The tracking chaos in B2B: Which analytics tool lies (or doesn't)?

That's why all your analytics tools show different values

Massive data gaps in GA4: How to still measure your B2B reach correctly

Anyone who runs a B2B website knows that frustrating moment: A look at the various analytics tools often reveals completely different realities. While Jetpack reports solid traffic in the WordPress backend, Google Analytics (GA4) suddenly shows 40 percent fewer users, Cloudflare presents much higher figures, and Semrush's traffic estimates seem to come from a completely different world. The obvious question then is usually: "Which tool is lying?"

The short answer: None – but each measures in a completely different way. This article debunks the myth of "one right number." It details why tag-based systems like Google Analytics 4 (GA4) have massive data gaps due to GDPR and the lack of cookie consent, why network-based (edge) solutions like Cloudflare are often closer to the actual reach, and why you should never mistake Semrush's traffic figures for real visitor data. Instead of getting bogged down in tool debates, you'll learn how to correctly assess the respective strengths of each system, avoid systematic errors (such as VPN biases in IP location determination), and build a strategic analytics setup that finally enables you to make sound decisions for your B2B marketing.

The illusion of the "exact" number

Anyone running a B2B website will sooner or later experience the same moment: You open Jetpack in the WordPress backend, check the visitor statistics, then open Google Analytics, Cloudflare, or Semrush – and are presented with three or four different realities. Sometimes the figures seem close, sometimes they differ by 30, 50, or even 100 percent. The spontaneous reaction is almost always the same: "Which tool is lying?" or, to put it more positively: "Which tool can I actually trust with my KPIs?" This question is particularly relevant in the B2B context because it often involves smaller, focused target groups, complex decision-making processes, and a strong interrelationship between marketing metrics and sales activities.

It is important to be transparent about what this article can – and cannot – achieve. The article deliberately presents only a selection from the overall complexity of web analytics, tracking technology, data protection, and the tool landscape. It examines selected components that are particularly relevant in practice: the different measurement approaches of Jetpack and Cloudflare, the GDPR-related limitations of Google Analytics in Europe, the accuracy of IP geolocation at the country level, and the model-like nature of Semrush data. Many other aspects – such as alternative tools, individual configurations, special cases in specific industries, or in-depth technical details – are only touched upon or not addressed at all.

Precisely because of its scope and clear focus on typical B2B questions, this article provides a very strong framework for a better understanding of the topic. It helps you internalize the fundamental thought patterns: that different tools answer different questions, that legal requirements and user behavior systematically distort measurements, and that modeled figures should be interpreted differently than actual log data. With this foundation, you can draw significantly more informed conclusions for your own company, your industry, and your specific product—for example, which key performance indicators (KPIs) you entrust to which tool, how you differentiate between human and bot traffic, what role SEO visibility plays compared to actual visitor behavior, and how heavily you want to rely on IP-based country data. While the article doesn't replace individual implementation or legal advice, it provides a solid foundation for making more conscious and strategic internal decisions, selecting the right tools, and developing your reporting logic.

The central challenge: Different tools don't simply measure "the same thing, just inaccurately," but rather they systematically measure different things – using different technical methods, legal frameworks, and assumptions about user behavior. Jetpack attempts to give WordPress users a quick overview but lacks transparent, finely controllable bot filters. Google Analytics offers in-depth marketing analyses but is limited in the EU by cookie consent, Consent Mode v2, and strict GDPR regulations, leading to sometimes significant data gaps. Cloudflare Web Analytics, on the other hand, measures at the network edge, is cookieless, and filters bots based on its own machine learning stack – thus delivering a different, often "cleaner" picture of actual requests. Finally, Semrush doesn't measure any actual visitors at all but models traffic from rankings, search volume, and clickstream data.

Anyone who treats all these tools like interchangeable thermometers will inevitably run into contradictions. This article addresses precisely this issue: It shows why the figures from common tools differ, where their respective strengths and weaknesses lie, and how you can combine them in a B2B environment to obtain reliable, decision-making metrics. The goal isn't to declare a "winner," but rather to understand the nature of each system: Jetpack as a fast editorial dashboard, Cloudflare as a robust source for real-world reach, Google Analytics as a marketing analytics engine within data privacy regulations, and Semrush as a strategic SEO and competitor radar. Once you clearly define these roles, many apparent contradictions disappear—and you can use data instead of endlessly debating it.

Why web statistics always differ

The first step towards making sound analytics decisions is a dispassionate look at the measurement logic. Three axes are crucial: Where is the measurement taken (server/edge vs. browser), how is it taken (event tracking vs. modeled traffic), and what is filtered (bots, aggregators, internal users). The simplest distinction is between server- or edge-based tools and tag-based systems. Edge-based solutions like Cloudflare see every HTTP request that passes through the CDN, regardless of whether the browser loads JavaScript or accepts cookies. Tag-based systems like Google Analytics or Jetpack rely on a JavaScript snippet being executed in the user's browser – anyone who blocks JavaScript, removes trackers via browser extensions, or leaves the page very quickly is excluded from the measurement.

Furthermore, there's the legal dimension: Google Analytics 4 (GA4) simply cannot operate in the EU without valid consent for analytics/cookies. This means that a significant percentage of actual traffic—between 30 and 70 percent, depending on the target group—remains completely invisible in the data. Cloudflare Web Analytics, on the other hand, works without cookies and with minimal personal data collection, so no explicit consent is required, and therefore no one gets lost "under the banner." Jetpack, however, occupies a gray area: While it uses a script, Automattic doesn't document the precise effects of ad blockers, script blockers, and privacy tools as transparently as GA4 does.

The third major aspect is the handling of bots, crawlers, and news aggregators. Google Analytics automatically filters out many known bots using the IAB bot list and its own algorithms, but without providing users with granular control options. This means that certain "benign" crawlers and aggregators disappear from the reports, even though they can be important distribution channels in a B2B context. Jetpack lacks a similarly well-documented bot strategy; anecdotal evidence shows that both bot traffic and legitimate but technically suspicious referrers can be excluded from the statistics. Cloudflare, on the other hand, relies on its own machine learning stack, combining IP reputation, behavior, JavaScript challenges, and bot scores. This allows you to decide very precisely which traffic classes you want to see, which you want to analyze further, and which you want to block completely.

Related to this:

• The market for news aggregation such as Google News and media monitoring is substantial for B2B decision-makers and is growing significantly

Semrush takes a completely different approach. It doesn't directly measure visitors to your site. Instead, Semrush aggregates search volumes, ranking positions, click probabilities, and external clickstream data into a model that estimates how much organic search traffic a domain or URL is likely to receive. Studies and practical experience show that these estimates can sometimes deviate by 30 to 60 percent for small and medium-sized websites, and even more in extreme cases – the relative trend (more/less than competitor X) is usually more useful than the absolute value. Comparing Semrush figures directly with GA4 or Cloudflare data essentially means comparing measurements from actual logs with model assumptions – the discrepancy is therefore not an error, but inherent to the system.

The practical consequence is this: Instead of pitting tools against each other, you should categorize them according to their measurement logic and consciously use them for different questions. Edge- and server-based solutions provide the most robust view of actual requests; tag-based tools with mandatory consent are ideal for marketing attribution and funnels, but only represent a slice of reality; model-based tools like Semrush are suitable for market and competitive analysis, not for operational KPI reports. If you clearly separate these roles and internally define which system is the "leading" one for which metric, many apparent contradictions in your dashboards will disappear on their own.

Jetpack vs. Cloudflare: What really matters?

For operators of WordPress-based B2B sites, Jetpack initially seems like the obvious solution: activate a plugin, log in, and the dashboard immediately displays visitor numbers, top posts, and referrers directly in the backend. This proximity to daily editorial work is convenient, but can easily lead to a dangerous false sense of security. Jetpack gives the impression of providing an objective truth about traffic without making transparent the technical and methodological decisions being made behind the scenes. Cloudflare Web Analytics takes a different approach: it captures data at the network edge and consistently focuses on cookieless, data-minimizing measurement. This results in two very different perspectives on the same traffic – and for B2B decision-makers, the question arises as to which one is more suitable for strategic KPIs.

Let's first look at Jetpack. The system combines server-side information from WordPress with client-side elements integrated via scripts. In practice, this means that only page views that are delivered cleanly and processed in the browser as Jetpack expects will reliably appear in the statistics. Logged-in users—such as editors, administrators, external authors, or agency partners—are often excluded by default to keep internal activity out of the statistics. This makes sense, but it skews the data if your B2B site relies heavily on internal usage, such as portals, knowledge bases, or partner areas. Additionally, Jetpack filters spam referrers and obvious bots relatively aggressively. While this keeps the dashboard clean, it can also cause legitimate but technically "unusual" sources—such as certain news aggregators, industry portals, or monitoring services—to disappear from the reports.

This is precisely where a core problem for B2B sites lies: Many important distribution channels behave technically like bots or crawlers, without actually being considered "spam" from a content perspective. An industry aggregator that retrieves your RSS feeds, a specialist portal that embeds teasers for your articles, or a monitoring service that regularly makes page requests – all these accesses can either be classified as "robots" and hidden in Jetpack, or recorded as unclear. At the same time, other, less obvious bots are sometimes not detected at all and are included in the visitor count as normal. The result: You get a mix of genuine user visits, undetected bots, and incompletely recorded, high-quality machine visits. The direction of the distortion is difficult to assess because the system only provides limited insight into its filtering logic.

Cloudflare Web Analytics takes a significantly more technical, "bottom-up" approach. The system sits at the edge of the content delivery network and, in principle, sees every HTTP request that reaches your domain, regardless of whether your WordPress site responds correctly or the browser executes JavaScript. This gives you a more robust foundation, especially for answering the question: "How many requests actually arrive at my infrastructure?" At this level, Cloudflare combines comprehensive bot detection, IP reputation, heuristics, and optional machine learning models to identify malicious or clearly automated traffic and—depending on the configuration—block it even before it's delivered. Therefore, what you see in the standard analytics tends to be more focused on human interactions than a simple log counter or an unconfigured JavaScript tracker.

Another key difference: Cloudflare Web Analytics is designed from the outset to operate without cookies and without user profiles. It doesn't track individual visitor activity over extended periods, but instead generates aggregated metrics based on page loads and requests. For you as a B2B operator in the EU, this means two things. First, you can generally measure reach without an explicit analytics cookie banner because no personally identifiable tracking cookies are set. Second, you won't lose users who reject tracking consent or automatically block cookie banners. This effect is particularly significant among tech-savvy target groups – IT decision-makers, developers, and technical buyers. While Jetpack and especially GA-based solutions will increasingly lose visibility in this area, Cloudflare remains close to reality when it comes to pure traffic data.

One advantage of Cloudflare, often underestimated in the B2B context, is its ability to segment machine and human traffic. Instead of simply "counting or blocking" all bots, you can use bot scores, user-agent signatures, and IP lists to fine-tune which types of crawlers are included in your standard report, which are reported separately, and which are completely suppressed. For example, you can define a "Website Reach (Humans)" metric that only considers traffic with a low bot score, and a separate "Ecosystem Reach (Crawling and Aggregation)" metric that deliberately groups together news aggregators, industry portals, price comparison services, and AI crawlers. Jetpack doesn't offer this level of differentiation – it usually remains a binary "show/hide" without allowing you to actively control this decision.

Of course, Cloudflare also has its limitations. In the free version, data is often collected and extrapolated on a sample basis, which limits the absolute precision of individual figures. Some very aggressive privacy setups or specific enterprise proxies can also lead to certain requests not being recorded as expected. Compared to Jetpack's lack of transparency, however, these limitations are easier to understand and calibrate for B2B KPIs. In practice, a pragmatic approach has therefore proven effective: Continue using Jetpack as a convenient editorial tool that gives editors a quick overview of popular content, but rely primarily on data collected at the network edge and cleaned with modern bot detection for reporting, budget decisions, and international reach analyses. For the question of "real, business-relevant visibility," Cloudflare is thus the more robust foundation in most B2B setups.

Benefit from Xpert.Digital's extensive, five-fold expertise in a comprehensive service package | R&D, XR, PR & Digital Visibility Optimization - Image: Xpert.Digital

Xpert.Digital possesses in-depth knowledge across various industries. This allows us to develop tailored strategies precisely aligned with the requirements and challenges of your specific market segment. By continuously analyzing market trends and monitoring industry developments, we can act proactively and offer innovative solutions. The combination of experience and expertise generates added value and provides our clients with a decisive competitive advantage.

More information here:

• Benefit from Xpert.Digital's 5 areas of expertise in one package – starting from just €500/month

Google Analytics vs. Cloudflare: GDPR, consent, and data gaps

Google Analytics, in its current GA4 version, is practically the standard in traditional online marketing. Campaign tracking, funnels, goals, attribution – all of this is very powerful in GA4. However, for a B2B website in Europe, the tool reaches its limits, which have nothing to do with technology in the strict sense, but rather with legal frameworks and user behavior. This is precisely where the difference to Cloudflare Web Analytics becomes particularly clear. While Google Analytics relies on a client-side, cookie-based tracking model that requires explicit consent, Cloudflare is optimized for cookieless reach measurement at the network edge. This means that in everyday B2B practice, GA4 often only captures a fragment of reality, while Cloudflare is closer to the actual number of page views and user interactions.

The biggest stumbling block for Google Analytics 4 (GA4) in the EU is the consent requirement. Because Google Analytics recognizes visitors via cookies and unique identifiers and analyzes their behavior across multiple sessions, it is legally considered to require consent. In practice, this means that unless a visitor explicitly agrees to Analytics in the consent banner, GA4 is either not allowed to fire its data or its functionality is severely restricted. Every rejected or ignored consent represents a lost session in your data. Rejection rates range from moderate to dramatic, depending on the industry and the banner design. In the B2B sector with data-sensitive target groups—such as IT, manufacturing, and the public sector—skepticism towards tracking tools is particularly pronounced. Even with Google's Consent Mode, which attempts to algorithmically model data loss, your figures ultimately become a mixture of measurement and estimation and no longer fully reflect what actually happens on the website.

Cloudflare Web Analytics circumvents this problem because it takes a fundamentally different approach. Instead of tracking individual users via cookies, it collects anonymized, aggregated metrics directly based on network requests. There are no persistent analytics cookies, no personal profiles, and no cross-device recognition mechanisms. This places the tool in a different legal category: For pure, data-minimizing audience measurement, explicit consent via a cookie banner is generally not required. The result: You also see visits where users never make a selection in the consent layer or reject tracking altogether. This effect is particularly significant for B2B decision-makers who reflexively dismiss cookie banners or manage them via privacy extensions. While Google Analytics 4 (GA4) is developing increasingly significant blind spots, Cloudflare remains stable and comprehensive at the page view and country level.

Related to this:

• SST Pioneers | The End of the Cookie Age: Why Companies Are Relying on Server-Side Tracking – Facebook, Pinterest & TikTok

Another structural difference lies in the technical implementation. GA4 relies on the JavaScript tag in the browser. It requires a fully loaded page and a functioning script to register a page view or event. However, many B2B users employ ad blockers, anti-tracking extensions, or strict company policies that block precisely these scripts. Performance issues—such as long loading times or early abandonment—also mean that the GA4 tag simply isn't executed in some cases. In all these situations, the user experiences the page, but you don't see them in GA. Cloudflare takes a more fundamental approach: As soon as a request reaches the CDN, it can potentially be included in the analytics count. Even if the browser blocks scripts or the user leaves the page extremely early, this interaction is significantly more likely to appear in your statistics.

The handling of bots and crawlers also differs significantly. Google Analytics automatically filters out many known bots based on predefined lists and its own heuristics, without offering users much control. This is convenient, but lacks transparency. It can lead to certain types of news aggregators, monitoring services, or search engine experiments silently disappearing from reports, even though they are relevant to your B2B distribution ecosystem. At the same time, "smart" bots that mimic human behavior often slip through the standard filters and distort engagement metrics. Cloudflare uses its own bot detection system, which combines IP reputation, request patterns, and optional additional challenges. The key advantage: You can control much more granularly what is blocked, what is made visible, and what is included in standard reports. This allows you to set up your reporting so that human traffic and machine access are analytically separated, but both are consciously considered.

Related to this:

• B2B information gathering with Google Alerts, Google News and Google Discover – Automated notification economy

Of course, this doesn't make Google Analytics 4 (GA4) "bad." On the contrary: as soon as it comes to marketing-specific questions—campaign performance, conversion path, attribution models, event tracking—GA4 really shines. You can track very precisely how a click from a specific ad leads to a conversion, how long users spend interacting with certain elements, and at which points in the funnel they drop off. Cloudflare Web Analytics doesn't offer this level of depth; its focus is more on an overview of visits, countries, devices, and paths. For a B2B organization in the EU, the pragmatic approach is therefore obvious: use Cloudflare as the "single source of truth" for reach and country distribution—that is, for the question "How much real visibility do we actually have?"—and use GA4 as a supplement where you want to delve deeper into campaigns and conversions with explicit consent. This way, each tool can play to its strengths in its respective area, and you avoid legal and technical limitations obscuring the bigger picture.

IP geolocation: How accurate is the country assignment?

When you look at country statistics in B2B reporting, they often appear very precise: 62% Germany, 14% Switzerland, 9% Austria, the rest distributed across other markets. Behind this is almost always IP geolocation – the attempt to deduce a visitor's country, region, or even city from their IP address. The obvious question is: How reliable is this, really? Especially if you're linking sales priorities, trade fair budgets, or account-based marketing to these analyses, you want to know if you can rely on the figures. The good news: At the country level, the technology is surprisingly good these days. The not-so-good news: Certain B2B-specific configurations, such as VPNs, corporate proxies, or central gateways, distort the picture – and below the country level, the accuracy drops significantly.

Let's start with the country level. Major geolocation data providers like MaxMind, IPinfo, DB-IP, and IP2Location report country-specific accuracy of around 99% and higher. Studies that compare real user locations with IP databases essentially confirm this: In Western Europe and North America, the hit rate is over 99% when considering standard residential and business connections without a VPN. The reason is structural: IP address blocks are typically assigned by regional internet registries like RIPE or ARIN with country-specific identities, and most internet providers primarily serve one country. In short, whether an IP address belongs to "DE," "FR," or "US" is, in most cases, a straightforward task. For broad market analyses—such as whether your traffic predominantly originates from the DACH region (Germany, Austria, and Switzerland) or whether one region is gaining more traction—country-level IP geolocation is therefore sufficient to support strategic decisions.

However, this high accuracy has its limits, and these limits are particularly relevant in the B2B environment. The most significant source of interference is VPN connections and corporate proxies. Many companies bundle all web traffic through central nodes, sometimes even located in other countries. An employee in Munich whose company routes their internet access through a central gateway in the Netherlands or the USA will then appear in the geolocation data as "NL" or "US". A similar situation exists with classic consumer VPNs used for data protection or compliance reasons: The IP address then belongs to a data center in the selected country, not to the employee's physical location. In industries with high VPN usage – IT, finance, globally operating industrial companies – this effect can lead to a portion of your actual German audience appearing statistically as international traffic. This cannot be completely eliminated; it is an inherent characteristic of IP-based location tracking.

The second major source of distortion is news aggregators, crawlers, and other machine access. If a US-based aggregator reads your German B2B articles, this access will naturally appear as US traffic in the geolocation data. This doesn't mean your reach there suddenly increases, but simply that a server in that country is accessing your content. Strictly speaking, this is "noise" for traditional marketing KPIs, but it can still be interesting for a technical and strategic analysis – for example, as an indicator of where your content is stored, mirrored, or processed by AI models. It's crucial that you clearly separate these types of access sources from human traffic in your analyses, rather than mixing them together in country-specific statistics. Tools like Cloudflare help you with this by separately identifying bots, known crawlers, and data center IPs, allowing you to decide whether to include them in country-specific reports or analyze them separately.

How exactly does Cloudflare position itself regarding geolocation? Cloudflare relies on an integrated IP geodatabase and now supplements this with data from specialized providers like IPinfo to achieve a high level of accuracy. Every request that travels across the network is enriched with attributes such as `CF-IPCountry`, `CF-Region`, and `CF-City`, which you can use in both your origin code and Cloudflare Analytics. In practice, developers report that the `CF-IPCountry` header provides valid country codes for almost all regular visitors and only rarely—for example, with Tor connections or very exotic network setups—does it return an "unknown" value. This suggests that Cloudflare performs similarly well at the country level as established geodatabases and offers a very solid foundation for analytics purposes. Cloudflare itself points out that even this system cannot “magically” penetrate VPNs, proxies and Tor – if a user intentionally conceals their origin, Cloudflare will only see the exit node.

You should be considerably more cautious with analyses below the country level, regardless of whether they come from Cloudflare, Google Analytics, or other tools. Studies on city and region accuracy show that hit rates at this level can drop to between 50 and 80 percent, depending on the region. In Western Europe, city accuracy is typically between 65 and 80 percent, according to comparative studies, and often lower in rural areas or with mobile connections. This is due to technical reasons: Many providers bundle huge IP blocks and assign them to large regions or entire states. Mobile networks also have carrier-grade NAT, where thousands of users share IP address pools that are sometimes collectively assigned to a major city or the provider's headquarters. So, if your analytics suggest you have twice as many visitors from city A as from city B, this should be considered a rough guideline rather than an absolute truth.

For your B2B reporting, this translates into a pragmatic approach to IP geolocation. At the country level, the mapping is usually accurate enough to align sales regions, language versions, and broad market strategies—especially in Europe and North America. You should factor in distortions caused by VPN and proxy usage, particularly when working with globally operating corporations or if your target groups are security- and privacy-conscious. Below the country level, however, you shouldn't overemphasize city or town-level data. Use it as an indicator, not as the basis for hard budget decisions. Wherever possible, supplement IP-based country data with first-party signals: information from forms, CRM data, account assignments, and sales feedback. Combining robust country metrics from tools like Cloudflare with the more granular, person-based information from your own system creates a picture that reflects the real B2B world far more accurately than any pure IP statistics could.

B2B support and SaaS for SEO and GEO (AI search) combined: The all-in-one solution for B2B companies - Image: Xpert.Digital

AI search changes everything: How this SaaS solution will revolutionize your B2B ranking forever.

The digital landscape for B2B companies is undergoing rapid change. Driven by artificial intelligence, the rules of online visibility are being rewritten. For companies, it has always been a challenge not only to be visible in the digital mass, but also to be relevant to the right decision-makers. Traditional SEO strategies and managing local presence (geo-marketing) are complex, time-consuming, and often a battle against constantly changing algorithms and intense competition.

But what if there were a solution that not only simplified this process but also made it smarter, more predictive, and far more effective? This is where the combination of specialized B2B support with a powerful SaaS (Software as a Service) platform comes into play, specifically designed for the demands of SEO and GEO in the age of AI search.

This new generation of tools no longer relies solely on manual keyword analysis and backlink strategies. Instead, it leverages artificial intelligence to more accurately understand search intent, automatically optimize local ranking factors, and conduct real-time competitive analysis. The result is a proactive, data-driven strategy that gives B2B companies a decisive advantage: they are not only found, but perceived as the leading authority in their niche and location.

Here's the symbiosis of B2B support and AI-powered SaaS technology that transforms SEO and GEO marketing, and how your company can benefit from it to grow sustainably in the digital space.

More information here:

​

• B2B Support & Blog for SEO, GEO and AIS – Artificial Intelligence Search
• Forget expensive SEO tools – this alternative dominates with unbeatable B2B features

Semrush: Strong rankings, weak traffic figures

Semrush is standard equipment for many SEO and marketing teams when it comes to keyword research, competitor analysis, and visibility comparisons. The temptation is strong to interpret the traffic figures displayed there in the same way as data from Google Analytics, Cloudflare, or server logs: as a measurement of actual visitors. This is precisely where one of the biggest sources of error in reporting lies. Semrush doesn't measure anything on your website itself, but rather models traffic from external signals – primarily rankings, search volumes, and clickstream data. For strategic questions ("Who is bigger than whom?", "Where are the market opportunities?") this works surprisingly well, but for operational B2B KPIs ("How many actual visitors did we have?") it's only of very rough use.

To understand Semrush's strengths and weaknesses, it's worth examining its data. Semrush monitors search engine results pages (SERPs) for millions of keywords, combines this data with estimated search volumes and typical click probabilities per position, and supplements it with clickstream data from panels or partnerships. This creates a model that estimates how much traffic a domain or URL is likely to receive when it ranks for specific terms in specific positions. This approach has two consequences. First, Semrush only sees a portion of reality—namely, the parts represented by search terms included in its own keyword set. Long-tail searches, niche terms, and many highly specific B2B queries may simply be missing. Second, direct traffic, referral visitors, email clicks, social media engagement, and paid campaigns are only captured very indirectly and with a high degree of uncertainty.

Numerous comparative tests using real analytics data demonstrate the practical implications of this model's nature. Agencies and SEOs who have compared Semrush traffic with Google Analytics or Google Search Console regularly report discrepancies of 20 to 50 percent – ​​both higher and lower. An analysis of 30 websites revealed that Semrush was within ±10 percent of the Search Console values ​​in only two cases, while for the remaining domains, the estimates were off by an average of +152 percent (overestimation) or -51 percent (underestimation). Other analyses conclude that Semrush figures for smaller sites with fewer than 10,000 visits per month often deviate from the actual traffic by 40 to 60 percent. Extreme examples are also documented: domains that Semrush claims have 110,000 organic visits per month, but Google Analytics shows only around 8,000 visitors.

It's important to understand that these discrepancies aren't "errors" in the traditional sense, but rather a consequence of the methodology. Semrush doesn't have access to your actual user data; it has neither access to your Google Analytics nor your server logs. It estimates based on external signals and can therefore only approximate reality. Nevertheless, the tool has some clear strengths. Semrush excels at making relative statements: If Semrush shows that Domain A has roughly twice the traffic of Domain B, this trend is accurate in about 80 percent of cases, according to tests. This is perfectly adequate and often extremely helpful for competitive analysis ("Are we bigger than manufacturer X?"), market analysis ("Which players dominate this keyword cluster?"), and trend monitoring ("When did the competitor experience a traffic peak?").

Even in terms of rankings themselves, Semrush is usually surprisingly reliable. Position tracking data often falls within one or two positions of what Google Search Console indicates as the average position. While rankings naturally fluctuate and tools only measure snapshots in time, this is more than sufficient for operational SEO work – monitoring keyword clusters, SERP features, and competitor activity. The situation becomes problematic when direct visitor numbers are "derived" from these rankings and communicated as hard KPIs – for example, in the form of statements like "According to Semrush, we have 12,000 visits per month." Such formulations obscure the fact that these are modeled estimates, which are notoriously unreliable, especially in the B2B environment with its high volume of long-tail traffic, niche keywords, and strong direct or referral share.

Semrush performs particularly poorly when it comes to breaking down traffic sources and smaller target groups. Analyses show that the tool's internal estimates for direct and referral visits can be off by 50 to 70 percent for many pages. This is logical: Without direct access to your logs or tag data, the tool can only very indirectly "guess" how much direct or referral traffic might be generated. In a B2B context, where a large portion of relevant traffic often originates from email newsletter links, personal referrals, partner portals, or internal intranet links, these estimates are correspondingly unreliable. Channel-specific interpretations ("Semrush shows that 60% of our traffic is organic") are therefore more like rough indicators than reliable decision-making tools.

For B2B websites, this results in a clear usage pattern. Semrush is not a replacement for true analytics tools like Cloudflare, GA4, or Matomo when it comes to measuring actual visitors, sessions, pageviews, and conversions. It's a complementary, strategic tool that allows you to analyze market and competitor data, keyword opportunities, and visibility trends. The relevant questions are therefore not "How many visitors did we have according to Semrush?", but rather: "How does our domain compare to competitor X for this keyword cluster?", "Which countries are gaining relative visibility?", or "Which pages are underperforming or overperforming compared to our competitors?". When using Semrush figures in internal reports, they should always be explicitly labeled as estimates—ideally supplemented by actual metrics from Cloudflare or GA4 for your own site.

In short: Semrush provides strong signals for rankings, market share, and SEO strategies, but weak, sometimes wildly inconsistent, figures for absolute traffic metrics. Those who respect this limitation and use the tool as intended gain valuable insights into visibility and competition in B2B marketing. However, those who try to use it to replace "external Google Analytics" are building their metrics on a foundation of model assumptions. The key is to combine Semrush data with real visitor data from Cloudflare or GA4: visibility and potential from Semrush, actual usage and conversions from first-party analytics – especially in the B2B environment, this combination creates a significantly sharper, more reliable picture than either approach alone.

Specific recommendations for B2B sites

After all the differences between Jetpack, Cloudflare, Google Analytics, and Semrush, the practical question arises: How do you, as a B2B website operator, build a setup that delivers strategically sound insights without getting bogged down in tool wars? The most important insight is this: You don't need a "perfect" tool, but rather a clear division of roles among your tools. Instead of asking "which tool is lying," you should define which system is leading for which question and how you can meaningfully combine the data from it. This is precisely where mature B2B analytics setups differ from ad-hoc tool collections.

First, you should define a hierarchy of your measurement goals. At the top is usually the question of actual visibility: How many people see our content, which countries do they come from, and which pages do they use to access it? Edge or server-side solutions like Cloudflare Web Analytics are particularly well-suited for this because they also capture users who reject cookie banners, use ad blockers, or block JavaScript. You can define Cloudflare as your "single source of truth" for reach and country breakdown—not because it's absolutely perfect, but because it provides the most comprehensive data in a GDPR-driven B2B environment. Below this, you can place tools like Google Analytics 4, which delve much deeper into marketing and conversion analysis but only see a portion of this actual reach.

The second step involves examining your legal and technical framework. If you're managing EU traffic and using a consent banner, you must automatically interpret Google Analytics 4 (GA4) data as a subset of reality – studies and case studies indicate data loss of 30 to 60 percent after implementing Consent Mode v2. This doesn't mean GA4 is "unusable." It simply means that you should use metrics like session counts, page views, or conversion rates there primarily for relative trends within the same tool, rather than for absolute reach comparisons between tools. You can use Cloudflare data as a parallel corrective: If, for example, Cloudflare consistently shows roughly twice as many page views from Germany, Austria, and Switzerland (DACH region) as GA4, it's clear that your funnel analyses in GA are primarily based on half of your actual audience. This gap can't be completely closed, but you can make it transparent and factor it into your decision-making.

A third key element is the conscious management of bots, aggregators, and automated traffic. For B2B sites, news aggregators, industry portals, and monitoring services are both Segen and a challenge: they increase visibility but distort raw traffic figures. Your tools handle this differently – Google Analytics 4 (GA4) filters out some traffic rigidly, Jetpack sometimes erratically, and Cloudflare allows for nuanced rules. In a mature setup, you define two metric levels: one for "human interaction" (e.g., only requests with a high human score in Cloudflare, possibly supplemented by GA4 sessions) and one for "automated reception" (crawls, aggregators, AI bots). This allows you to report, for example: "This month, we had 8,000 human user interactions from the DACH region and an additional 2,500 technical requests from aggregators and crawlers." This transparency is much more helpful for stakeholders than hiding everything in a number and then arguing about its plausibility.

The fourth point concerns the effective use of Semrush and similar SEO suites. Instead of viewing Semrush as an "alternative analytics tool," you should clearly position it as a strategic visibility and competitive intelligence instrument. Use Semrush to answer questions like: "How do we perform organically compared to competitor X?", "Which topic clusters are underserved in the market?", or "Which countries are showing growing organic interest?"—not to simply state, "We had exactly 12,300 visits." In your reporting, you can explicitly label Semrush data as estimates ("Semrush Visibility Index," "Estimated Organic Traffic vs. Competitors") while sourcing actual visitor metrics from Cloudflare or Google Analytics 4. This prevents model values ​​and metrics from being inadvertently mixed up.

Finally, you should consistently integrate your web analytics with CRM and sales data. Especially in the B2B environment, simply looking at traffic figures remains abstract as long as they aren't linked to accounts, opportunities, and revenue. Tools that connect web events with CRM records (for example, via UTM parameters, first-party tracking, or IP mapping to firmographic data) provide the missing link: Which companies are visiting the site, what content are they consuming, and how does this correlate with the pipeline and closed deals? Cloudflare and Google Analytics 4 (GA4) provide you with the raw signals, while CRM and marketing automation make the business relevance visible. In your setup plan, this means consciously embedding analytics tools in an architecture where website data doesn't remain in silos but communicates with sales and marketing systems.

For practical application, keep the following guidelines in mind: Use Cloudflare Web Analytics as a reliable foundation for reach, countries, and technical quality; use Google Analytics (GA4) where you conduct funnel analysis and campaign optimization with consent-based tracking; keep Jetpack as a lightweight editorial extension for everyday WordPress use; and use Semrush specifically for SEO visibility and competitor analysis, not as a simple visitor counter. Complement this with tight integration with your CRM so that traffic data becomes genuine pipeline insights. If you clearly document and communicate this role distribution internally, most of the confusion surrounding "conflicting numbers" will disappear – and your B2B organization can finally use web analytics for what it should be: a decision-making tool instead of a source of endless tool debates.

☑️ Our business language is English or German

☑️ NEW: Correspondence in your native language!

Konrad Wolfenstein

I and my team are happy to be available to you as your personal advisor.

You can contact me by filling out the contact form here or simply call me at +49 7348 4088 965. My email address is: [email protected]

I'm looking forward to our joint project.

☑️ SME support in strategy, consulting, planning and implementation

☑️ Creation or realignment of the digital strategy and digitization

☑️ Expansion and optimization of international sales processes

☑️ Global & Digital B2B trading platforms

☑️ Pioneer Business Development / Marketing / PR / Trade Fairs

Market vs. Marketing Knowledge: Why SMEs Block Their Own Growth - Image: Xpert.Digital

A persistent, pragmatic misconception exists among small and medium-sized enterprises (SMEs): that those who know their customers and the market also know how marketing works. However, this very equation is increasingly becoming a strategic trap for many SMEs.

The following article analyzes the often overlooked tension between operational market knowledge (looking in the rearview mirror) and strategic marketing knowledge (the high beam for future market share). Learn why a sole focus on sales targets leads to interchangeability in the long run and how SMEs can mature from "short-distance runners" to distinctive brands by consciously separating and realigning these two disciplines. Because those who understand marketing merely as "colorful pictures for sales" surrender 95 percent of tomorrow's potential customers to the competition without a fight.

More information here:

• The 95/5 Problem: Why Sales Knowledge Alone Hinders Growth in Medium-Sized Businesses