Buying an Affiliate Site Instead of Starting One: The Math, What to Verify Before Paying, and the Deals to Walk Away From
A new Amazon affiliate site spends its first year earning nearly nothing, and a bought one earns from the day the transfer clears. That's the whole case for buying, and it's a good one when the site is what the listing says.
The case against is that you're buying a stranger's claims about traffic that belongs to Google and earnings that belong to Amazon, and both can move the month after you pay.
This post covers the buy-or-build math in plain numbers, the six things to verify live before any money moves, the deals to walk away from however good the multiple looks, where to buy and what each marketplace's vetting is worth, and what to do in the first ninety days of owning a site someone else built.
It's the buyer's side of the valuation post, which covers the price from the seller's end.
The short answer
Buy when you can verify twelve months of earnings live, the traffic is spread across many pages, and you can see work the seller hasn't done. Pay a multiple of the trailing twelve months' profit, not of the best quarter.
Walk away from a single-page earner, a site that moved on the last core update, content the seller can't prove they own, or any earnings you can only see in a screenshot. Then spend the first ninety days refreshing, not expanding.
The math
| Build | Buy at 30x | |
|---|---|---|
| Cost to a site earning $1,000/month | 12 to 18 months of writing and hosting; little cash, much time | $30,000 up front |
| Earnings in year one | Close to zero for most of it | $12,000 if earnings hold |
| Payback | Whenever the site ranks; no guarantee it does | 30 months at flat earnings; faster if you improve it, never if it drops |
| Main risk | The site never earns | The site stops earning after you've paid |
| Where the edge is | You know every page and every pick | Someone else's year, bought in a day; the upside is the work they didn't do |
The number that decides it is the one in the last row. A site bought at 30x that you can lift by half in a year, by refreshing stale guides and filling the hubs the seller never built, pays back in under two years. A site that's already polished pays back in two and a half if nothing goes wrong, and something usually does.
The six things to verify live
- Associates earnings, on the seller's screen. Twelve months, with tracking IDs visible, on a video call where they drive. The tracking IDs post covers what the split by ID shows; here it shows whether the earnings come from the site's pages at all.
- Search Console, same period. Clicks by page. If the top three pages are most of the traffic, the site is three pages with a domain attached, and the price should say so.
- Clicks against orders. A month where orders rose without clicks is a bounty, a seller's own purchase, or someone else's link using the site's tag.
- The buttons, yourself. Twenty, on the top guides, in a private window. Dead picks mean the earnings are already falling and the report hasn't caught up.
- Content ownership. Who wrote it, on what terms, and whether any of it is lifted. A paragraph pasted into a search engine answers the last question; the first two need the contracts or the seller's word, which is worth what the rest of the call has shown.
- Transferability. The domain, the hosting, the Associates account's pages (the site moves to your account; the earnings history doesn't), the email list if any, and the social accounts. Anything that can't transfer isn't in the price.
Walk away when
- One page earns most of it. A single guide at position one is a single update away from zero, and no multiple prices that risk.
- Traffic moved on a core update and the seller's explanation is "seasonality". Date the Search Console graph against the update timeline; a drop that lines up is a site Google has already judged.
- Earnings only in screenshots. There is no reason for a seller with real reports to refuse a live view, and a few reasons for one without.
- The seller can't say who wrote the content, or the content reads as one AI run with no edits, no verdicts and the same seven picks in every guide. You're buying a rewrite.
- The price is set on the last three months. Sales seasons inflate a quarter; the trailing twelve is the number.
- The niche is one Amazon commission cut away from unprofitable, or a category Amazon has moved to a lower rate before.
Where to buy
Flippa lists the most sites with the least vetting; the floors are low and so is the signal, and every verification above is yours to run. Empire Flippers and Motion Invest vet listings and verify earnings before they go up, which removes the screenshot problem and raises both the floor and the price.
Private deals through niche forums and newsletters are cheapest and least checked. In all three, "verified earnings" means the money arrived; it doesn't tell you which page it came from or whether that page is about to lose its position. That part is the live call.
The first ninety days
- Change nothing structural for a month. No redesign, no URL changes, no theme swap. Google is watching the ownership change; give it a site that looks the same.
- Refresh the top twenty guides: dead picks replaced, prices current, verdicts added where there were none. The updating old posts post covers the method; this is where the bought site's upside is.
- Set your own tracking IDs per topic, so from month two you can see what earns under you.
- Build the hubs the seller skipped, if any, and link the guides into them.
- Only then expand: new guides in the gaps the keyword list shows, at the site's existing pace, not faster.
Doing it with AffSERP
The bought site's upside is almost always the refresh, and that's what the tools do. AffSERP's Refresh option regenerates a stale guide from current Amazon product data on the same URL, which turns the top-twenty refresh from a month of rewriting into a week of reviewing.
The Keyword Explorer and Keyword Grouping tools map what the seller left uncovered for the expansion in month four, and the X Best Review tool writes those guides. The verification is yours; the tools start the day the transfer clears.
FAQ
Is it worth buying an affiliate website?
Sometimes, and only for a site whose earnings you've verified live and whose traffic isn't about to move. At 30x monthly profit a site pays back in two and a half years if earnings hold, which they often don't; the buyer's edge is skipping the twelve to eighteen months a new site spends earning nothing. It's worth it when you can improve the site (stale guides, no hubs, half the picks dead) and the price reflects the current state, not the potential. It isn't when the seller has already done the easy work and the price assumes it continues.
How do I verify an affiliate site's earnings before buying?
Live, on the seller's screen, never from screenshots: the Associates earnings report for the last twelve months with the tracking IDs visible, Search Console for the same period, and analytics if there is any. Match clicks to Amazon against orders, check that the earnings come from the site's own pages (tracking IDs by page or topic make this possible), and look for a month where earnings spiked without traffic, which usually means the seller's own purchases or a bounty. Then click twenty buttons yourself and see what they land on.
Where do you buy affiliate websites?
Marketplaces and brokers: Flippa for smaller sites with light vetting and more noise, Empire Flippers and Motion Invest for vetted listings with verified earnings and higher floors, and private deals through niche forums and newsletters where the price is lower and the checking is entirely yours. The marketplace's vetting is a start, not a substitute for your own; verified earnings tell you the money arrived, not where it came from or whether it stays.
What's a fair price for an affiliate site?
Twenty-five to forty times average monthly net profit over the last twelve months, with the multiple set by age, traffic diversity, earnings trend, how many pages do the earning, and how much work the site needs. A site at 40x has to be growing, diversified and clean; a site at 25x has something wrong that you've found and priced. Pay for the trailing twelve months, not the last three, and never for what the seller says it could do.
A bought site earns on the guides the seller let go stale. AffSERP's Refresh regenerates each one from current Amazon data on the same URL, so the top-twenty refresh is a week of reviewing, not a month of rewriting.
Start with free wordsKeep reading
