Unmasking Anonymous Visitors: The Tech Behind Person-Level Website Identity
How tools like Rb2b turn website visits into actionable opportunities
If you were forwarded this newsletter, consider joining over 200 weekly readers—some of the smartest GTM devs—by subscribing here:
When we first adopted Clearbit at SpatialChat, my eyes nearly popped out. Having never used ABM software before, I suddenly had a tool that could identify customers' corporate accounts visiting our website. It felt both intimidating and powerful.
But looking back, I realize we weren’t getting as much value as we could have. Seeing which companies visited our page was interesting, but as a horizontal PLG tool, the data wasn’t always actionable.
Say someone from Sony’s engineering or marketing team lands on your website.
What next?
Unless you have a workflow to act on that insight, it’s just an interesting observation.
It was a bit creepy to say “We noticed you watched our landing page 4x.”
Then, we started thinking deeper and come up with a few workflows :
What if we tracked visitors to our Terms page? Nobody reads terms just for fun—they check them when evaluating software for procurement.
What if we monitored our pricing page? If someone from a target company (even if they’re not our primary contact) visits, it’s not necessarily a buying signal, but it could indicate internal discussions.
Same for the /solutions page.
That made me realize something: traditional web analytics tools like Google Analytics or PostHog weren’t enough. For B2B SaaS, where traffic is lower than in B2C, every visitor is a potential goldmine—if you know how to use the data.
The concept isn’t new. Many vendors have shaped this space, primarily through an ABM lens. It's about granularity and workflow, not scale.
Historically, buying signals were promoted as a tool deeply embedded in Account-Based Marketing (ABM). ABM focuses on high-value accounts through tailored strategies, targeting them with personalized experiences and aligning sales, marketing, and customer success efforts toward meaningful engagement.
Traditional marketing automation tools incorporate signal-based marketing that utilizes first-party data—like website visits and product usage. Later, the tools expanded to include second-party data (for example, interactions with partners) and third-party data (such as industry news and competitor research).
Website visitors are part of the first-party data signals, which work well alongside other providers. Modern platforms enable earlier engagement by deanonymizing visitors and integrating these signals into workflows.
The history starts in the 2013 :
6sense ($200M ARR, 2024) built a deep predictive analytics platform for enterprise ABM, combining multiple intent signals.
Clearbit ($41.1M ARR, 2023) focused on SMBs with a self-service model for leaner teams.
Leadfeeder entered the European market, addressing similar needs.
Today, Apollo, Snytcher, and People Data Labs offer accessible versions of these tools, with pricing as low as $49/month—democratizing access to website visitor intelligence.
And buzzy RB2B is redefining the category by mapping anonymized visitors to LinkedIn profiles—a simple yet highly effective approach that is already helping it scale past $22M ARR (2024)
I’ve invited Imran from Syft Data to collaborate and explore the complexity of current IP de-anonymization tools.
Can you build this in-house without relying on vendors?
What’s the best approach?
Let’s break it down.
How Website De-Anonymization Works
Technical Mechanisms
Modern de-anonymization relies on multiple mechanisms:
IP Address Matching & Reverse DNS: Every visitor’s device uses an IP address that is logged when they visit a site. By performing a reverse IP lookup, which involves querying internet registries or IP databases, the visitor’s organization can be inferred. However, basic IP matching reliably identifies only those visitors coming from corporate networks or office locations.
Cookies and Tracking Pixels: Cookies embedded across multiple publishers help recognize visitors within networks, even if their IP changes. Over time, this builds a behavioral profile tied to a unique cookie ID. If that cookie ID is ever linked to a known user (for example, if the person fills out a form to submit a comment or clicks an email), all past anonymous visits can be retroactively attributed to that person or company.
Device Fingerprinting: In the absence of cookies, especially as browsers enhance privacy, fingerprinting can identify repeat visitors. This technique collects a combination of the device’s attributes, such as browser version, OS, screen size, and installed fonts, to generate a unique “fingerprint.” If the same fingerprint appears again, it’s likely the same visitor. Fingerprinting can also help correlate visits across sites in a data network.
Data Enrichment & Third-Party Data: Once an initial identifier, such as a company name, domain, or personal email, is guessed, de-anonymization tools enrich that data with additional context.
They access firmographic databases and other external sources to gather details like company size, industry, revenue, office locations, and relevant contacts.
Accuracy and Challenges: it's important to note that no mechanism is 100% accurate. Pure reverse IP lookup has inherent limitations—it traditionally identifies only 20% of visitors (those on known corporate networks) and fails entirely for employees at home on ISP addresses. Work-from-home and mobile hotspots have made identification harder since many visitors appear as "Comcast" or "Verizon" rather than a company name. Vendors have responded by layering techniques: the best systems combine multiple signals (IP, cookies, device IDs) to improve match rates and accuracy.
Despite these advances, certain factors can still obscure identification:
Shared IPs and NAT: Many small companies or remote workers share IP addresses with unrelated users, such as two different companies using the same coffee shop Wi-Fi. In these cases, an IP alone won’t uniquely identify one company.
VPNs and Privacy Tools: If a visitor uses a VPN or privacy proxy, their apparent IP might point to a VPN provider’s server or a cloud host, not the true origin. This disrupts IP matching, and even the smartest algorithms struggle with this.
Browser Restrictions: Browser-level privacy features, such as Safari’s ITP and Firefox’s ETP, may limit cookie tracking and fingerprinting. Third-party cookies are being entirely deprecated, and even first-party cookies may have capped lifespans. This results in shorter windows to recognize returning visitors unless server-side tracking or fingerprinting is used.
Data Quality & Freshness: IP-to-company databases require constant updates as companies change office locations, ISP allocations shift, and new IP ranges come into use.
Privacy and Legal Constraints: From a compliance standpoint, de-anonymization exists in a gray area that must be navigated carefully. In regions governed by laws like GDPR (EU) and CCPA (California), an IP address or a cookie can be considered personal data if it can be linked to an individual. That's why most B2B "visitor intelligence" tools limit themselves to company-level information to avoid directly handling personally identifiable information (PII).
Key Providers in the Data Supply Chain
Many major ABM and intent platforms—6sense, Demandbase, etc.—don’t operate in isolation. Instead, they rely on or augment their data with specialized third-party providers. The de-anonymization value chain involves multiple layers:
IP Data Providers: These companies specialize in mapping IP addresses to organizational data. Examples include KickFire, IPinfo.io, Digital Element/Digital Envoy, and Neustar.
Data Enrichment Services: Once an IP is matched to a domain, these providers add firmographic and contact-level data to flesh out the insights. PeopleDataLabs (PDL) and Clearbit are examples (Syft uses them) but there are many more.
Data Co-ops and Networks: A newer trend in the ecosystem involves collaborative data sharing to improve identification accuracy. Companies like 5x5coop (Vector) and LiveIntent (Rb2b) aggregate anonymous data from multiple sources to enhance match rates. I’ve also heard that CTV networks, such as Sirista, provide a solid and stable data source.
Historical Evolution
What started with basic IP lookups from LeadLander and Demandbase (2004-2006) has evolved into sophisticated intent platforms.
Early 2000s: Basic IP tracking tools identified ~20% of office-based visitors.
2010s: The space saw rapid growth with companies like Clearbit and PeopleDataLabs launching large-scale enrichment APIs, while 6sense and Lattice Engines introduced AI-powered predictive analytics.
2014: LinkedIn’s $175M acquisition of Bizo validated the space and its future potential.
2020s: Remote work disrupted traditional IP-based tracking, making home/VPN identification more difficult. At the same time, GDPR, CCPA, and cookie deprecation forced vendors to innovate with privacy-first tracking methods. Clearbit, Demandbase, and others quickly adapted, developing new approaches for tracking users while remaining compliant.
As privacy laws tighten and remote work becomes the norm, the future of de-anonymization will depend on multi-layered tracking strategies. Companies will need to combine IP data, behavioral signals, and cooperative data networks to maintain match rates and generate meaningful insights.
Buying Signals – Why They Matter
Stepping back, let’s look at why internet signals are critical and how to use them effectively.
These signals suggest a company or individual—who fits your ICP—is actively considering or ready to buy a product like yours. They help companies prioritize accounts, personalize engagement, and align sales and marketing efforts.
Buyers don’t simply wake up and decide to purchase. Their journey includes research, comparison, and internal discussions. They analyze their pain points, explore solutions, and work with stakeholders before making a decision.
But not all customers are alike. Their intent typically falls into three categories:
Rational Intent: These methodical researchers seek the best solution. They identify challenges, explore options, and build a business case internally. They compare features, read whitepapers, and ask detailed questions. To engage them, position yourself as a trusted advisor—share in-depth content, case studies, and insights to help them navigate internal approvals. They aren't swayed by aggressive sales tactics; they need a reliable partner to guide them toward the right decision.
Urgent Need: These buyers are driven by time-sensitive challenges—such as a new regulation, competitive pressure, or an internal crisis. Their evaluation process is fast and unstructured. They prioritize quick solutions, clear answers, and seamless implementation. Speed is crucial: the faster you address their pain points, showcase your solution, and remove barriers, the higher your chances of closing the deal. They don't have time for long pitches—they need actionable solutions now.
Fake Priority: These prospects show initial enthusiasm—requesting demos and pricing—but lack the internal drive or ROI justification to proceed. Their interest often stems from executive curiosity or internal brainstorming. However, without a clear champion or timeline, their intent is fleeting. Qualification is essential: ask what problem they’re solving, who’s driving the initiative, and their decision-making timeline. If responses are vague or inconsistent, focus on prospects with stronger buying signals.
Recognizing these different types of intent isn’t just about categorization—it’s about knowing how to engage when to push forward, and when to step back. Engagement stages ensure that sales and marketing efforts align with actual buying interest rather than just static ICP fit. They also facilitate the establishment of rules for engagement and the handoff from marketing to sales.
Within each account tier, look at the engagement stages to figure out how to prioritize where marketing and sales should focus efforts.
Here is some advice from Imran at Syft Data.
Combine Signals: Don’t analyze an identified visit in isolation. The real power comes from combining firmographic, behavioral, and intent data into a lead-scoring model. At Syft, we use a combination of on-site behavior and off-site intent to predict when a prospect is “in the market.” The more signals align, the stronger the buying intent.
Act Fast: Intent goes cold quickly. Set up real-time alerts for key events—if a visitor from a target account spends time on your pricing or demo page, notify the responsible sales rep immediately to engage while interest is high.
Stack your data providers: No single deanonymization provider is perfect. Using multiple sources improves match rates, but accuracy remains challenging since providers pull from different datasets. Combining them helps increase reliability.
Ensure Privacy and Compliance in Use: If a user opts out of tracking, your de-anonymization script must honor that. And when reaching out, don’t be overly intrusive—never say, “We saw you visited our site three times.” That’s a surefire way to erode trust (or worse, trigger complaints).
That being said, it's important to approach the data thoughtfully. Avoid diving in headfirst by creating campaigns for every website visit from your ICP you can track. Start by identifying the most promising opportunities, then test and scale those that show potential.
I hope this text clarifies the accuracy levels the technology behind it, and whether the technology is worth its hype.
Btw, Imran and the team at Syft Data are continuously benchmarking the accuracy of the major providers. Check this out!