How to do a viral launch on X

A viral launch looks like luck. It isn't. Every launch that looks effortless is usually the result of weeks, even months, of preparation, and the preparation is surprisingly very mechanical. I know because we've done it repeatedly: our launches have generated over 100 million views with no agency and no ad spend.

When we launched Okara, the launch tweet did 14 million views in a day, and we got over 3M visitors during that launch.

https://x.com/askOkara/status/2033562024651968657

But what if I told you just three weeks before this launch, we had another launch which got 70M views in 3 weeks, all organic. And just 6 weeks before this launch, we had done another launch with 1 million organic views.

In 2025, we were going through the deepest hell of pivots and we were building in public. Our loop to test ideas was simple: build a product, do a launch, get feedback, decide on the fate of the product. A launch is everything on steroids, you get thousands of users and real feedback that without a launch would usually take months or years. That's why as a team we swear by public and direct launches.

People assume there must be some secret behind a successful launch. There isn't. There's just a sequence of steps, most of which are unglamorous, and very tedious.

In the initial days of these launches, I asked my teammate to write simple list of what are the things we need to do before a launch. Once the list is done, add sub-tasks and so on till you get to the atomic tasks. We have sweared by this process and it has helped us a lot. Here is my simple sequence of steps to a successful launch.

the launch requisites.

product press release launch video day and time influencers, users, friends copywriting skills X thread launch day newsletter afterward

it starts with the product.

https://x.com/naval/status/1946792401433161869

No launch technique saves a product people don't want. Ask me, because we experimented with multiple product angles before Okara really landed. Our most viral launch got 70 million organic views but no product users.

But even good products fail at launch anyway, for reasons that have nothing to do with the product's quality and everything to do with the first 60 seconds of using it.

A launch sends thousands of strangers to your product in a short span of time. They arrive skeptical and impatient because they are coming from X and their attention span is hardly 30 seconds. You have to make a statement in 30 seconds. If they have to answer a survey, invite their team, and do multiple steps of onboarding before anything happens, they leave. So the rule is: show value in the first 30 seconds, and get to the "aha" within 60.

At Okara, you type in your website, log in, and agents start working on it. We ask for nothing until the product has already demonstrated itself. If you want to know whether your product passes this test, don't guess. Watch five strangers use it and time them. If it takes them more than a minute to understand why the thing exists, fix that before you plan anything else.Your product is ready for the launch when you have nailed the first 30 seconds and 60 seconds experience.

press release.

At Amazon there is a culture of writing 1-page press releases and FAQs before the product is built.

Before the launch, write that one-pager and FAQs: what exactly you're launching, what problem it solves (in the customer's words, not yours), who it's for, and what it costs. A one-page TLDR of the product, bells and whistles included. Write frequently asked questions related to the product.

This page becomes the source of truth for everything downstream: the video, the thread, the influencer packet, the newsletter. It will keep the team and internal and external stakeholders aligned. It's also a test. If you can't write this page crisply, you're not ready to launch. You're not confused about marketing; you're confused about the product.

This press release is exactly what you would pass on to a journalist if you were looking to get featured in press media.

the video.

The script is the life of the video, and this is where your press release earns its keep. Use it as the guide for what the video says. Then decide the outcome before you get to production: is this a product demo, awareness, conversions? A video made without a decided outcome ends up optimizing for nothing.

We've done high production, animation, and plain screen recordings, and the surprising conclusion is that the format barely matters. What matters is the fit between the product, the problem, and the channel. If your product is solving a hair-on-fire problem in a novel way, a simple screen recording is enough to excite the users.

If you do animate, keep it simple: a clean demo, music that fits, under 2 minutes. The moment viewers notice the animation or production instead of the product, you have lost.We did a high-production launch in the past and spent considerable time on the quality of the video. Guess what! There was more discussion on the quality of the video than the real product. Seeing all those comments piling up praising the video quality really made me uncomfortable, because the product was overshadowed by the ad.

One warning if you're launching an AI product. Eighteen months ago, everything built with AI went viral, because anything made with AI looked novel. That novelty has worn off. "We added AI" is no longer a launch; it's a feature note. You now have to think deeply about the problem and the outcome, and lead with those. Our video doesn't open with agents. It opens with: enter your website, get traffic.Whatever the format, the first three seconds show the product or state the problem. No logo animation. No "hey guys." Get to the point.

pick a day, and recruit before it.

The first hour of the launch is the most important time, it defines whether your product's reach will expand or decay. If more people are engaging, the algorithm will show it to more people. If there is no engagement, the reach will contract. Algorithms reward early velocity, and early velocity comes from people who were aligned to participate in your launch.

So weeks before, reach out to friends, power users, ex-colleagues, VCs, and influencers who are posting actively on X to participate in the launch. (The influencer part is the most time-consuming one. I'll talk about it in a separate section later.)

You can email something like this:

Anyone who says yes gets a calendar invite. This sounds absurdly small, but the calendar invite is half the trick. People aren't unwilling to help; they're busy. A notification at launch minute is the difference between a launch that goes viral and a launch that dies.

Ask beforehand. Get commitment to participate in your launch. Organize about fifty people, influencers, friends, power users to retweet and quote-tweet within thirty minutes of launch.

As for the day itself, launch when your target audience is awake and idle and don't volunteer to compete with the news. You can't control whether OpenAI or Anthropic announces something that day, but you can at least check the calendar.

We personally feel Tuesday 9am EST is the best time to launch a product for the US audience, especially if you are in an early stage with no traction on social channels.

recruit influencers.

We have never outsourced recruitment of influencers to agencies. Firstly because the launch experience is very core to a product and we don't want to outsource that experience to an agency. Secondly, agencies are expensive: they charge a commission which is usually 15% to 30% of the campaign budget, they add extra pricing on top of influencer rates, and they take 4 to 6 weeks minimum to close influencers.

We usually engage with X influencers who have more than 10K followers, are consistently posting related to our niches, and have a baseline reach of a couple of thousand consistently.

In the initial few launches, we reached out to them directly, negotiated the rates, signed deals, followed up on launch day individually, and then cleared all payments one by one.

This was a very time-consuming and boring process, but we had to do it because we wanted to own and control the narrative and outcome. We eventually automated this whole process into Okara's influencer agent, mostly out of exhaustion.

https://x.com/askOkara/status/2066899565484413070

If you are doing it on your own, here is the process.

1. DM X accounts in your niche with 10K+ followers, who are active and getting minimum 3K reach per tweet.
2. Ask them to share rates for retweets and quote tweets.
3. Negotiate with them.
4. Important: follow the TOS of the X platform, add a sponsored tag to paid tweets.
5. Sign the deal.
6. Close at least 50 influencers, for this you will have to reach at least 500 accounts.
7. Share your product with them to use.
8. On launch day, individually DM all of them to tweet.
9. Once the launch is over, ask them to share proof of tweet.
10. Clear their payments one by one.

Give these influencers access to the product a week early, so their posts are specific instead of generic; specific posts are the ones that perform. Tell them you want their honest review, and readers can smell coordinated copy from a mile away.

Tip:

We're noticing that quote tweets with memes and brief one-line content increase reach.

The most effective trick in the whole brief is to show what good looks like.

Screenshot quote-tweets you saw going viral from other people's launches and paste them into the doc. People are bad at writing from instructions and good at writing from examples. "Something like this, in your voice" gets you ten times better posts than "please support our launch."

the thread.

This is where launches are won or lost, and within the thread, the hook is most of it. People decide whether to stop scrolling within the first few words. The attention span on X is extremely low. If the opening isn't bold and doesn't create curiosity, surprise, or emotion, nothing after it matters, because people will not engage with it.

The way to learn hooks is not to be clever but to be empirical. Pull up the twenty biggest product launches of the past year and read only their first lines. Don't copy the wording; notice the patterns. A claim that sounds impossible. A number. A villain. A before and after. Then write twenty hooks of your own and kill nineteen.

The copy should be cut-throat and straightforward, and it should be at least one of three things:

1. Emotional, so the reader feels the pain you're solving.
2. Divisive, so it takes a side.
3. Or a hot take, saying the thing people believe but won't post.

These sound like risks. The actual risk is safety. If nobody could possibly disagree with your tweet, nobody will reply to it either, and replies are fuel.

The first tweet of your launch will have the video attached and a CTA to your website. Keep the URL clean. Your link will be typed from memory by someone who saw a tweet an hour ago on their phone.

launch day.

Congrats! It's launch day, and if you've done everything above, 90% of the work is already behind you. Phew.

The moment your launch tweet is live, ping your power users, friends, family, and influencers with the direct link.

Over the next few hours, repost every quote tweet around the launch, ship the newsletter, and watch what's landing.

the newsletter.

If you have an email list, even a small one, use it. Email is the only channel where no algorithm sits between you and your audience. Keep the email short: what you launched, why it matters, one link. Ask your audience to engage with the launch tweet in the next hour. Your subscribers are your warmest audience. They just need to be told exactly what helps.

Your newsletter is actually your one-page press release, but you attach your launch thread and ask people to support your launch if they like your product.

afterward.

Virality decays in 48 hours. Follow up with everyone who engages.

And decide, before launch, what a visitor is supposed to become: a customer, a lead, a signup. A launch that produces 100,000 visitors and no revenue is a fireworks show. Ours convert because pricing is visible and signing up takes less than a minute.

Which brings up the biggest misconception of all: that you get one launch. You don't. Every feature, milestone, and integration is a launch, and every launch makes the next one bigger, because the audience compounds. None of this is magic. The companies that win at distribution aren't the ones that launch well. They're the ones that launch constantly.

https://x.com/askOkara/status/2075515973520752798