New AI feature unlocks work-from-home B2B visitor data

Written by Dr. Elena Drozd, Bombora SVP Data Science
January 24, 2025

B2B sales and marketing teams benefit significantly from knowing when visitors from their target companies interact with their web content. But when it comes to determining who’s visiting websites, there’s a problem that’s been difficult to solve – identifying professionals who don’t work from a corporate office location.

During and after Covid-19 , the way in which employees “go to work” changed dramatically. For years, offices remained empty, and some businesses went permanently virtual. Between 2019 and 2021, the number of people working from home tripled – from 9 to 27.6 million individuals. ¹

Even after the Covid pandemic, work-from-home is going strong. A 2023 Pew Research Center study found that 35% of workers with jobs that can be done remotely are working from home all the time. ² While more companies are now encouraging onsite attendance, 53% of U.S. workers still report they’re working hybrid in 2024. ³

Missing B2B web traffic identification

With over a half of U.S. employees doing work from a place other than a business address, marketers are struggling to track the companies where their anonymous website visitors work. That’s because the IP address in a person’s home will show up as owned by an internet provider like Comcast, Verizon, AT&T, Xfinity, etc . The result is that marketers have been missing millions of visitors in their B2B web traffic. That’s a lot of powerful data that can help direct marketing, sales, and content strategy.

Another issue is that in many homes the internet is accessed by several household members, including children. In 2021, 97% of 3- to 18-year-olds had home internet access. ⁴ In addition, it’s very common for multiple household members to share the same location while working for different companies. Filtering out the noisy data of household internet use has been a struggle as marketers seek to look for quality Intent signals from their ICP.

For employees who use the WiFi in cafes and co-working spaces, the IP address of the business will be associated with site vistors, further thwarting efforts to glean insights from B2B web traffic.

Bombora’s AI model cracks the work-from-home code

Today, the ability to resolve identity in the remote world is vital to provide businesses with a full picture of their website visitor data. Starting in early 2020 Bombora has been responding to customer needs by building an AI model incorporated into our Visitor Insights solutions that could recognize when employees of large, small, and medium-sized corporations work from home .

This best-in-class AI technology uses behavioral modeling powered by Bombora’s Data Co-Op and robust deterministic traffic signals. Our proprietary IP-to-company data, used in combination with external vendors, feeds a composition-based model to maximize both reach and accuracy of identity resolution. Today, our AI solution has achieved high reliability and operates continuously behind the scenes, delivering large-scale visitor data resolution to our clients while maintaining exceptional quality.

Neural networks contextualize traffic and resolve identity

Identifying web traffic as originating from a company employee rather than an unrelated individual is a complex challenge. To solve this, we utilize signals from our proprietary Data Co-op alongside a custom deep learning model to resolve anonymous web traffic back to specific companies . Our new deep learning AI model connects multiple anonymous profiles and generates a densely connected graph of events that link unknown visits to known ones.

Our neural network models, which have been evolving and learning for over three years, make high-quality determinations about the organizations engaging with our customers’ websites. With these mature models, our customers can now experience substantial benefits through more accurate visitor identification.

The future of work and B2B Intent signals

Despite highly publicized return-to-office mandates, many business leaders expect the remote work trend to continue and even increase. A Harvard Business Review report shared that U.S. executives expect hybrid work to increase 15% between 2023 and 2028, with remote work expected to increase about 10% in the same period. ⁵

For sales and marketing leaders, the remote work trend has been a source of frustration in their visitor tracking efforts. But with the new AI features in Bombora’s Visitor Insights solution, Intent data can be gleaned from B2B web visitors no matter where they get their work done.