A client asked me this in March, right after Google’s core update had finished chewing through the internet. He’d read somewhere that AI killed SEO, that links don’t matter anymore, and that he should stop wasting money on them.
Fair question. He’d spent $400 on a backlink package the year before and gotten exactly nothing for it. He’s not unusual. Half the small business owners who come to us have a story like that.
So here’s my honest answer, the same one I gave him: yes, backlinks still work in 2026. But about half of what gets sold as “backlinks” stopped working years ago, and most small business owners can’t tell the difference between the two. That gap is where all the money gets wasted.
Let me show you what the data actually says, what died, and what’s still worth your budget.
The numbers haven’t gone soft
WebFX published a study a few weeks ago that looked at 1,462 domains ranking on page one across 15 industries. The median site holding a page-one spot has 907 referring domains pointing at it. In finance, the median jumps past 3,000. In apparel, it drops to 76.
Two things jump out of that study for me.
First, the range. Anyone who tells you “you need X backlinks to rank” without asking what industry you’re in is guessing. A local service business and a fintech startup are playing completely different games.
Second, and this is the number I keep quoting to clients: 92.2% of the links held by page-one sites are editorial. Links earned inside real content, on real websites. Directory submissions and resource pages combined make up less than 8%.
Read that again if you’re still buying directory packages.
Backlinko’s older research found the #1 result in Google carries about 3.8x more backlinks than positions 2 through 10, and Ahrefs has shown the same pattern for years. Correlation isn’t causation, sure. Pages that rank well also attract links because they’re visible. But when every independent study points the same direction for over a decade, the argument is settled for practical purposes. Google itself still lists link spam in its spam policies, and you don’t write rules against manipulating something that doesn’t matter.
What actually changed this year
Google’s March 2026 core update rolled out between March 27 and April 8, and it hit hard. We audited several sites afterward that had lost serious traffic. The pattern was consistent: large volumes of generic content, thin link profiles built from marketplaces, nothing a reader couldn’t find on a hundred other sites.
One blog we reviewed lost over 60% of its organic traffic. Their links weren’t penalized. They were simply ignored, which for your wallet is the same thing.
That’s the real shift in 2026. Google rarely punishes bad links anymore. It devalues them silently. You pay, nothing happens, and you conclude backlinks don’t work. They do. Yours just never counted.
The AI angle nobody saw coming
Here’s the part that surprised even people in my industry.
Everyone assumed AI search would make links irrelevant. The opposite happened. AI Overviews, ChatGPT, Perplexity, all of them lean on authoritative sources when deciding what to cite, and authority on the web is still measured largely through the link graph. One Ahrefs analysis found brand mentions correlate with AI search visibility roughly three times more strongly than backlinks alone. But here’s the catch: those mentions and citations flow to sites that already built authority the old way.
So a strong backlink profile now pays you twice. Once in traditional rankings, and again when AI systems decide whose content is trustworthy enough to quote. I didn’t expect that two years ago. Almost nobody did.
What stopped working (save your money here)
Some things I’d tell any small business owner to stop paying for immediately:
Bulk link packages. Five hundred links for fifty dollars means five hundred links from pages Google has already written off. Almost every cheap batch we audit has the same hidden flaw: the pages holding those links were never even indexed. A link on a page Google hasn’t indexed is a vote that was never cast.
Profile and forum links. Mostly nofollow, mostly ignored.
Exact-match anchor text. If every link pointing at you says “best plumber Dallas,” you’ve handed Google a signed confession.
Sites with inflated authority scores. There’s a whole market of domains with impressive DA numbers and no actual visitors. The score is manufactured. The traffic isn’t there. And traffic is the tell.
That last one trips people up constantly, so we built a habit around it: before accepting any link placement, run the site through a simple three-filter test for judging link quality. Is the site relevant to your topic? Does it get at least a couple thousand real organic visitors a month? Does the link sit inside genuine content, not a footer or author bio? Fail any one, walk away. It takes five minutes and it has saved our clients more money than any other single check we do.
What still works, and what it did for one client
Editorial links from relevant sites with real audiences. That’s it. That’s the whole secret, and it’s boring, which is why nobody wants to hear it.
Earlier this year we ran this exact playbook for a home services company whose money page had been stuck deep on page 4. We disavowed the junk they’d bought, then spent three months earning a small handful of links through guest contributions on trade and local business sites. Real sites, real readers. Under four months later the page was sitting on page one, and the owner’s phone had started ringing.
A handful of earned links beat an entire folder of bought ones. Not because of some trick. Because a few votes from voters Google trusts outweigh hundreds of votes from voters it stopped listening to years ago.
If you don’t have time to do the vetting and outreach yourself, that’s essentially what agencies like LinkHarbor SEO exist for. But whether you hire it out or do it yourself, the standard is the same: fewer links, better sources.
So, should a small business still invest in backlinks?
Yes, with two conditions.
First, fix your content before you buy a single link. Links amplify what’s already there. They can’t rescue a thin page.
Second, adjust your expectations to reality. A quality link takes 6 to 12 weeks to show up in rankings. A campaign takes a quarter or two to change your trajectory. Anyone promising page one in 30 days is lying to you, and in 2026 Google is better than ever at making sure their shortcuts lead nowhere.
Backlinks aren’t dead. Lazy link building is. There’s a real difference, and knowing it is worth more to your business than any package you’ll ever buy.
Read More From Techbullion
