How to track traffic from a Product Hunt launch
Product Hunt's Visit website button sends no referrer. It tags the link with utm_source=producthunt instead, so whether launch visits land under Product Hunt or under Direct depends on your tool. What to check before launch day, what the live view is good for, and how to add it up afterwards.
In this article
A Product Hunt launch sends people to your site through several doors at once, and the most obvious one works in the least obvious way. The Visit website button on a Product Hunt page sends no referrer. It tags the link instead, so whether your analytics files those visits under Product Hunt or under Direct depends on whether it reads the tag.
This guide covers where each kind of launch visit shows up, what to check before launch day, what the live view is good for while the launch runs, and how to add it all up afterwards. It works with any analytics tool. The Product Hunt details come from its help center and from the links on its product and launch pages, as of November 2026.
Where launch visits come from
Every launch visit arrives with a referrer, a tag, both or neither, and that decides where it lands in your reports:
| Where the click happens | What reaches your site | Where it usually shows up |
|---|---|---|
| The Visit website button on your Product Hunt page | No referrer. The link carries ref=producthunt and utm_source=producthunt. | As producthunt in tools that read utm_source. As Direct in tools that only read referrers. |
| A link to your site in the comments under your launch | The referrer https://www.producthunt.com/, without the page path, and no tag. | As producthunt.com or www.producthunt.com, among your referrers. |
| A post on X that links straight to your site | Often the referrer t.co, X's link service. | As t.co, or under a social channel, unless you tagged the link. |
| Your link in an email, a chat app or a document | Usually no referrer. | As Direct, unless you tagged the link. |
| A link to your Product Hunt page, shared anywhere | Nothing yet: the visitor goes to Product Hunt first. | Through one of the first two rows, if they click on to your site. |
The last row matters more than it looks. A UTM tag on a link to your Product Hunt page describes a visit to Product Hunt, not to your site, and it doesn't travel any further. People who follow it reach you through Product Hunt's own button or a comment, and show up the way those do.
What Product Hunt adds to your link
Open any product's page on Product Hunt, right-click Visit website and copy the link. As of November 2026, it has this shape, with the maker's own address where this example has example.com:
https://example.com/?ref=producthunt&utm_source=producthuntThe same link carries rel="noreferrer noopener ugc", and noreferrer, as MDN describes it, tells the browser to leave out the Referer header. So the tag is the only thing that says where the click came from:
- Tools that read utm_source show these visits as
producthunt. Some readrefas well: Plausible's docs, for one, listrefandsourcenext to theutm_tags. - Tools that only go by referrer have nothing to go on and file the visit as Direct. That alone can make Direct jump on launch day.
- There's no utm_medium. Tools that sort sources into channels use the medium for most of their rules. GA4, for one, only puts a visit in its Referral channel when the medium is referral, app or link, and
producthuntisn't on its lists of social, search, video or shopping sites, so with no medium these visits end up in Unassigned, the channel GA4 uses when no rule matches. Look forproducthuntas a session source in the Traffic acquisition report rather than in the channel totals.
Product Hunt's help page on building your product page asks for a direct link to where people can use or download the product, and no link shorteners. Product Hunt adds its own tag to that link, so the plain address of the page you want people on already arrives tagged, without you adding anything.
Before launch day
- Check every page people can land on. The button sends people to the address you submitted, but links in comments and shares can point anywhere: pricing, a demo, the sign-up page. Open each one in a private window, with blockers off, and look for your visit in the tool's real-time view. On a single-page app, click between pages and check that each one counts as a pageview, not only the first. Do this a day or two before, not at midnight.
- Test that a tag survives. Open your site with
?utm_source=teston the end, let any redirect happen (to https, to www, to a language folder) and check that the tag is still in the address bar and that the visit shows up with it. This is the same check as the callout above, done before it matters. - Tag the links that point straight to your site. An email to your own list, a post that links to your site rather than to Product Hunt, a message in a community. Without tags, most of these end up in Direct or under a generic social referrer. A small UTM naming system keeps the values consistent, and the same article explains why links between your own pages should never carry tags.
- Write down a baseline. Visitors, pageviews and top sources on the same weekday over the past few weeks. Without one, you can see the spike but not how much of it the launch added.
While the launch runs
A live or real-time view shows who is on the site right now. On launch day it's good for three things, and judging the launch isn't one of them:
Confirming that visits arrive
As soon as people click through, visits should appear: tagged producthunt in a view that shows tags, as Direct in one that only shows referrers. If your page is collecting upvotes and the live view stays empty, check the script and the redirect before anything else.
Seeing where people land
The button sends people to the address you submitted. Comments and shares can send them anywhere. If many land on a page you didn't plan for, make sure that page explains the product too.
Catching broken links
If visitors pile up on a not-found page that carries your script, a link somewhere is wrong. The referrer or the tag on those visits tells you where to look.
Expect Direct to rise alongside Product Hunt. Besides tools that ignore the tag, people pass launches on in chat apps and emails, and those clicks usually arrive with no referrer. What direct traffic really is covers what else hides in that bucket. Posts on X that link straight to your site often show up as t.co: according to X's help center, every link shared on X goes through that link service.
The live view tells you the launch is running. It can't tell you if it worked.
One launch, two calendar days
Product Hunt's help center says a launch is visible from midnight Pacific Time and runs for the entire day. Midnight Pacific is 08:00 UTC while the US is on standard time and 07:00 UTC in summer, which is nine in the morning in Central Europe for most of the year.
If your dashboard draws days in another time zone, the launch lands across two of your calendar days, and neither looks as big as the launch was. Compare a 48-hour window instead, or switch the report to Pacific Time if your tool lets you. Why no two analytics tools agree on your traffic has more on how each tool decides when a day starts.
Add it up afterwards
Launch traffic is the sum of a few parts, and they aren't equally certain:
- Visits tagged producthunt, plus visits with the producthunt.com referrer. Firm: these came from Product Hunt.
- Visits from your own tagged launch links. Firm too, under the source and campaign values you chose.
- Social and Direct above their usual level for that weekday. An estimate. You're subtracting a baseline, and some of it would have come anyway.
Report the firm part and the estimate separately, and say which is which.
Then look past the count:
- Entry and exit pages. Entry pages show where launch visitors started, exit pages where they left. If most entered on the home page and left from it, that page did all the work.
- Bounce rate, with care. In a tool that counts every one-page visit as a bounce, a launch visitor who reads one page and goes back to the list is a bounce, so a high rate on launch day says little on its own. GA4 decides differently, by engagement. A bounce isn't a bad visit explains what the number does and doesn't measure.
- Countries and devices. A launch can bring a different crowd from your usual one. Compare launch day's split with your baseline, and if phones are a bigger share than usual, try your key pages on one.
What the launch left behind
The launch-day spike is the least useful number. Compare the week after the launch with the week before it. Whatever stays above the old baseline is what the launch added beyond the day itself: people who came back, links from other sites, searches for your name.
Check how your tool counts unique visitors before you add days together. One that counts without cookies, from a key that changes every day, counts someone who comes back the next day as a visitor on both days. How to count unique visitors without cookies goes through the methods and where each one miscounts.
If you take payments, put revenue on the same timeline as visitors. A chart like that shows whether launch week brought money or only visitors. It can't show which source earned it, as Stripe revenue and website traffic on one dashboard explains.
Why the numbers won't match Product Hunt's
Upvotes, comments and follows count what people did on Product Hunt. Your analytics counts visits to your site. Someone can upvote without ever visiting, and someone can visit three times without upvoting, so the two never line up and aren't meant to. Two analytics tools on the same site won't agree with each other either, for reasons covered in why no two analytics tools agree on your traffic.
The short version
- Know that the Visit website button sends no referrer, only a tag.
- Find out whether your tool reads utm_source, and where it files a source with no medium.
- Test every landing page, and test that a tag survives your redirects.
- Tag the links that point straight to your site, and none between your own pages.
- Note a baseline for the same weekday.
- Use the live view to confirm visits arrive, not to judge the launch.
- Compare a 48-hour window, or work in Pacific Time.
- Report the firm part and the estimate separately, then compare the weeks around it.
Lomen reads the five utm_ tags on a link and nothing else in the query string, so a click on Product Hunt's button is stored with the source tag producthunt and the ref is ignored. Its Sources list and its live view go by referrer, though, so both show that visit as Direct, while links from comments show as producthunt.com. To count the tagged visits, ask Lomen AI, which can break launch day down by source, landing page and UTM tag against the same weekday in earlier weeks; each answer counts toward the three AI insights a month the beta includes. The live view shows who has opened a page in the last five minutes, with the page they're on and the page their visit began on. Once a site has a few days of history, the traffic alert can email you when the last 24 hours are far busier than usual. Days follow your site's own time zone, so a launch can span two of them. Lomen has no custom events or funnels, so it counts the visits a launch brings, not the sign-ups. Lomen is free during the public beta.