You're about to spend fifteen grand. Maybe fifty. The listing shows a traffic graph climbing nicely to the right, the seller seems straightforward, the multiple looks fair, and everything feels fine.
Here's the uncomfortable bit. That traffic graph is the single easiest thing in the entire listing to manipulate, and the problems that will actually hurt you don't show up in it at all. A website can be sitting on a Google penalty, a collapsing link profile, or content that's one algorithm update away from vanishing, and still produce a beautiful chart on the day you look at it.
This is what to check before you buy, roughly in the order that matters. Works the same whether you're on Flippa, Empire Flippers, Acquire, or buying directly from someone's DMs.
The Question the Listing Can't Answer
Every listing tells you what the traffic is. Hardly any tell you why it exists, and that's the only thing that predicts whether it survives once you own it.
Traffic from one post that went viral, or one partnership that ends at closing, or one lucky ranking the site doesn't really deserve, is a completely different asset from traffic spread across two hundred deliberately built pages. Identical number on the chart. Wildly different purchase.
So the lens for everything below is: is this traffic durable, and is it actually what the seller says it is?
Step 1: Verify the Traffic Independently
Never accept screenshots. A screenshot is a photograph of a claim, nothing more, and editing one takes about ninety seconds.
Ask for temporary read-only access to Google Analytics and Google Search Console. Any serious seller grants this during due diligence. Refusal is itself a complete answer.
Once you're in, check these:
- Do Analytics and Search Console agree with each other? They're separate systems. Faking a match across both is a lot of work, and most people who fudge numbers only bother with one.
- Where's the traffic coming from? Look at channels in Analytics. A site sold as an "SEO asset" that turns out to be 70% direct and social traffic is not the thing you were shown.
- Does it look human? Suspiciously flat plateaus, visitors who leave in under two seconds, or a surge from a country the business has nothing to do with all suggest bought traffic. Real audiences are lumpy and seasonal.
- What does 24 months look like, not 3? Listings show the window that flatters. Ask for the longer view and watch whether that changes the story.
Step 2: Check for Penalties
This is the highest-stakes check here, because a penalised site can look completely healthy and be close to worthless.
In Search Console, open Security & Manual Actions. A "manual action" means a human at Google has looked at this site and decided to punish it, usually for spam or dodgy links. It should say "No issues detected". If it doesn't, either walk away or price in a recovery you might not win.
There's a quieter version too. Algorithmic damage comes with no notification at all. Pull up 24 months of Search Console history and look for a sharp drop that lines up with one of Google's core updates and never recovered. That's a content quality problem, and it transfers to you along with the domain name.
Step 3: Look Properly at the Backlinks
Backlinks are just links from other sites pointing at this one. Google reads them as votes of confidence, which is why they're the part of a website you can't fix quickly if they're bad or rebuild cheaply if they're gone.
Two questions:
Are they real? Sample twenty and go and look. You want links inside genuine articles on sites with actual readers. What you don't want: networks of blogs that exist only to sell links, pages that are nothing but lists of outbound links, or three hundred links from the same handful of domains. A site built on bought links is one enforcement action from losing everything, and by then it'll be your site. Our guide to identifying toxic backlinks walks through the patterns.
Are they growing or dying? A link profile that peaked two years ago and has been quietly shedding ever since is a site steadily losing its ranking power. That decline doesn't stop because the owner changed.
Step 4: Check the Content Is Actually Theirs
Take five of the best-performing articles and paste a distinctive sentence from each into Google, in quotation marks.
If that exact text turns up on other websites, you've found one of three things: content scraped from somewhere else, content the owner syndicated without mentioning it, or spun content churned out at scale. All three reduce what you're buying. The first one can wipe it out entirely.
Also worth asking: does the top content depend on something you won't inherit? An article that ranks because of the author's personal reputation, or a partnership that ends at closing, is traffic you're paying for but can't keep hold of.
Step 5: Find Out What Got Cleaned Up Before Listing
Sellers tidy up before selling. Some of that is fair presentation. Some of it buries history.
Use the Wayback Machine (web.archive.org) to look at the site 6, 12 and 24 months back. You're looking for:
- A recent redesign or site move. Migrations are the number one cause of traffic collapse, and the damage usually shows up two to four months later. A redesign shortly before listing means you might be buying numbers that haven't caught up with reality yet.
- Content that disappeared. Pages that existed and now don't can mean a penalty cleanup. Compare the old page count with what
site:domain.comreturns today. - A completely different niche. Domains that used to be about gambling, adult content or pharmaceuticals can carry baggage that follows them for years.
Step 6: Check the Technical Foundations
Technical problems are the good news category, because they're the fixable one. You just want them priced into the deal rather than discovered the week after closing.
What to look at: mobile page speed, whether the important pages are actually in Google's index, redirect chains left behind by old migrations, orphan pages that nothing links to, and duplicate or missing page titles across the catalogue.
Doing that by hand across a few hundred pages is genuinely tedious. This is the one place where a proper automated crawl earns its money, because it looks at every page rather than the five you happened to click on.
Step 7: Judge the Keywords, Not Just the Volume
Look at what the site ranks for and ask whether you can make money from those specific people.
A site with 50,000 monthly visitors on pure curiosity searches ("what is X") is usually worth less than one with 5,000 visitors on buying searches ("best X for Y", "X pricing"). The first is an audience you still have to convert. The second already has a wallet out.
Then check concentration. If one page brings 60% of the traffic, you're not buying a website. You're buying one article, its ranking, and every risk attached to that ranking.
The Fast Version: What to Run Before You Even Make an Offer
Everything above is worth doing on a deal you're serious about. You obviously don't want to spend three hours on every listing you glance at, and buyers who try that end up giving up and trusting the traffic graph, which is the exact outcome this whole process exists to prevent.
So the efficient order is: run a technical and ranking check first to see if the thing is structurally sound, then request Analytics access and do the deep manual work only on the ones that survive that first pass.
That first pass is what a $4.99 TrackSEO report does. You put in the domain, you need nothing from the seller because it works entirely on public data, and you get back the rankings, the backlink profile and its health, the technical issues, and whether AI search tools ever cite the site. If you're screening a handful of listings, the 10-pack works out around $3 each, which against a five-figure purchase is a rounding error.
Red Flags That Should Just End the Conversation
- Won't give read-only Analytics and Search Console access during due diligence
- A manual action sitting in Search Console
- Traffic that fell on a core update date and never came back
- A link profile that's mostly paid or networked links
- Top articles appearing word for word on other sites
- A migration or redesign in the last three months with no data since
- Most of the traffic riding on one page or one keyword
Use What You Find to Negotiate, Not Just to Walk
Most due diligence advice treats every finding as a yes or no on the whole deal. In practice, the useful outcome is usually somewhere in between, and what you find becomes bargaining power.
Technical problems are the clearest example. A site with slow pages, broken redirects and no structured data isn't damaged goods. It's a site with fixable upside, and you now know something the seller either didn't know or hoped you wouldn't check. Both are useful positions to negotiate from. "I'm still interested, but there's about three weeks of technical work here that wasn't in the listing" is a reasonable thing to say and it's backed by a document.
Some findings genuinely should end it. A manual penalty, a link profile built on bought links, or content copied from elsewhere are all things you cannot fix by working harder, and you'd be buying someone else's problem at full price.
And some findings change what you're buying rather than whether you buy. If 60% of the traffic comes from one article, the site might still be worth owning, but it's a different asset with a different risk profile and it should be priced like one. The mistake is paying portfolio prices for a single lottery ticket.
The First 30 Days After You Buy
Worth planning before you close, because the handover window is short and things go missing.
Get your own tracking in place on day one. Your own Analytics, your own Search Console. Don't rely on inheriting the seller's accounts, because access can vanish and you want a clean baseline that starts the day you take over.
Change nothing structural for a month. The temptation to redesign immediately is strong and it's how new owners destroy the thing they just bought. URLs, page titles and site structure are load-bearing. Leave them alone until you understand which pages are actually earning.
Do fix the technical problems. Speed, broken links, missing descriptions. These are safe improvements that don't touch what's already ranking.
Re-run your audit at day 30. Compare it against the one you ran before buying. If rankings or links have moved sharply in either direction, you want to know early, while you can still do something about it.
Can You Check a Website's SEO Without the Seller's Permission?
Mostly, yes. Rankings, backlinks, technical health, page speed, indexed pages and content originality are all visible from the outside using public data. You can screen a listing thoroughly before you ever message the seller.
What you can't see without access is their own Analytics and Search Console: the real traffic numbers, the actual search queries, and crucially that manual actions panel. Which is exactly why the order matters. Public checks first to decide what deserves your time, then ask for access on the ones worth pursuing.
How Long Does SEO Due Diligence Take?
A public-data screen takes a few minutes per site with a tool, or an hour or so by hand. The full version with seller access, reading two years of Search Console history and sampling the backlinks properly, is a half day of real work.
Spend that half day only on sites you're genuinely considering. It's not a process designed for browsing.
What's the Biggest SEO Risk When Buying a Website?
Traffic concentration, and almost nobody weights it properly. A site earning 60% of its visits from one page isn't a portfolio, it's a single ranking with a website attached to it. Google can adjust that ranking any time it likes, and if it does, the business you bought is gone while the invoice very much isn't.
Second place goes to recent migrations, for the reason above: the damage lands months after the move, so the numbers you're looking at may simply not have caught up yet.
What Should SEO Due Diligence Cost?
Consultants charge from a few hundred dollars up to several thousand for a full pre-acquisition review, and on a large acquisition that's money well spent. For screening listings, or for a deal in the low five figures, that maths doesn't work.
The practical answer: run the automated checks yourself for a few dollars a site, spend your own time on the judgement calls software can't make (is this content actually good, would a real person link to it, does this niche have a future), and bring in a paid expert only for the deal you've already decided you want.
What you shouldn't do is skip it. The traffic graph in a listing is a claim, not evidence, and the gap between those two things is where buyers lose money.
Screen a site for $4.99 before you wire anything. Few minutes, works on any domain, and the seller doesn't need to know or agree.