For a content site, the keyword rankings are the business. Everything else — the WordPress install, the logo, the email list — is packaging. Here's the exact keyword audit I run before I let anyone wire money for a content-driven acquisition.
Deal Alert AI is reader-supported. We earn commissions from affiliate links at no cost to you.
This post is based on a video from our Deal Alert AI YouTube channel. Watch the original or read the full breakdown below.
I've watched buyers spend three weeks arguing over a $4,000 difference in the multiple, then skip the keyword audit entirely. Six months later the site is down 60% in traffic and they're posting in a Facebook group asking whether they got scammed. They didn't get scammed. They bought a foundation problem they never inspected.
Content sites are unusual assets. When you buy a SaaS product, you buy code and customers. When you buy an ecommerce store, you buy inventory, suppliers, and a brand. When you buy a content site, you are buying a set of rankings inside a search engine owned by a company that owes you nothing and can change its mind on a Tuesday. That's not a reason to avoid content sites — I still buy them — but it is a reason to do the work.
This guide is the keyword research and SEO due diligence framework I use before making an offer on any content-driven business. It works whether you're looking at a $30,000 starter site on Flippa or a $900,000 affiliate portfolio on Empire Flippers. The scale changes. The questions don't.
Brokers price content sites on a multiple of trailing twelve-month profit. That's a convenient shorthand, but it's backward-looking. A 38x monthly multiple on a site earning $5,000/month gives you a $190,000 price tag. Fine. But that number assumes the $5,000 keeps arriving. The only thing that determines whether it keeps arriving is the durability of the keyword rankings underneath it.
Think about what actually produces revenue on an affiliate content site. A visitor types "best noise cancelling headphones under $200" into Google. Your page ranks third. They click, read, click an affiliate link, buy. You earn $8. That single transaction depends entirely on one thing: your page holding position three for that query. Lose the ranking and the $8 vanishes. Not gradually. Immediately.
So when I evaluate a content site, I'm not really valuing the business. I'm valuing a portfolio of ranking positions and estimating how likely each one is to survive the next 24 months. A site with 400 keywords in the top ten spread across three related topical clusters, backed by 90 referring domains from real publications, is a fundamentally different asset than a site with 12 keywords in the top ten that all sit in one category and got there on the back of a link-building campaign that ended in 2022. Both might be earning $5,000/month today. Only one is worth 38x.
Key insight: A content site's multiple should be a function of ranking durability, not just profit. Two sites earning identical revenue can deserve multiples 15x apart depending on keyword concentration, backlink quality, and content cadence. Most sellers price on revenue alone. That gap is where buyer edge lives.
We scan Empire Flippers, Flippa, Acquire.com and Quiet Light daily — scoring every listing. Start free.
Start by pulling the full organic keyword footprint. Drop the target domain into Ahrefs Site Explorer or Semrush and export the complete organic keywords report to a spreadsheet. If the site is small enough that third-party tools show incomplete data, ask the seller for read-only Google Search Console access. Any legitimate seller will grant it during due diligence. If they refuse, you've learned something important for free.
Once the export is in a spreadsheet, filter to positions 1 through 10. In most niches, the first page captures roughly 90% of clicks for a given query, and positions 1 through 3 alone typically pull somewhere between 50% and 65%. Keywords ranking at position 24 are not assets. They're lottery tickets. Include them in your notes, exclude them from your valuation.
Now sort by estimated traffic and look at the shape of the distribution. What you want to see is a long, gradual curve — hundreds of keywords each contributing a modest slice. What you don't want is a cliff: three keywords doing enormous volume and then a sharp drop to nothing. I'll cover concentration risk in detail below, but you can usually feel the problem the moment you look at the sorted list.
One practical note: reconcile the tool's traffic estimate against actual Google Analytics data from the seller. Ahrefs and Semrush estimates are directionally useful but often off by 30-50% in either direction. If the tool says 40,000 monthly organic sessions and Analytics says 11,000, that discrepancy needs an explanation before you go further. Sometimes it's benign — a lot of low-CTR informational rankings. Sometimes it means the site's high-value rankings dropped recently and the tool hasn't fully updated.
Not all traffic is equal, and on affiliate sites the gap is enormous. I've audited sites where 80% of traffic came from informational queries that monetized at roughly $2 RPM, while 20% of traffic came from commercial queries monetizing at $60 RPM. Guess which 20% actually mattered.
Filter your keyword export for commercial and transactional intent modifiers: "buy," "best," "review," "vs," "alternative," "coupon," "deal," "cheapest," "for [use case]," and category-plus-product patterns. These are the keywords where a reader is holding a credit card. On a display-ad site, the equivalent filter is different — you're looking for high-CPC categories like finance, insurance, legal, and home services — but the principle holds.
Once you've isolated the money keywords, calculate what percentage of total organic traffic they represent and — more importantly — what percentage of revenue. Ask the seller for affiliate dashboard exports broken out by page or subID if they track it. Match those top-earning pages back to the keywords that drive them. Now you know the actual revenue engine of the business, usually 15 to 40 pages out of a site that might have 600.
This is the moment where a lot of deals get reframed. I looked at a home-improvement site last year listed at $310,000 with 220,000 monthly sessions. Impressive number. But when I mapped revenue to keywords, 71% of the affiliate income came from nine review pages targeting a single product category. The other 590 pages were essentially a very expensive traffic decoration. That's not a $310,000 content portfolio. That's nine pages with a website attached.
Here's my hard threshold: if the top ten keywords drive more than 60% of total organic traffic, the site has dangerous concentration risk and I either walk or demand a significantly lower multiple. If the top three keywords drive more than 40%, that's near-disqualifying unless there's a compelling structural reason.
The math on why is simple. A single Google core update, a competitor publishing a better resource, or a brand deciding to build its own comparison page can take out one keyword. If that keyword is 25% of your traffic and 35% of your revenue, you've just lost a third of the business in a week with no recourse. Diversification isn't an abstract portfolio-theory concept here — it's the difference between a bad month and a dead asset.
Run the same concentration analysis on three dimensions. First, keyword concentration: what share of traffic comes from your top 10, 25, and 50 keywords. Second, page concentration: what share of traffic and revenue comes from your top 5 and top 20 URLs. Third, topical concentration: are the money keywords spread across multiple product categories or clustered in one? A site earning from headphones, laptops, and monitors is more resilient than a site earning entirely from headphones, even at identical revenue.
Warning: Sellers sometimes mask concentration by presenting traffic at the site level and revenue at the site level, never connecting the two. Always request revenue broken out by page or by affiliate link. If a seller says they "don't track that," they either aren't running the business seriously or they don't want you to see the answer. Both are reasons to price the deal lower.
A ranking can exist for two very different reasons. Either the site has genuine topical authority — deep, comprehensive content in a niche, earning links naturally over years — or the ranking has been purchased through link building that may or may not be sustainable and may or may not be within Google's guidelines.
To distinguish between them, pull the backlink profile and analyze it at the page level, not just the domain level. Take your top ten money pages and check referring domains for each. What you want to see is a natural mix: some editorial mentions, some resource-page links, some forum and social citations, varied anchor text, links acquired steadily over time. What concerns me is 40 referring domains to a single money page, all acquired within a three-month window, all with exact-match commercial anchor text, all from sites that exist to sell links.
Check the referring domains themselves. Are they real publications with their own organic traffic, or are they thin sites with no audience that link out to 200 unrelated commercial pages? Ahrefs' Domain Rating is a crude proxy — I'd rather see 30 links from sites with genuine traffic than 200 links from DR 40 private blog networks. If a large share of the profile looks purchased, assume those links carry risk and that the rankings they support could evaporate in a spam update.
Also ask the direct question: "What link building have you done in the last 24 months, through what channels, and at what cost?" Honest sellers will tell you. Some will say they spent $2,000/month on guest posts, which is useful information because it means your cost structure includes $24,000/year you may not have budgeted. Others will claim they've never built a link while the backlink profile clearly shows otherwise. That's the answer you actually needed.
Content sites decay. Rankings that were earned in 2021 with a 1,800-word article get overtaken in 2024 by a competitor publishing 3,000 words with original testing photos. If the seller stopped publishing 14 months ago and traffic is still flat, that doesn't mean the site is durable — it usually means the decay hasn't shown up in the data yet. Traffic decay from a publishing stop typically becomes visible somewhere between 6 and 12 months out.
Pull the publication history for the last 24 months. You can do this from the WordPress admin, from an XML sitemap with lastmod dates, or by crawling the site. Chart posts per month. What you're looking for is whether the seller wound down operations before listing — a very common pattern. They stop paying writers six months before the sale to inflate profit, list the site at a higher multiple because margins look great, and hand the buyer a content pipeline that's already dry.
Separate new publications from content updates. A site publishing four new articles a month is investing in growth. A site republishing old posts with new dates and minor edits is maintaining. A site doing neither is harvesting. Each of those deserves a different multiple, and the difference is worth tens of thousands of dollars on a mid-six-figure deal.
Then estimate what it costs to keep the cadence going. If the site needs eight articles a month at $180 each plus $400 for editing, that's $1,840/month of real cost. If the seller's P&L shows $200/month in content spend, you're looking at a profit figure that isn't reproducible. This is one of the most common ways stated SDE overstates actual owner earnings on content sites, and it's exactly the kind of thing we flag in deal breakdowns at Deal Alert AI.
Work through this list in order before you submit an offer. It takes about three to five hours for a typical site, which is a rounding error against a six-figure purchase. I've killed deals on item four more than any other.
Over time, the failure patterns become recognizable. The most common one is single-category dependence: every money keyword lives inside one narrow product vertical. When Amazon cuts the commission rate for that category — as they did in 2020, taking home improvement from 8% to 3% overnight — the entire business drops by more than half in a single day with zero warning.
The second pattern is mismatched traffic. The site shows 300,000 monthly sessions, but 60% comes from India, Pakistan, and the Philippines while the monetization is US-only affiliate offers. The traffic number looks great in the listing. The revenue per session tells the real story. Always pull the geographic breakdown and calculate revenue per US session specifically.
The third is unexplained ranking drops. Traffic peaked at 90,000 sessions 14 months ago, sits at 52,000 today, and the seller's explanation is "seasonality" — except the drop happens to line up perfectly with a core update date. Sellers who can't or won't explain a decline usually know exactly what caused it. Ask directly, then verify against update timelines yourself.
The fourth is the link profile that tells on itself: hundreds of referring domains, almost all DR 30–50, almost all with the same handful of anchor texts, almost all acquired in tight clusters. That's a purchased profile. It might hold for years. It might get neutralized in the next spam update. Either way, you should be paying a lower multiple than a site with organically earned links, and you should be budgeting to keep buying links if you want to keep the rankings.
Key insight: The strongest signal of a durable content site isn't traffic volume — it's the ratio of branded search to non-branded search. If people are typing the site's name into Google, you're buying an actual brand with a moat. If branded search is under 2% of total, you're renting rankings from an algorithm.
The obvious problem with a twelve-step audit is that you can't run it on every listing. There are hundreds of content sites live at any moment across Empire Flippers, Flippa, and the smaller brokerages, and running a full keyword audit on each would consume your entire month.
So triage first. Before you touch the spreadsheet, do a five-minute screen: check the traffic trendline shape, check whether the niche is one where big brands are moving in, check the age of the domain, and check whether the listing discloses traffic sources at all. Any listing that fails that screen doesn't earn your five hours. That single filter cuts my funnel by roughly 80% before real analysis begins.
This is the exact problem I built Deal Alert AI to solve. We aggregate listings across the major marketplaces and score them on traffic source quality, concentration signals, and the SEO patterns that historically precede a collapse — so you spend your due diligence hours on the 20% of deals that could actually work, instead of discovering on hour four that a site's entire revenue rides on two keywords. You still do the deep audit yourself. We just make sure you're doing it on the right deals.
One last thing. The point of this framework isn't to find a perfect site — perfect sites don't get listed, or they get listed at multiples that eliminate your return. The point is to price risk accurately. A site with 55% top-ten concentration and a purchased link profile might still be a good buy at 28x. It's a terrible buy at 42x. Your job as a buyer isn't to avoid every flaw. It's to find them all before the seller does, and to make sure the price reflects what you found. Start with the keyword export at Deal Alert AI, and don't wire a dollar until you know exactly which queries are paying for the business.
We scan Empire Flippers, Acquire, Flippa, and Quiet Light daily. The best sub-$500K businesses are gone within 48 hours.