Three months after launch, Shipfolks has 115 accounts, 59 public builder profiles showing 122 shipped projects, and it has not made any money. Nearly all the sign-ups came in the first week. No one has sent a message through the site since August.
Most "three months in" posts are written when the numbers are good. These are not, and that is the reason to publish them: a directory that asks clients and builders to trust it should not be vague about its own size. Everything below was counted from the production database on 11 October 2026.
The whole picture in one table
| What | Count |
|---|---|
| Accounts created | 115 |
| Builder profiles live and public | 59 |
| Builder accounts started but never submitted | 48 |
| Client accounts (people signed up to hire) | 6 |
| Projects on builder profiles | 122 |
| Products in the directory | 29 (28 added by us, 1 by a member) |
| Conversations started through the site | 4 |
| Messages sent | 16 |
| Paid placements bought | 0 |
| Revenue | $0 |
Two of those need a sentence each. The 59 profiles exclude the founder's own. And "added by us" means the product directory was seeded by hand with well-known tools so it was not empty on day one, which is why 28 of its 29 entries are things like Supabase and Linear rather than member submissions.
Sign-ups were one good week
| Month | Accounts created | People who signed in |
|---|---|---|
| July (from the 22nd) | 106 | 104 |
| August | 6 | 13 |
| September | 3 | 8 |
| October (to the 11th) | 0 | not counted yet |
The launch was the traffic. The launch posts brought a wave of people in the first days, and nothing since has replaced it. A directory needs a reason to come back, or a way to be found that does not depend on someone posting about it, and at three months it has neither in any meaningful amount.
The sign-in column is the more honest of the two. Accounts are cumulative and only ever go up. People signing in is the closest thing we have to "people still using it", and it fell from 104 to 8 in two months. It is a rough measure, because someone who stays signed in is not counted again, but a fall that size is not a measurement error.
Almost half of builder sign-ups never finished
Of 109 builder-side accounts, 59 went live, 48 started and never submitted a profile, one was removed, and one is the founder's own.
We do not know why the 48 stopped, and guessing would be inventing a finding. Two things are true about the form, though. It asks for at least one shipped project with a permanent link (your own domain or a store listing, not a preview deployment), which is a real bar for someone who has not shipped yet. And every profile is reviewed by a person before it goes public, so finishing does not mean appearing straight away. Both are deliberate. Both cost sign-ups.
The demand side is the problem
A directory of builders is only useful if someone comes to hire one. So far, almost nobody has.
- 6 client accounts, against 59 builders.
- 13 clicks on "contact" in three months: 9 in July, 4 in August, none since.
- 4 conversations and 16 messages, all in July and August.
That is the number that matters, and it is close to zero. Everything else on this page follows from it: builders do not come back because nothing happens when they do, and paid placement has nothing to sell, because more visibility in front of no one is worth nothing.
Revenue: none
Paid placements (Boost, Featured, Spotlight and Partner) have been live since 24 July. Nobody has bought one. The two ad slots have run an unpaid house ad the whole time. The free tools on the tools page take no money by design.
None of this is a surprise given the section above. Paying to be seen only makes sense once there is someone to be seen by.
Page views look healthy, and mostly are not people
The database records 8,769 project page views and 4,801 profile page views since launch, and September was the busiest month for both. That looks like a site that is growing. It is not.
In a 15-day sample of our server logs, 29% of page requests came from software that openly identifies itself as a crawler, before counting anything that hides it. Views are also suspiciously even: across the 110 projects listed in July, the 10th percentile has 44 views and the 90th has 70. Human attention does not spread that evenly. Crawlers visiting every page in turn do. How much of a small site's traffic is bots has the full count.
Until 11 October, the view counters on our own pages, the ones builders see on their profiles, counted this automated traffic along with everyone else. Since then they skip anything that names itself as a crawler and requests that do not look like they came from a browser. The 30-day figures will be clear of the old traffic by 10 November; the all-time totals keep it. Either way, read them as an upper bound, not a headcount, because a bot that imitates a browser well still counts.
What these numbers mean if you are deciding whether to list
If you build things and want a page that shows them: a Shipfolks profile works as a portfolio link you can send to anyone, it costs nothing, and your work sits on a permanent page with its real links. That part works today, whatever the traffic does.
Do not list expecting clients to find you. At this size, that would be a promise we cannot keep. Where freelance developers actually find clients covers the channels that do the work, and none of them is a directory three months old.
If you are hiring, the 59 profiles are real people with real shipped work, every one checked by hand before it went up. There are not many of them. Browse the ones open to work and judge the work yourself.