How to Track Which Directories Actually Send You Traffic
Attribution for directory listings: UTM parameters, the referrer fallback when platforms strip them, and the 30/90-day window that reflects reality.
Published 2026-09-07 · 4 min read
Founders submit to a dozen directories and then cannot say which one worked, because the tracking was not set up before the submissions went in. Attribution is retroactively impossible — which is why this is worth an hour before the first submission rather than a day after the tenth.
The other problem is timing. Directories ramp slowly, so measuring at day one always shows zero and always produces the wrong conclusion.
The three layers of attribution
| Layer | How it works | Reliability |
|---|---|---|
| UTM parameters | You append source, medium and campaign to the link you submit | Best where honoured — but many directories strip or rewrite links |
| HTTP referrer | The browser reports which page sent the visitor | Works regardless of UTMs, but is stripped by some privacy settings and redirects |
| Landing page correlation | A directory-specific URL path or page | Most reliable, most work; worth it for the two or three that matter |
Setting up UTMs properly
- Use a consistent scheme. `utm_source=<directory>&utm_medium=directory&utm_campaign=ai-launch-2026`. Consistency matters more than the specific values, because you will be reading these in aggregate months later.
- Lowercase and hyphenate everything. `Product_Hunt` and `product-hunt` are two different sources in your analytics.
- Do not put UTMs on the link you give to humans in conversation. They are ugly and they leak your tracking scheme. Use them on directory submissions, not in emails.
- Record every UTM string you submit. In the same tracking table as the submission. If you cannot look up what you sent, you cannot interpret what arrives.
When UTMs get stripped
Many directories rewrite outbound links — for click counting, for affiliate attribution, or to add their own parameters. Some strip yours entirely. This is common enough that referrer-based segmentation is not a fallback but a requirement.
- 1
Segment by referrer domain
Every analytics tool can group sessions by referring site. Build a saved view of all referrals from directories, and check it against your submission list.
- 2
Match referrers to submissions
Any directory you submitted to that does not appear in referrals either sent nobody, or is stripping the referrer too. Those need the third layer.
- 3
Use a distinct landing path where it matters
For your two or three most important listings, submit a link to /from/<directory> rather than your homepage. Crude, unambiguous, and immune to both stripping problems.
- 4
Check for redirect chains
A directory that redirects through its own domain may drop the referrer. Look at the actual link in the HTML rather than assuming it points at you.
The measurement window
The mistake is judging a directory in its first week. Listings are found through search and category browsing, both of which take time to build. A realistic schedule:
| When | What to check | What it means |
|---|---|---|
| Day 7 | Did the listing go live? Is it in the right category? | Process, not performance. Fix problems now |
| Day 30 | First traffic numbers, baseline established | Early signal. Do not conclude anything yet |
| Day 90 | Cumulative visitors, signups, activation rate | The number that matters. Compare directories here, not earlier |
| Day 180 | Trend direction, and whether the listing still exists | Decides which ones to maintain and which to drop |
What to measure besides visits
- Activation rate per source. A directory sending fifty visitors who sign up beats one sending five hundred who do not. This is the metric that should decide where you invest next.
- Position within the category page. If you can see it, note where your listing sits. Position changes with new submissions and with how the directory orders things.
- Whether the listing still exists. Directories prune, restructure and shut down. A quarterly check takes ten minutes.
- Comments and questions received. Some directories produce conversations rather than clicks, and those are worth more per unit.
A minimal setup that works
- 1
Before the first submission
Decide the UTM scheme. Create the tracking table with one row per planned submission. Set up a saved referrer segment in your analytics.
- 2
At each submission
Record the exact URL you submitted, the UTM string, the date, and the confirmation details.
- 3
Weekly for the first month
Check which listings went live and whether any traffic has appeared. Fix placement problems while they are easy.
- 4
At day 30 and day 90
Record cumulative visitors, signups and activation per directory. Compare them to each other, not to your hopes.
- 5
Quarterly
Verify the listings still exist and still describe the product accurately. Update what has changed.
What to do with the answer
The point of tracking is a decision, and it is usually one of three:
- One or two directories clearly work. Find five more like them — same audience shape, same category depth, same link behaviour. This is the highest-value output of the whole exercise.
- None work, but activation is fine. Your listings are not being seen. Revisit category choice and listing quality before adding more directories.
- Traffic arrives but does not activate. The problem is not the directory. Stop submitting and fix the product or the landing page.
You cannot improve a channel you cannot see. An hour of tracking setup before your first submission is worth more than every hour spent submitting afterwards without it.
More in Launch platforms & directories
- Is Paying for Visibility Worth It for a New AI Product?When paid placement pays and when it does not: the four conditions, what it can and cannot buy, and how to test it with the smallest spend.
- What to Do Before, During and After a Directory SubmissionA submission is a small campaign, not a form fill. The preparation, the fifteen minutes of the form, and the follow-up that determines whether it works.
- Niche vs General Directories: Where a Small AI Tool Gets SeenA general directory has more visitors and more competition; a niche one has fewer of both. How to decide, and why the smaller one usually wins early.