement velocity early, so you want the post live when the largest share of your buyers are awake and available to reply.
How far in advance should a launch date be set?
Three to ten days gives the widest creator selection and comfortable briefing time, and most launches go live inside five. A 48-hour turnaround is possible when the date cannot move, and it buys whoever is free rather than whoever is right, which shows up in the report. Anything under 48 hours is a booking exercise rather than a plan.