Waitlist launch strategy: how the referral queue and the gated waitlist work, what converts, and when a waitlist backf

A waitlist launch strategy collects signups before a product opens, then controls who gets in and when, through a referral-ranked queue or an onboarding-gated list. Plan on confirmation, not sharing: LaunchList's median confirmed-signup rate was 46.4%, while median referral share ran from 0.6% to 1.3% depending on list size. Budget launch traffic from your own promotion and treat referrals as upside.

What a waitlist launch strategy is, in short

A waitlist launch strategy collects signups before a product opens, then controls who gets in and when. Two models are documented. In the referral-ranked queue, each signup gets a place in line and a personal link, and sharing moves them forward. In the onboarding-gated list, the company admits people in batches as it has room to serve them.

Robinhood's United Kingdom launch in November 2023 ran both at once. People who joined got "a unique referral link to share with friends and family to move up the queue," and the company said it would "grant access to our app on a rolling basis over the following weeks, reaching full availability in early 2024" (Robinhood).

Model	What the user does	What the company controls	Fits when (our reading)
Referral-ranked queue	Shares a personal link to climb	Order of access	Early access is the reward people want
Onboarding-gated list	Waits for an invite	Pace of access	Support or infrastructure has a hard ceiling

When a waitlist suits the launch

KickoffLabs says a waitlist works well when the launch has scarcity, such as "limited inventory, beta seats, early-bird pricing, a Kickstarter window" (KickoffLabs). For software as a service (SaaS), it notes a waitlist "can shape onboarding capacity and beta waves" (KickoffLabs). We think a waitlist with no real limit behind it reads as a delay, not a privilege.

The headline figures

LaunchList measured its own waitlists created January 2023 to September 2026. Across all list sizes, the median confirmed-signup rate was 46.4% on the 128 waitlists with confirmation emails switched on (LaunchList).

Referral numbers sit much lower. For lists of 250 to 999 signups, the median referral share was 0.6%, and the top quarter started at 6.3% (LaunchList). Of the 610 waitlists with referral links on, 270 (44%) got no referral signups at all (LaunchList).

A worked example

Take 10,000 signups as an example input, not a benchmark. At a 46.4% confirmed rate, 10,000 signups leave 4,640 confirmed addresses to launch to. At the 1.3% median referral share for lists of 1,000 or more signups, 130 of 10,000 signups would have come through a referral link (LaunchList).

That gap is the planning point. LaunchList describes referral share as "also a rough viral coefficient" (LaunchList), so a median list grows mostly from its own promotion, not from sharing. Budget your launch traffic on that basis, and treat referrals as upside you design for with the referral program playbook and track as a viral coefficient.

How Robinhood's waitlist worked

Robinhood's referral queue had three moving parts, and each one is cheap to build. According to a GetWaitly case study drawn from public reporting, the 2013 page asked only for an email. Then, after signing up, you saw your spot in the queue and the total number of people on the list. The third part was the share link that moved you up.

The scale is a reported figure, not a Robinhood one. GetWaitly, a waitlist tool vendor, describes its piece as based on publicly reported information, and it says the list grew to roughly one million users before the Robinhood app was even available.

The same mechanic, ten years later

The UK launch in November 2023 reused the queue and the referral link covered above. It also set out the exit from the queue clearly: UK residents can join the waitlist today and will be notified when they can sign-up for early access. In our reading, that line did real work. It told people the wait would end and how they would hear about it.

Where the Gold Card waitlist differed

The Gold Card waitlist ran on different rules, and they are worth studying because they break the ranked model:

• Position was partly hidden. People who joined online were told you won’t be able to view your place on the waitlist.
• Order was not strict. Robinhood says someone who joined later may have been randomly selected to be given the opportunity to apply before you.
• Timing was open. Its note reads: This is a gradual rollout and the wait time to join may vary.
• Referrals stopped. The solid gold (invite friends) referral program ended July 28, 2025.

We think this is the gated model: access depends on factors the company controls, such as credit approval, instead of on how many friends you recruit.

A template for your own queue rules

Write these down before the page goes live. Publish them on the confirmation screen.

Rule	Referral-ranked (our reading)	Gated (our reading)
What the signup sees	Position and list size	Confirmation only
What moves you up	Each confirmed referral	Company selection
What you promise	How access is granted	That a rollout is gradual
When referrals end	A stated date	Not offered

Pick one column per launch. Mixing them, for example showing a position but then admitting people out of order, is the point where we think a queue starts to feel unfair. For the referral side, see the referral program playbook and how a viral coefficient is measured.

How Superhuman's onboarding-gated waitlist worked

Superhuman, the email client Rahul Vohra founded, ran the opposite of a referral queue. A person's position on the list did not decide when they got in. Readiness decided it: first the product's readiness, then the user's.

Superhuman's very disappointed score rose from 22% to 32% after segmenting, which still left it below the 40% threshold.

Superhuman's very disappointed score against the 40% thresholdPer cent

Superhuman's very disappointed score against the 40% threshold

Item	Value, %
Before segmenting	22%
After segmenting	32%
Threshold	40%

The launch gate was a survey score

Vohra described facing a multi-year build and needing a way to show his team that Superhuman was not yet ready to launch. He took his measure from Sean Ellis. Ask users "How would you feel if you could no longer use the products?" and count the share who answer very disappointed.

By Vohra's account, Ellis found that companies that struggle to grow always get less than 40% very disappointed. Superhuman's score rose from 22% to 32% once the team narrowed the market segment it surveyed. That was still short of the threshold, so the gate stayed shut.

Entry was through onboarding, not queue position

Vohra onboarded the first several hundred customers himself, often in sessions an hour long. He watched each new user for the first half hour. He usually found 10 or 15 bugs, and the team fixed them before the next week's cohort came in.

He said the process did not come from a wish to manufacture scarcity. Users who came through it, he said, "had really low churn."

What the gate trades away and protects

Model	Gives up (our reading)	Protects (our reading)
Referral queue	Control over who arrives first	List growth and launch speed
Onboarding gate	Speed and list size	Retention, first impressions, word of mouth

Here is a worked example, not a benchmark. Take a list of 10000 signups. If all 10000 get in at once, every one of them meets the same early bugs. A weekly onboarded cohort meets those bugs once, and the next cohort gets the fixed version. Vohra drew the same contrast with a typical tech launch that brings in 10s of thousands of people to try a product.

A gate you can run this week

1. Send the Ellis question to active users and record the very disappointed share.
2. Re-cut the results by segment and keep the segment that scores highest.
3. Hold open access until that segment clears the 40% line.
4. Book a small weekly cohort of onboarding calls from the waitlist and watch the first half hour of use.
5. Log every bug from those sessions and fix it before the next cohort, so time-to-value improves with each wave.

Which waitlist model fits your product

The four models differ in one thing above all: who decides the order. Pick the model whose ordering rule matches the constraint you actually have. The table below is a starting point; the last three columns are our reading, not measured results.

Model	How position is set	Who it suits (our reading)	Effort to run (our reading)	Time to first signal (our reading)	Metric to watch (our reading)
Plain list	Signup order, nothing else	A launch with a fixed date or fixed stock	Low: a form and a launch email	Slow: you learn at launch	Confirmed signups and launch-day opens
Referral-ranked queue	Points from referrals and actions	Consumer apps with a social or bragging angle	Medium: links, position display, fraud checks	Fast: sharing shows within days	Share rate and referred signups per signup
Reward tiers	Referral count against fixed thresholds	Products with cheap perks to give, such as founder pricing	Medium to high: perks must be delivered	Fast, then uneven around each threshold	Share of subscribers reaching the first tier
Onboarding-gated access	The team picks who gets in, cohort by cohort	Software as a service (SaaS) with a hands-on setup	High: founder or team time per cohort	Slow at first, then specific	Activation and retention per cohort

A few details from the tools that run these models. On KickoffLabs, queue position comes from points, and ties are usually broken by signup time, so early supporters still get credit. Its reward-level example is to refer 3 friends for early access, 10 friends for a discount, 25 friends for VIP perks, and it frames tiers as the choice when you want predictable sharing behavior. A visible leaderboard is framed for an audience that is competitive, community-driven, or likely to share publicly.

Set expectations for the referral models

LaunchList reports that most pre-launch waitlists land somewhere between K=0.2 and K=0.6. As an example, not a benchmark: on a list of 10,000 signups, a K-factor of 0.2 adds about 2,000 referred signups in the first round, and a K of 0.6 adds about 6,000. Either way the list grows but does not compound on its own. The same guide says improving share rate by 10 percentage points usually moves the needle more than improving conversion rate by 10 percentage points, so test the post-signup share prompt before the landing page.

Match the scarcity to the model

We think each model needs a different kind of scarcity to feel honest:

• Fixed stock or a fixed date justifies a plain list. Order by signup and say when the batch ships.
• Status scarcity, for a private community, or a product people want to say they discovered first, justifies a referral queue, because moving up is the reward.
• Perk scarcity, such as a capped founder price, justifies reward tiers.
• Capacity scarcity, meaning onboarding hours you cannot stretch, justifies gated access. Size each cohort to the hours you have, and watch activation before widening the gate.

If you cannot name which of the four you have, a plain list is the safer default in our reading.

How to build a waitlist, step by step

Six steps, each with an owner, an output and a check. The owners assume a founder, one marketer and one engineer. On a smaller team, the founder takes the marketer's rows.

Step	Owner	What it produces	How to know it worked (our reading)
Single-field signup page	Marketer	A page that asks for an email address only	Visitors who reach the form mostly submit it
Post-signup page	Engineer	Queue position and share link on screen	New signups share before they leave
Referral credit after verification	Engineer	Credit that counts only confirmed friends	Positions move only when a friend confirms
Fraud and spam check	Engineer	Rules that block fake and self-referrals	No sudden bursts on the leaderboard
Emails while people wait	Marketer	A confirmation email plus a regular update	Opens hold steady between sends
Invite codes for batches	Founder	One code per batch of invites	Each batch can be traced to who used the product

Notes on the steps that break most often

The first minute. LaunchList's guide sets the target as a post-signup page and first email that prompt sharing in the first 60 seconds. Put the position and the share link at the top of that page, then repeat both in the confirmation email.

Test the referral link before you announce the list. Click it from a clean browser, sign up as a friend, and confirm the original account moves. The same guide says this check should be one of the first things you verify works.

Credit only after the friend verifies. LaunchList describes its own rule this way: referral credit is given only once a friend confirms, and self-referrals and repeated referrals from the same IP address are refused. Copy that logic whatever tool you use. Its screening also targets bots, honeypot fills and bursts of signups. A honeypot is a hidden form field that humans never see, so a filled one marks a bot.

If you add side tasks such as following your social account, the guide notes they run on the honor system and you should treat them as engagement rather than proof.

Emails and batches. LaunchList's pitch is that people who hear from you while they wait are still interested on the day you let them in. When you open the doors, issue one invite code per cohort. You can then compare activation batch by batch and size the next one.

For reward design, see the referral program playbook. To track sharing over time, see the k-factor entry.

Setup checklist

• Signup form asks for email only
• Confirmation email switched on
• Post-signup page shows position and share link at the top
• Share link tested end to end from a clean browser
• Referral credit counts only confirmed friends
• Self-referrals and repeat referrals from one IP address blocked
• Update email scheduled while people wait
• One invite code created per batch
• Owner named for every line above

What percentage of a waitlist converts

Four rates get blended into one "waitlist conversion" number, and each one has a different denominator. Track them separately or you will compare your list to a figure that measures something else.

Rate	What it divides	Reference figure
Visitor to signup	Signups ÷ page visitors	No LaunchList benchmark
Signup to confirmed	Clicked confirmation link ÷ signups since confirmation was on	46.4% median
Signup to referrer	Signups who brought in at least one person ÷ all signups	0.3% median
Referrals per referrer	Referred signups ÷ signups who referred	3.34 median (lists of 250 to 999)
Invite to active user	Free signups ÷ people let off the list	Around 50% if released in under a month

The middle three come from LaunchList's benchmarks, drawn from 701 waitlists created January 2023 to September 2026 and measured on October 15, 2026. A confirmed signup there is one who clicked the confirmation link. LaunchList does not benchmark visitor to signup because page visits were not recorded for most of the period. It also found that 270 of 610 waitlists with referral links on got no referral signups at all, which is why the referrer median sits near zero.

A worked funnel

Start with an example from getwaitlist.com: 12,000 visitors produced 276 signups, a 2.3% visitor to signup rate.

1. Confirmed. 276 signups times the 46.4% median gives about 128 confirmed addresses.
2. Referrers. 276 signups times the 0.3% median gives fewer than one person who refers. At 3.34 referrals each, that one person adds about three signups.
3. Active users. Invite the 128 confirmed within a month. As an assumption, half of 128 is 64 active users. Wait past three months and the same list could drop to around 20% of 128, about 26.

So 12,000 visitors became roughly 64 users. In our reading, the confirmation step and the release date move that final number more than referral mechanics do on a list this size, because the referral term rounds to almost nothing.

What to do this week

• Put the five rows above in one sheet with your own counts beside them.
• Measure confirmed rate only from the day confirmation emails went on.
• Count invite to active as a separate activation metric, not as waitlist conversion.
• Multiply referrer share by referrals per referrer to get your K-factor, then decide whether referral work is worth this quarter.

How long a waitlist should stay open, and when it backfires

A waitlist works when the line exists for a reason. It backfires when the line is the only thing on offer. Three failure cases show up again and again, and each one has a visible signal you can check this week.

Demand has to exist first

Robinhood is the case people copy, and they often take the wrong lesson from it. GetWaitly's case study notes that free stock trading was a genuinely exciting idea in 2013, and that the waitlist amplified demand and did not manufacture it. The same write-up puts the risk plainly: "If the underlying offer is weak, a viral mechanic just spreads indifference faster."

KickoffLabs reads the early signals the same way. If nobody joins, fix the positioning before you pour money into paid traffic. If people join but do not share, the incentive may be weak. A referral queue built on a soft offer gives you a low viral coefficient, and no reward tier fixes that.

The three ways a queue goes wrong

Failure case	What you see	What to do (our reading)
Long wait, no contact	Invite emails go unopened	Send an update every week the list stays closed
Cold list never emailed	Launch email lands as a surprise	Email the list before launch day, or re-confirm it first
No real scarcity	Signups ask why they cannot just start	Drop the queue and open signup

On the first two, GetWaitly's advice is to email the list during the wait, because momentum, not silence, keeps a queue alive. The third case is the one founders resist. If you can serve everyone today, a queue is friction you added on purpose, and we think it costs you more signups than the perceived exclusivity gains.

A rule for wait length

Tie the wait to invite batches, not to a launch date you picked. KickoffLabs notes that for software, a waitlist can shape onboarding capacity and beta waves. Use that the other way round: let your onboarding capacity set how long anyone waits.

The method:

1. Decide the weekly batch you can onboard and support without breaking time to value.
2. Divide your list size by that batch to get the number of weeks until the last person gets in.
3. If that number runs past the point where conversion drops (covered in the conversion section above), either raise the batch size or stop growing the list.

For example, not as a benchmark: a list of 10,000 signups with a batch you can only raise slowly means the last people in line wait the longest and convert the least. Invite your most engaged referrers first, treat each batch as a cohort, and fix what breaks before the next batch goes out.

A six-week waitlist timeline with the metrics to track

This is our plan, built for a list you intend to open within six weeks. Each phase has one job and one or two numbers to watch. The thresholds come from LaunchList's waitlist benchmarks for lists of 250 to 999 signups. Confirmed signups are the exception: that size group has too few lists, so we use all sizes.

Week by week

• Week 1: page live. Turn on confirmation emails, click your own referral link from a second inbox, and count daily signups. The median LaunchList waitlist got 17 signups in its first 7 days.
• Week 2: first update email. Ask people to share, state the reward, and start tracking who referred.
• Weeks 3 and 4: push the sharers. If each sharer brings few friends, LaunchList suggests you give them a goal to reach, such as earlier access at a set number of referrals. By day 30, the median list had 57 signups.
• Week 5: first invite batch. Invite a slice of the list and measure how many invitees reach your activation event.
• Week 6: final batch. Invite the rest and start the free-to-paid clock.

Weeks	Metric	Top quarter (75th percentile)	Top 10% (90th percentile)
1 to 2	Confirmed signups (all sizes)	60.7%	75.3%
2 to 4	Signups who referred	2.7%	4.9%
3 to 4	Referrals per referrer	5.75	10.79
3 to 4	Referral share	6.3%	25.3%
5 to 6	Invite-to-activation	Not in LaunchList data	Not in LaunchList data

A worked example

Take a list of 10000 signups as an example, not a benchmark. Lists that size fall in LaunchList's group of 1,000 or more, where the median referral share was 1.3% and the top quarter 19.1%. At 1.3% of 10000, 130 people arrived through referral links. At 19.1% of 10000, 1,910 did.

Referral share also works as a rough viral coefficient. LaunchList's own example: a referral share of 8% means each signup brought in about 0.08 more people on average. Track it weekly. If it stays flat through week 4, keep the goal mechanic and change the reward.

After the doors open

LaunchList notes that signup-to-customer conversion happens after launch and is not in this data. For weeks 5 and 6, compare invite-to-activation against activation rate benchmarks, which Lenny's Newsletter built from a survey of 500+ products. For the paid step, use free-to-paid conversion benchmarks, drawn from a survey of 1,000+ products.

We think these last two numbers decide whether the waitlist worked. A strong referral share with weak activation means you grew a queue of people the product did not hold.

What this page does not cover

This page does not benchmark visitor-to-signup rates or signup-to-customer conversion after launch, and it does not compare or choose between waitlist tools. It also leaves out reward and promotion rules for specific markets, and nothing here is legal or financial advice.

Related on Race to Scale: Referral program examples that worked, and the mechanics behind them; A/B test sample size and duration calculator, with the formula; What is a good activation rate? Benchmarks by product type and how to define yours.