The Rel Attribute Disappeared: We Tracked 20 'Dofollow' Links for 90 Days
As of November 10, 2026, the links in this audit are 90 days old. A dofollow link is not a permanent physical object. It is a line of HTML on a page someone else controls. We wanted to know how often that line changes after the invoice clears, so we bought 20 packages marketed as dofollow placements from 9 vendors. Prices ranged from $79 to $499. We saved the raw HTML at delivery, then re-checked at day 30, 60, and 90. This is not a vendor takedown. It is a receipt audit.
How we checked
Each link got a folder: live URL, anchor text, target page, HTTP status, rel attribute, canonical tag, robots meta, and a dated screenshot. We used view-source and curl, not just a browser plugin. A link counted as dofollow only if the anchor had no rel='nofollow', rel='sponsored', or rel='ugc'. We also noted the odd placements that used rel='dofollow'. That is not a standard value, but it was clearly meant to signal the opposite of nofollow. We checked the exact page, not the homepage.
The scoreboard
| Checkpoint | Dofollow | Nofollow | Sponsored | UGC | Missing or redirected |
|---|---|---|---|---|---|
| Delivery | 20 | 0 | 0 | 0 | 0 |
| Day 30 | 18 | 1 | 1 | 0 | 0 |
| Day 60 | 15 | 2 | 2 | 0 | 1 |
| Day 90 | 10 | 4 | 3 | 1 | 2 |
Missing or redirected includes 404s and pages that redirected to a URL without the link. We did not count a link as alive just because the page loaded.
What changed in the HTML
Eight links changed their rel attribute, and two vanished. One started as a plain anchor and became rel='nofollow' on day 30. Another was delivered with rel='dofollow' and switched to rel='sponsored nofollow' on day 60. One turned rel='ugc' on day 90, which is a strange choice for a guest post but not unheard of. Two pages 404ed or redirected to a homepage with no link. We found no case where a nofollow link turned back to dofollow.
The timing matters. Most of the switches happened after the typical 30-day replacement window. That is not a coincidence. It is also not proof of bad faith. Publishing sites update CMS settings, change editorial policies, and reorganize old posts. But from the buyer's side, the result is the same: you paid for dofollow and you are now hosting a link that tells search engines not to count it in the usual way.
How vendors explained it
Nine vendors, eight affected links. We asked one simple question: why did the rel change? Three blamed the publisher's CMS or SEO plugin. Two said the site had updated its sponsored-content policy. One said the link was always sponsored and the initial report was wrong. One offered a replacement link. One offered account credit. One stopped replying after day 45.
Nobody said, 'Yes, we switched it to save money.' The explanations ranged from plausible to creative. One vendor argued that rel='sponsored' is not a downgrade because it is more honest. That is a fine editorial position, but it is not what the package page said. If a vendor sells a dofollow placement, the buyer should not have to discover a sponsored tag three months later.
The invoice is not a warranty
Most order forms promise 'dofollow placement' at the time of publication. Very few promise 'dofollow for 12 months.' That distinction matters. A link can be live, indexable, and still be useless for the thing you bought it for. If you are buying links as assets, you need contract language that covers re-checks and remedies. If you are buying them as rentals, price them like rentals. The messy middle is where budgets go to die.
This is also why screenshots from delivery day are not enough. They prove what the page said once, not what it says now. A vendor can be perfectly honest on day one and still have a link that changes on day 60 because the publisher installed a new plugin. The only way to know is to check again.
What to check before you call it a win
- View the raw HTML, not just the rendered page.
- Confirm the anchor has no rel='nofollow', rel='sponsored', or rel='ugc'.
- Check HTTP status and canonical. A 200 page with a canonical pointing elsewhere is not a win.
- Confirm the page is indexable and not blocked by robots meta.
- Save a dated screenshot and, if possible, an archive snapshot.
- Re-check at day 30, 60, and 90. Put it on the calendar.
- Read the vendor's replacement and refund policy before you pay.
- Ask whether the guarantee covers rel changes, page removal, or sitewide policy shifts. Get it in writing.
The HackLink Shop take
We sell backlink packages, so it would be easy to wave this away. We are not going to do that. The audit shows a boring truth: a dofollow link is a temporary condition, not a permanent title. Some changes are legitimate compliance. Some are sloppy. Some are bait. You cannot tell which from an invoice. You can tell from the HTML, checked over time.
After 90 days, 10 of our 20 links still passed the dofollow test. That is a 50% survival rate, which is not a great ad for the category. The eight rel switches and two disappearances are the reason we now ask vendors for a rel audit clause. If they will not put it in writing, we treat the placement as a rental and budget accordingly. That is less exciting than an 'instant authority' package, but it is a lot easier to defend to a client.
