A $99 Ahrefs subscription can save you a $300,000 mistake. Most buyers open Site Explorer, glance at the traffic graph, and call it diligence. Here is the actual workflow — five reports, in order, with the exact thresholds that should make you walk away or renegotiate.
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 have looked at thousands of online business listings. The single biggest difference between buyers who make money and buyers who buy someone else's problem is not capital, patience, or deal flow. It is whether they know how to read a traffic profile.
Most first-time buyers treat SEO diligence as a vibe check. They open the listing, see "organic traffic: 85,000 monthly visitors," see a screenshot from Google Analytics, and move on to the P&L. That is exactly how people end up owning a site that lost 60% of its traffic in the March 2024 core update and is still bleeding.
Ahrefs is the closest thing buyers have to an X-ray machine. Used properly, a 15-minute audit surfaces risks that would take you weeks to find manually — and in many cases risks the seller genuinely does not know exist. Below is the exact workflow I run before any deal gets serious attention.
Let's do the math that matters. A standard Ahrefs plan runs about $99 per month. A mid-market content site on Empire Flippers trading at a 40x monthly multiple with $7,500 in monthly profit will sell for roughly $300,000. If your Ahrefs audit catches one structural problem — a fake backlink profile, a traffic peak that has already rolled over, a single keyword carrying half the revenue — you have earned back the subscription cost roughly 3,000 times over.
I am not exaggerating for effect. In deals I have reviewed, Ahrefs data has produced price reductions of $20,000 to $85,000 on six-figure acquisitions, purely because the buyer could point to a chart and say "this traffic line broke in September 2023 and never recovered — here's the report." Sellers argue with opinions. They rarely argue with a dated traffic graph overlaid against a known algorithm update.
If you are not ready to commit to a paid plan, start with Ahrefs Webmaster Tools, which is free but only works on domains you control — useful post-acquisition, not much help pre-LOI. A better move is to buy one month of a standard plan while you are actively evaluating three to five deals, run every audit in that window, and cancel if you go quiet. Treat it as a per-deal expense, not a subscription.
We scan Empire Flippers, Flippa, Acquire.com and Quiet Light daily — scoring every listing. Start free.
Start by dropping the target domain into Site Explorer. You are looking at four numbers on the overview screen: Domain Rating (DR), estimated monthly organic traffic, total ranking keywords, and referring domains. Write all four down in a spreadsheet next to the listing's claimed figures. This takes 90 seconds and it is the fastest filter you will ever run.
The critical comparison is Ahrefs' traffic estimate versus the seller's claimed traffic. Ahrefs is an estimate, not gospel — it undercounts branded traffic, direct traffic, email, social, and anything behind a login. A site with 40,000 real monthly sessions might show 25,000 in Ahrefs and that is completely normal. But if a seller claims 90,000 organic sessions and Ahrefs estimates 18,000, that is a gap you need explained in writing. My threshold: anything over a 30% discrepancy in the organic channel specifically requires a written explanation and a screen-shared Search Console session.
The other overview signal people skip is the ratio of referring domains to ranking keywords. A site with 12,000 ranking keywords and 90 referring domains is either an extremely efficient content operation or the beneficiary of a topical authority window that may close. A site with 4,000 keywords and 2,800 referring domains has almost certainly been aggressively link-built, and you should assume a portion of those links were paid for and will decay or be devalued. Neither is automatically disqualifying. Both change what you should pay.
Now go to the Organic Keywords report and export the top 500 to 1,000 rows. Filter for positions 1 through 10 — those are the keywords actually producing traffic. Everything ranking on page three is a vanity metric, not an asset.
Here is the calculation that matters: what percentage of total estimated organic traffic comes from the top 10 keywords? Sum the traffic column for those ten rows and divide by the site total. If the answer is above 60%, you are not buying a website — you are buying ten rankings. One core update, one competitor with a better page, one SERP layout change that pushes organic below an AI Overview, and the business is materially impaired.
My rough bands after looking at a lot of these: under 25% concentration is healthy and diversified; 25% to 45% is normal for a niche site and manageable; 45% to 60% is elevated risk that should compress the multiple; above 60% means you either walk or you underwrite the deal assuming those keywords disappear. I also check whether the top traffic keywords are commercial or informational. A site where the top ten keywords are all "best X" affiliate terms is more monetizable but more contested. A site where the top ten are informational how-to terms is more stable but harder to monetize at higher RPMs.
The Backlinks and Referring Domains reports are where sellers get caught. Nearly every site above DR 40 has had some link building done. The question is not whether links were built, but whether they were built in a way that survives the next Google spam update.
Sort referring domains by "first seen" date and look at the acquisition curve. Organic link growth looks like a slow, lumpy staircase. Manufactured link growth looks like a wall — 400 new referring domains appearing in a six-week window two years ago, then nothing. That pattern almost always means a link package was bought. Then check the anchor text distribution report. Natural profiles are dominated by brand names, bare URLs, and junk anchors like "click here" or "this article." If 35% of anchors are exact-match commercial phrases like "best ergonomic office chair," someone was buying placements.
Three more red flags I look for every time. First, foreign-language sites linking to English content — a Ukrainian recipe blog linking to a US personal finance article is not an editorial link. Second, clusters of referring domains that all share hosting footprints or near-identical site structures, which signals a private blog network. Third, links from pages with zero organic traffic themselves; Ahrefs shows referring page traffic, and a link from a page with no traffic transfers very little real value and carries real spam risk. Document what you find with screenshots, because this becomes your negotiating ammunition later.
Diligence is not just about finding reasons to walk. It is about knowing what you can realistically do with the asset in year one, because that determines what you can afford to pay and still hit your return target.
Run the Content Gap report with two to three genuine competitors — not category giants like Healthline or NerdWallet, but sites of comparable DR and size. The output is a list of keywords those competitors rank for and your target does not. Filter for keyword difficulty below the target's demonstrated capability (if the site already ranks top 10 for KD 30 terms, filter to KD 35 and below) and search volume above 200 per month. What is left is your realistic content roadmap.
Now put a number on it. If the gap report surfaces 180 viable keywords with 95,000 combined monthly search volume, and the site converts organic traffic at a $22 RPM, capturing even 20% of that volume at a 30% average CTR adds roughly 5,700 monthly sessions and about $125 in monthly revenue per thousand sessions — call it $600 to $800 per month in incremental profit within 9 to 12 months, against a content cost of maybe $18,000. That is a defensible expansion thesis, and it is exactly the kind of underwriting we build into every listing score at Deal Alert AI. Vague upside is worthless. Quantified upside justifies paying a fair price for a good asset.
This is the most important report in the entire workflow, and it takes 60 seconds. In Site Explorer, open the Organic Traffic graph and extend the date range to "all time." You are looking for the shape of the line, not the current number.
Sharp vertical drops matter more than gradual declines. Cross-reference every drop date against known Google update windows: the September 2023 Helpful Content Update, the March 2024 core update and spam update, the August 2024 core update, and the ongoing rollout of AI Overviews across commercial queries through 2024 and 2025. If a site lost 55% of traffic in a two-week window that lines up exactly with a core update and has been flat ever since, you are looking at a structurally damaged asset. Sites hit by helpful-content-style demotions have historically taken 12 to 24 months to recover, if they recover at all.
Also examine the shape before a spike. If traffic tripled in three months with no corresponding content or link investment, that is often a competitor being penalized rather than the target improving — and those gains reverse. My rule: I underwrite deals on the trailing 12-month traffic trend line, not the trailing 3 months. Sellers list at the top of a spike. That is rational behavior on their part. Your job is to price the mean, not the peak. Marketplaces like Flippa have plenty of listings where the last 90 days look spectacular and the 24-month chart tells a very different story.
Run this in order, on every deal, before you send an LOI. It takes about 20 minutes once you have done it a few times. I keep a spreadsheet with these twelve fields and fill it in for every listing that clears my initial financial screen.
The point of a fixed checklist is that it removes emotion. When you like a business, you unconsciously stop looking for problems. A checklist does not care whether you like the business.
Documentation converts into dollars. A buyer who says "I'm a little worried about the backlinks" gets nothing. A buyer who says "412 of your 890 referring domains were acquired in a seven-week window in Q2 2023, 38% of anchors are exact-match commercial phrases, and 61 of those domains have zero organic traffic — here is the export" gets a price concession or a structured deal.
Match the remedy to the finding. Algorithm damage with no recovery justifies a multiple reduction, typically 4x to 10x depending on severity and how much of the traffic base is affected. Backlink manipulation justifies an earnout or holdback, because the risk is future-dated — structure it so 20% to 30% of the purchase price is contingent on traffic holding for six to twelve months post-close. Keyword concentration justifies a lower multiple plus a longer seller transition period, since you will need help understanding exactly what is holding those rankings.
Be fair about it. If your audit comes back clean — diversified traffic, natural link growth, survived multiple core updates, real content gap upside — say so and pay a fair price quickly. Good assets attract multiple buyers, and sellers on Empire Flippers and other curated marketplaces talk to each other. A reputation for rigorous but reasonable diligence gets you first look at off-market deals. A reputation for using diligence purely as a re-trade weapon gets you excluded.
Ahrefs is an estimate engine, not a source of truth. It does not see direct traffic, email lists, paid social, push notifications, app installs, or revenue. I have seen buyers pass on excellent businesses because Ahrefs showed modest organic numbers, when 70% of the revenue came from a 90,000-person email list with a 34% open rate. That is an asset Ahrefs is structurally blind to.
It also cannot tell you about content quality, editorial process, or whether the site is built on 400 AI-generated articles that have not yet been caught. It cannot tell you about affiliate program stability, whether the Amazon commission rate in that category is about to be cut, or whether the single supplier behind an ecommerce store is about to raise prices 18%. It cannot tell you whether the owner's personal brand is doing the heavy lifting and walks out the door at closing.
So use Ahrefs as your first filter, not your only one. Layer it with Search Console access, Google Analytics with verified property ownership, revenue platform logins viewed live on a screen share, and a genuine read of ten random articles on the site. The tool tells you where to point the flashlight. You still have to look.
The honest problem with this workflow is time. Twenty minutes per deal is fine when you are evaluating five listings. It is impossible when 300 new businesses hit the major marketplaces every week and maybe eight of them are worth serious attention.
That is the specific problem Deal Alert AI was built to solve. We ingest listings across the major marketplaces, pull Ahrefs-grade traffic and backlink signals automatically, and score every listing against the same criteria in the checklist above — traffic trend versus algorithm dates, keyword concentration, referring domain growth patterns, and content gap upside. Instead of running twelve manual checks on 300 listings, you open your inbox to the ten that already cleared them.
Nothing in this replaces your own judgment. The score tells you which deals deserve your twenty minutes, and which deals deserve none. You still run the reports yourself on anything you plan to bid on, you still get Search Console access, you still read the content. But you stop wasting evenings on listings that were structurally broken from the first chart. If you want the filtering done for you across Flippa, Empire Flippers, and the other major marketplaces, that is what Deal Alert AI does every single day — so the only deals you evaluate are the ones that already survived the screen.
By Sophal Lanh, Founder of Deal Alert AI
We scan Empire Flippers, Acquire, Flippa, and Quiet Light daily. The best sub-$500K businesses are gone within 48 hours.