What $387 of bought links actually did for a new site

$387.11. Three links, bought over two days in July. One of them still doesn’t exist.

Almost every article about buying links is written by somebody who sells them, and the numbers are always vague on purpose. These are mine, off my own ledger, and I still have the invoices.

Here’s the spend:

• $70, guest post on a tools blog, one-time, dofollow. Live and still live.
• $150, listing on a scraping-tools site, six-month term. Paid on 12 July. Eight days later, nothing on the page.
• $167.11, guest post placed through an agency onto an IP lookup site. Live, dofollow, with the anchor I asked for.

I’m not naming the placements. Publicly labelling a specific URL as a paid link devalues it and drops the publisher in it, and I’d want the same courtesy. The numbers are what’s useful here anyway.

The site I pointed them at

This matters more than the links do.

The site had 40 pages indexed, zero ranking keywords, and one click from Google search in the previous 180 days. One click. In six months, across 40 pages Google had already crawled and filed away.

So the constraint was never authority. The constraint was that nothing on the site answered a question anyone was actually typing into a search box. I knew that when I bought. I did it anyway, partly to test the vendors and partly because I wanted placements banked before the content caught up. Whether that was smart is genuinely arguable, and I’ll come back to it.

The $150 that vanished

The failure is the most useful part, so let’s start there.

Paid 12 July. Invoice in hand, confirmation email in hand, and eight days later there was still no listing on the page. The vendor had said they’d email when it went up. The email never came.

That’s 39% of the entire spend sitting in someone else’s account with no asset attached to it. Not lost exactly, just parked somewhere I couldn’t point a link from, for an amount of time nobody would commit to. When I chased, the reply was friendly and vague, which is its own kind of answer.

Calling it a scam would be wrong. It was a small team with a queue and my order sat in that queue behind other people’s. Ordinary operational slippage. That failure mode is far more common than fraud, and nobody warns you about it because it’s boring.

So: put a date in your calendar for every placement you pay for, about ten days out, and chase on that date. Nobody else is tracking it.

The invoice was not one number

Worth pausing on how that $167.11 assembled itself, because agency link pricing is never the figure on the page.

List price was $157.90. A 10% discount came off, $15.79. Then three days later, a $25 add-on fee to get it onto the specific site I wanted rather than whatever was in rotation that week.

Advertised: $158. Actually left my account: $167.11. About 6% over sticker, which is small, but it arrived as a separate charge on a separate invoice three days after the first one. Budget twenty placements and that gap stops being a rounding error.

Ask what the site-selection fee is before you order. It’s almost never in the headline price.

What “verified live” has to mean

When the other two went live I didn’t take the vendor’s word for it.

Verified means you personally opened the page and read the rel attribute on your link. It takes about thirty seconds. If it says nofollow or sponsored, you bought an advert, which is fine if that’s what you agreed to pay for. On both of mine the attribute carried only noopener, which is a security attribute and passes nothing on, so both were genuinely followable.

Check the anchor text in the same pass. That’s the other thing that quietly drifts.

The spam scores, and the one that surprised me

I vetted all three on backlink data before paying rather than on gut feel:

• $70 site: 569 referring domains, 16% spam score
• $167.11 site: 7,800 referring domains, 677,100 backlinks, 26% spam score
• $150 listing: 938 referring domains, 44% spam score

44%. Comfortably the dirtiest profile of the three, and going in it felt like the most legitimate purchase of the lot. Real company, real product, real audience, a name you’d recognise. The spam score doesn’t care about any of that. It looks at who links to them, and a lot of who linked to them was junk.

I bought it anyway. I decided audience relevance was worth it and I’d probably make the same call again. But I made it with the number in front of me, which is the whole point. Treat vetting as pricing information rather than as a gate. The score doesn’t tell you to walk away, it tells you what you’re actually buying, so when a placement underperforms you already know why and you paid accordingly.

Domain rank is close to useless as a buying signal

This is where I’ll lose some of you.

My three sites had ranks of 307, 210 and 464. Line those up against what happened. The best rank number belonged to the listing that never went live and carried a 44% spam score. The worst rank number belonged to the placement that shipped clean, dofollow, correct anchor, on a site with nearly 8,000 referring domains.

The score told me nothing useful. Referring domain count, spam percentage and whether the page is genuinely about my topic told me everything.

Vendors lead with the authority score because it’s the number that’s easiest to inflate and hardest for a buyer to argue with.

Relevance, briefly

The two that shipped sit on sites where a mobile proxy link makes topical sense. One covers automation tooling, the other covers IP lookups and network topics. A reader on either page could plausibly want what I sell.

Relevance is the one factor that has survived every algorithm update anyone has written about, and it’s also the factor vendors will quietly ignore if you let them. Their inventory is their inventory, and they’ll place you wherever there’s a slot this week unless you name the constraint. Name it in the order, and be willing to wait for the right site instead of taking the available one.

The renewal trap

The listing was sold as a six-month term with automatic renewal. That’s the standard shape for directory placements and it’s where money leaks. Six months later you get billed again for something you’d forgotten you owned.

I had the renewal cancelled and got it in writing before the money went out: an email stating the renewal was paused and I wouldn’t be charged automatically. Keep that email. A verbal assurance in a support chat is worth nothing in six months when a different person has the account.

What I can’t tell you

Here’s the part these articles usually skip.

I can’t tell you what these links did to my rankings. It’s been about three weeks. Anyone showing you a ranking chart three weeks after a link purchase and drawing a causal arrow between them is selling you something. Link effects, where they exist, take months and get tangled up with every other change you made meanwhile.

It’s worse than that. I can’t show you a traffic graph either, because the analytics project behind that site was deleted in an unrelated platform mess and reporting has been broken since. So even the observational data I should have isn’t there.

I’m telling you that because leaving it out would make this piece look more rigorous than it is.

So was it a good spend

Premature, and I’ll say that plainly.

A site with zero ranking keywords doesn’t have a link problem. It has a nothing-to-rank problem. Links make an existing position stronger, they don’t conjure a position out of 40 thin pages nobody searches for. Spending $387 on authority before there was anything to point it at was solving the wrong constraint first, which is a mistake I’ve watched other people make and then made anyway.

Hand me the same $387 today and it goes on content targeting things people actually search for, getting pages onto the second or third page first. Then links pointed at those specific pages. Links applied to a page already ranking eleventh do something you can measure. Links applied to a page ranking nowhere do nothing you can see.

And I’d still buy them. This site sells placements too, so you know exactly where I stand. Just in that order, after the content rather than before it.

The vetting checklist I run before paying anyone is here.