Due Diligence 8 min read

Stop Guessing: How to Use Cloudflare Analytics to Vett Online Business Traffic

Sellers can easily fake Google Analytics numbers, but they rarely clone IP reputation data. Learn the technical steps to verify if the traffic driving your potential acquisition is legitimate or bot-generated.

2026-08-28  ·  By Sophal Lanh, Founder of Deal Alert AI

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.

The Problem With Trusting Seller-Provided Analytics

Buying an online business is one of the most lucrative side hustles or full-time careers you can undertake today. The asset class is booming, with platforms like Flippa listing hundreds of thousands of sites ranging from niche blogs to high-volume e-commerce stores. However, the primary risk for buyers is not finding a deal; it is buying a deal that does not actually exist. The most common metric that gets falsified is traffic. Sellers have every incentive to inflate their page views, sessions, and unique users to justify a higher multiple of EBITDA. Traditional web analytics tools like Google Analytics have become so sophisticated and easily manipulated that their data should be treated with extreme skepticism during the due diligence phase.

When you receive a data room or access to a seller’s dashboard, the numbers often look impressive. You see month-over-month growth, stable seasonal trends, and high engagement rates. On the surface, this looks like a healthy, growing digital asset. But if you have purchased a business before, you know that "looks good" is not the same as "is good." A sophisticated fraudster or even an inexperienced seller can use bots, self-referral tricks, or third-party traffic exchange services to pump the numbers on standard analytics platforms. If you pay $100,000 based on 50,000 monthly visitors, but only 10,000 of those are human beings with disposable income, your business model collapses the moment the bot-farming service stops.

This is where independent verification becomes critical. You cannot buy what you cannot verify. In the world of digital assets, the IP address is the closest thing to a fingerprint for a visitor. While you cannot identify the specific person behind every IP, you can determine the technical characteristics of the traffic source. Is it coming from residential IPs? Is it concentrated in a specific geographically implausible region? Is it hitting the server at impossible speeds? This is where Deal Alert AI and its integration with technical metrics become invaluable. We help buyers understand that traffic quality is a technical audit, not just a financial review. You need to look beneath the surface of the dashboard to see the raw data of who is actually visiting the site.

Why Cloudflare is the Gold Standard for Verification

Get Free Deal Alerts Every Morning

We scan Empire Flippers, Flippa, Acquire.com and Quiet Light daily — scoring every listing. Start free.

Cloudflare is one of the largest web infrastructure providers in the world, serving over a quarter of the entire web. They act as a reverse proxy and a CDN (Content Delivery Network). This means that when someone visits a website hosted on Cloudflare, the actual request does not go directly to the origin server; it hits Cloudflare’s edge server first. This architectural detail is what makes Cloudflare data so powerful for due diligence. Because Cloudflare sits in the middle, it sees every single packet of data attempting to enter the website. It logs the IP address, the geolocation, the user agent string, and the timing of the request with a granularity that standard scripts like Google Analytics often miss or interpret incorrectly.

Unlike Google Analytics, which relies on JavaScript running in the browser, Cloudflare logs data at the network layer. This is a crucial distinction. If a seller uses a bot to visit their site, the bot might execute the JavaScript and send a ping to Google Analytics, making it look like a real user. However, the bot’s IP address is often labeled as "datacenter" or "hosting" by IP intelligence databases. Cloudflare aggregates this data and provides insights into the reputation of the traffic. By analyzing the distribution of IP types, you can instantly spot anomalies. Human traffic comes from residential ISPs (Internet Service Providers). Bot traffic often comes from datacenters, cloud providers, or shared proxy networks. If 40% of your traffic is coming from AWS datacenter IPs for a consumer-facing blog, you have a serious problem.

Furthermore, Cloudflare provides data on engagement speeds that are difficult to fake. Real humans take time to load pages, scroll, and click. Bots tend to request static assets in rapid succession or fail to complete the entire page load sequence in a way that mimics human behavior. While advanced bots exist, the sheer volume of low-level bot traffic that inflates casual analytics dashboards is easily filtered out by Cloudflare’s firewalls and analytics suite. Using Cloudflare data allows you to move from "trust but verify" to "verify, then trust." It removes the ambiguity of whether the traffic is real, which is a key component of the valuation model we discuss on Deal Alert AI for any serious acquisition.

Setting Up Access During Due Diligence

Before you can analyze the data, you need access. This is a negotiation point that should be addressed early in the term sheet. Most professional intermediaries and sellers understand that a serious buyer will want third-party verification of traffic. You should request either read-only access to the Cloudflare dashboard for the specific domain or a comprehensive export of the Cloudflare analytics logs for the last six to twelve months. If the seller refuses this access or claims the site is not hosted on Cloudflare, treat it as a red flag. You can ask them to provide DNS records to prove where the site is hosted, or you can suggest a short trial period of migrating the site to Cloudflare under your supervision if the domain transfer hasn't happened yet, though this is risky and should be handled with caution.

If the site is already on Cloudflare, the seller should add your email as a user with read-only permissions. This is the safest and most professional approach. It allows you to see the real-time data without giving you the ability to change firewall rules or delete the account. When you log in, look for the "Analytics" or "Under Attack" modes. If a site is frequently in "Under Attack" mode, it indicates heavy bot pressure, which could mean the traffic is being actively generated by bots trying to bypass security, or that the site is a target for DDoS attacks. Both scenarios require deeper investigation into the quality of the remaining "clean" traffic.

You must also check the subdomains. A seller might show you clean analytics for the main domain, but if the ad network or affiliate links are hosted on a subdomain that is full of bot traffic, the revenue might still be inflated. Ensure you have visibility into all subdomains associated with the business operation. This level of thoroughness is what separates a casual buyer from a professional investor. We often see buyers on Empire Flippers and other marketplaces who jump straight to the financials without checking the technical foundation. Do not make that mistake. The technical foundation supports the financials. If the foundation is cracked, the numbers will eventually show it.

Analyzing Geographic Traffic Anomalies

One of the most telling signs of fake traffic is geographic data that makes no sense for the business model. If an English-language blog targeting the US and UK markets shows 30% of its traffic coming from Indonesia or Brazil, there is a problem. It might not be malicious; sometimes, ad networks run campaigns that overspend in cheaper geos to meet KPIs. However, if the monetization model is display ads with high CPMs (Cost Per Mille) for US traffic, paying for Indonesian users who fall below the ad network’s quality thresholds can result in rejected earnings or lower payouts. The discrepancy between reported traffic and actual monetizable traffic is where the value lie hides.

Look at the "Top Countries" report in Cloudflare. Compare this directly with the "Top Countries" in the seller’s Google Analytics or AdSense reports. If Google Analytics says 90% of traffic is US, but Cloudflare says 90% is from three Asian countries, one of them is lying or, more likely, the Google Analytics data is being skewed by self-referrals or ad fraud. Cloudflare’s geolocation is based on the IP address, which is much harder to spoof than a browser’s locale settings. A bot can set its browser language to "en-US," but it still connects from its actual physical location, which Cloudflare sees right away.

Additionally, pay attention to the "Top Regions" within countries. For a local service business, you expect traffic from the specific metro area. If a local SEO business claims to serve Miami, but Cloudflare shows traffic spread evenly across the entire globe, it is likely purchased traffic or bot traffic. The distribution should be organic and clustered. Use this data to cross-reference with the buyer’s expectations. If the traffic is global but the content is highly localized, the conversion rates will be abnormally low. Low conversion rates combined with high traffic volume are a classic sign that the traffic is not qualified. This analysis should be part of your standard due diligence checklist, as detailed in our guides on Deal Alert AI.

Checking for Bot Activity and Security Events

Cloudflare has a robust firewall that classifies traffic as "Bot Managed" or "Challenges." Before you buy, look at the volume of traffic that was challenged or blocked. If a significant portion of the traffic is being blocked because it looks like a bot, the "clean" traffic numbers the seller provides might be an overestimate of real, human interest. If the seller had "Bot Fight On" mode enabled, many of those hits would have been dropped before they ever reached the analytics counter. If the seller was running the site with "Under Attack" mode rarely, they might have allowed a lot of garbage traffic in.

Check the "Firewall Events" or "Security" logs. Look for spikes in HTTP 403 or 429 status codes. These indicate that users or bots were being rejected. If you see a consistent pattern of rejections during the same hours every day, it might indicate a scheduled bot script. If you see a sudden massive spike in blocked traffic, it could be a DDoS attack, but it could also be a competitor trying to manipulate the perception of the site’s stability. In either case, the integrity of the visitor experience is compromised. A stable business should have a relatively low and consistent rate of security events. High volatility in security events suggests an unstable technical environment.

Furthermore, examine the "Real User Agent" distribution. While it is harder for beginners to manipulate user agents, sophisticated fraud networks rotate them. However, you can still spot anomalies. If 60% of traffic is coming from outdated browsers or specific versions of Googlebot that are known to be used by spammers, that is a warning sign. Legitimate user agent strings should be a diverse mix of modern browsers. A homogeneous look suggests automation. This level of technical scrutiny is not required for every small blog, but for any business with meaningful revenue, it is non-negotiable. The goal is to find the needle in the haystack of data to ensure you are not buying a ghost town.

Correlating Traffic With Revenue Data

The ultimate test of traffic quality is not just whether it is human, but whether it is profitable. You must correlate the Cloudflare traffic data with the revenue reports from ad networks, affiliate programs, or e-commerce platforms. If Cloudflare shows a steady 100,000 page views per month, but the ad network revenue only reflects 20,000 fillable impressions, the remaining 80% of traffic is likely bot traffic that ad networks have identified and filtered out. Ad networks like Google AdSense, Ezoic, or Mediavine have their own fraud detection systems. They will pay you for quality traffic, not total traffic. If the seller is showing you total traffic to calculate a multiple, but the revenue is based on net traffic, you are overpaying.

Look for the "Fill Rate" and "ECPM" (Effective Cost Per Mille) in the ad network dashboard and compare it to the traffic levels in Cloudflare. If ECPM drops significantly when traffic spikes, it often means the additional traffic is low-quality. Human traffic usually drives up ECPM because it leads to higher click-through rates and better ad matches. Bot traffic dilutes this. You can also look at the "Click-Through Rate" (CTR). If CTR is extremely low despite high impressions, it suggests users are not engaging with the content. This could be due to poor content, but it can also be a sign of passive or bot-driven browsing. The correlation between these three data points—Cloudflare counts, Ad Network counts, and Revenue figures—is where the truth lies.

For e-commerce businesses, the correlation is with the order count. Compare the Cloudflare traffic for specific product pages with the number of orders generated. If a product page has 10,000 views in Cloudflare but only 5 orders, the conversion rate is 0.05%. This might be normal for a niche product, but if you compare it to industry benchmarks or the seller’s own claim of "high conversion," you will see the discrepancy. Use this data to negotiate a lower price based on the actual quality of the lead flow. If the traffic is cheap and low-quality, the lifetime value (LTV) of the customer is lower, and the business valuation should reflect that. Always remember that you are buying the cash flow, not the vanity metrics. The cash flow is only as real as the traffic that generates it.

Common Red Flags in Cloudflare Data

Here are the specific indicators that should trigger an immediate deep dive or a walk-away decision. First, if the "Top IPs" list shows 10 or more IPs accounting for more than 5% of total traffic, this is a strong indicator of bot farms. Real human traffic is distributed across millions of unique residential IPs. A small number of IPs generating high volume is classic bot behavior. Second, if the traffic graph shows a sharp, vertical line up at a specific second of the minute, rather than a smooth curve, it demonstrates automated scripting. Human browsing creates a Gaussian distribution of loads. Bots create spikes. Third, if the average response time from Cloudflare is extremely low (under 50ms) for a complex site, it may indicate that the traffic is not actually fetching the full page, but just hitting the headers. This is a technique used by some bots to inflate view counts without consuming bandwidth.

Another red flag is the "Cacheability" of the traffic. If a huge portion of the traffic comes from CDNs or cloud providers that do not require authentication, and the site is serving static content, it is easier to cache-bust and inflate numbers. Look for high volumes of 304 (Not Modified) responses. While this is normal for performance, an abnormal spike in 304s from datacenter IPs suggests bots are checking for content updates. Finally, check the "TLS" handshake data. If a significant portion of traffic is using older, insecure TLS versions, it suggests the clients are not modern browsers, which again points to scripting tools that do not maintain up-to-date security libraries.

These technical details might seem granular, but they are the difference between a business that performs and one that fails post-acquisition. A buyer who ignores these signals is essentially buying a black box. Once you have identified these red flags, document them. Use them as leverage in negotiations. If you find 20% of the traffic is bot-generated, you should request a 20% reduction in the purchase price, or at least a warranty that protects you if the traffic quality drops below a certain threshold after closing. The data will tell you everything you need to know if you know how to read it.

Warning: Do not rely solely on the "Bot Protection" dashboard in Cloudflare. It only tells you what was stopped. It does not tell you what slipped through *before* protection was tightened. You must analyze the historical data from when the seller was operating the site under their default settings. If they recently switched on aggressive bot protection, the historical data may be skewed, and the current "clean" data might not represent the long-term average. Always ask for logs from a period where bot protection was minimal to get a true baseline of raw traffic quality.

Building Your Technical Due Diligence Checklist

Due diligence is a process, not a one-off check. You need a systematic approach to ensure you don’t miss anything. Below is a comprehensive checklist that you should use for every acquisition over a certain threshold of revenue. This checklist is designed to be practical and technical, focusing on the data that cannot be easily faked. You should export these reports, save them as PDFs, and include them in your due diligence file. This documentation protects you legally and financially if the seller claims ignorance of the traffic issues later on.

This checklist is not just for large acquisitions. Even for a $20,000 micro-niche site, you need to ensure the traffic is real. The cost of verifying traffic is zero if you use the free tier of Cloudflare analytics, or minimal if you hire a technical due diligence expert. The cost of buying fake traffic is 100% of your investment. The asymmetry is so extreme that skipping this step is inexcusable. Use this list to guide your conversation with the seller, as well. Asking these questions shows you are informed and serious. It often makes sellers who are hiding data more nervous, which can reveal the truth faster.

  1. Verify IP Source Distribution: Export the "Top IPs" report from Cloudflare for the last 90 days. Ensure that no single IP or group of 10 IPs accounts for more than 5% of total requests. If you see clusters, investigate the geographic origin of those IPs.
  2. Analyze Geographic Spread: Compare the "Top Countries" in Cloudflare with the seller’s target market. Flag any country that is outside the intended audience and accounts for more than 10% of traffic. Investigate if this traffic is monetizing or if it is junk.
  3. Check for Bot Challenges: Review the "Firewall Events" for the last six months. Look for spikes in HTTP 403 or 429 errors. High volumes of these errors indicate heavy bot activity or DDoS attacks that may have inflated real traffic counts in previous months.
  4. Correlate with Ad Network Data: Cross-reference the monthly session counts in Cloudflare with the impression counts in your ad network (e.g., AdSense, Mediavine). A ratio closer to 1:1 is ideal. A ratio of 3:1 suggests significant bot filtering by the ad network.
  5. Inspect User Agent Strings: Look at the "User Agent" breakdown. If one specific user agent version accounts for an unusually high percentage (e.g., 20%) of traffic, it likely indicates a specific bot signature. Real user agents should be diverse.
  6. Evaluate Traffic Velocity: Check the "Request Rate" per second. Look for graphs that show unusual spikes, such as vertical lines, rather than smooth curves. Vertical spikes are characteristic of automated scripts sending requests in bursts.
  7. Review Subdomain Analytics: Do not just look at the main domain. Check all subdomains (e.g., store.domain.com, app.domain.com). Bots often target subdomains because they are less monitored. Ensure the traffic quality on subdomains matches the main domain.
  8. Verify TLS Handshake Data: Look at the encryption details if available in your specific Cloudflare plan. A high percentage of older TLS versions indicates non-browser clients. Modern human traffic should predominantly use TLS 1.2 or 1.3.

How to Negotiate Based on Traffic Quality

Once you have your data, it’s time to negotiate. If you find significant bot traffic, do not be shy about it. Present the data to the seller. Say, "Our technical review shows that approximately 15% of the traffic originates from datacenter IPs, which are typically non-human. We need to adjust the valuation to reflect the net human traffic, which is X." This is a factual statement, not an insult. It shows you have done your homework. Many sellers will accept a reduction in price because they do not want to deal with the scrutiny of a full audit.

You can also negotiate the structure of the deal. Instead of a lump sum, consider an earn-out structure where a portion of the payment is tied to maintaining certain traffic quality metrics for the first 6-12 months. This shifts the risk back to the seller. If the bot traffic stops and the human traffic remains, you pay more. If the traffic quality drops, you pay less. This aligns the incentives. It also gives you time to integrate the business and monitor the traffic with your own tools. This strategy is particularly effective for businesses with high ad dependency, where traffic quality is the single biggest determinant of profit.

Finally, do not let perfection be the enemy of good. You will never find 100% human traffic on the open internet. There is always some noise, some data center traffic, some occasional bots. The goal is to ensure the majority of the traffic is high-intent and high-quality. If 90% of the traffic is clean, that is a good sign. If 50% is clean, that is a bad sign. Use the threshold to make your decision. And remember, you can always ask for a second opinion. The community of buyers who use platforms like Deal Alert AI is vast and connected. Sharing insights on which traffic sources are currently known to be spammy is a form of collective intelligence that saves investors millions of dollars. Stay informed, stay technical, and stay safe.

Pro Insight: The most successful buyers do not just check the analytics once. They set up alerts. If you can, set up a Cloudflare alert for the domain during the transition period (if you have admin access) or monitor the public performance metrics closely. If the traffic pattern changes abruptly immediately after the sale, it may indicate that the seller was running a temporary promotion or using a specific traffic source that will not continue. Monitoring the "Before" and "After" is the ultimate test of traffic sustainability.

Final Thoughts on Technical Due Diligence

Buying an online business is no longer just a marketing or finance exercise; it is a technical operation. The days of buying a site based on a screenshot of a Google Analytics dashboard are over. Smart buyers are going deeper. They are using tools like Cloudflare to look at the raw data of the internet. They are connecting the dots between IP addresses, geolocation, and revenue. This technical proficiency is your moat. It is your competitive advantage in a market filled with amateur buyers and professional manipulators.

As you progress in your journey as a buyer, you will want to automate as much of this process as possible. Manual analysis of thousands of logs is tedious. That is why we built the integrations and AI models at Deal Alert AI to help streamline this. But you must understand the fundamentals first. You must know what you are looking for. You must know why IP reputation matters. You must know that a clean dashboard does not equal a clean business.

Take the information in this guide and apply it to your next deal. Do the work. Ask the hard questions. Demand the data. The truth is in the logs. And the logs never lie. If the data looks good, and the technical verification backs it up, you have a solid foundation for a profitable investment. If it does not, walk away. There will always be another deal. But your capital is finite. Protect it with knowledge. Good luck on your next acquisition.

By Sophal Lanh, Founder of Deal Alert AI: Sophal built Deal Alert AI after years of analyzing online business acquisitions and missing time-sensitive deals. The platform tracks and scores 100+ listings daily across Empire Flippers, Flippa, Acquire.com, and Quiet Light. Learn more →

Get Deals Before Other Buyers

We scan Empire Flippers, Acquire, Flippa, and Quiet Light daily. The best sub-$500K businesses are gone within 48 hours.