Six Months, Zero Links Bought: A Cold-Turkey Experiment That Exposed Our Spending Addiction
Photo by Photo by Matt Bero on Unsplash on Unsplash
Every agency owner has a number in their head. Mine was $1,400 a month — the average I'd been spending on links per site across my portfolio for the better part of three years. Some months it bought niche edits. Some months it bought guest posts. Some months it bought, honestly, nothing I could explain if you woke me up at 3 a.m.
So in January, I did something that felt genuinely reckless: I stopped. Completely. Three sites, six months, zero link purchases. Here's what the spreadsheet says happened.
The Three Sites and the Rules
The test wasn't scientific — I run real businesses, not lab conditions — but I picked sites with different profiles to make the results useful:
- Site A: A home services affiliate site, roughly 40 referring domains gained over two years, most purchased.
- Site B: A local service business site with a mixed profile — half earned, half bought.
- Site C: A B2B SaaS blog with strong organic links and only light paid activity.
The rules were simple. No new purchases, no trades, no 'favors.' Existing links stay put. Content publishing continues as normal. Rankings tracked weekly for 40 keywords per site.
Months One and Two: The Panic That Wasn't
Nothing happened.
That's the headline for the first 60 days, and I want to underline it because it contradicts the anxiety every link buyer carries. Site A's money keywords didn't budge more than two or three positions. Site B didn't move at all. Site C, predictably, kept climbing slightly on the strength of content momentum.
If you're spending $500 to $1,500 a month on links and your rankings are stable, the uncomfortable question this experiment raises is: what exactly were you maintaining?
Month Three: The First Real Slip
Site A blinked first. Around week 11, two mid-funnel keywords — terms that had sat at positions 4 and 6 for nearly a year — slid to positions 9 and 13. Nothing catastrophic. But the pattern was clear: the keywords with the thinnest content support and the most paid-link dependency were the ones that softened first.
Site B held. Site C held. My working theory: sites with genuine brand signals and earned links have a buffer. Sites running mostly on purchased authority are renting their positions, and the rent comes due the moment you stop paying.
Months Four Through Six: The Divide
The back half of the experiment split the sites into two camps.
Site A lost ground steadily. By month six, it had dropped an average of 4.2 positions across tracked keywords, and traffic fell roughly 30% from its January peak. This was the site where nearly every link had a receipt attached. Google, apparently, doesn't care about your receipts.
Sites B and C finished the six months within a position or two of where they started. B actually gained one keyword into the top three, likely from a competitor's own link freeze. Total spend on links across both sites during the experiment: $0.
The Maintenance Budget Framework
What I took away isn't 'stop buying links.' It's that most of us conflate two very different budgets:
A growth budget buys new authority — new referring domains, new pages, new rankings. It should be tied to specific keyword targets and judged by movement.
A maintenance budget protects what you've built. For most sites, that number is dramatically smaller than what they're currently spending — because most of what they're buying isn't protecting anything. It's just feeding the cadence.
My rule now: if a site's rankings are flat for two consecutive months despite active link spending, that spending isn't maintenance. It's a subscription to a feeling. Cancel it, bank the difference for three months, and see if the floor drops. For two of my three sites, it didn't — and the money is now funding content that actually compounds.
The link hamster wheel is real. The first step is checking whether you're actually on it, or just running in place next to it.
