How Kernel works
Introduction
Kernel works differently from traditional data vendors and credit bureaus (see Kernel vs. traditional data vendors and credit bureaus).
Instead of maintaining a static database, Kernel gathers and analyzes company information in real-time as it is needed. This "just-in-time" process involves collecting data from a variety of sources, including websites, news articles, and public filings. This contextual information helps ensure the account data in your CRM is accurate, comprehensive, and customized to your needs.
The process begins by connecting to your CRM to analyze account information (see Salesforce integration).
Kernel's AI-powered modules then identify and correct inconsistent data, find and merge duplicates, and uncover corporate parent-child relationships. To protect your master data, all data is written into separate, custom objects within your CRM - see Deployment.
Kernel vs. traditional data vendors and credit bureaus
The following table compares Kernel's approach to that of a traditional data vendor or credit bureau. The comparison below is for the credit bureau, Dun & Bradstreet.
Data Structure
Uses a flexible data structure centered on the trade names and domains that sales reps use in CRMs.
Focuses narrowly on legal entity names and countries, with each record assigned a D-U-N-S number. This can result in the creation of duplicate entries in CRMs.
Data Collection & Analysis
Collects data in real-time from across the web, including unstructured sources like news, websites, and PDFs, to support AI-driven analysis. This allows for an understanding of complex situations where internal CRM data may be contradictory.
Relies on a static database of information. Changes to company addresses or ownership can take months to be reflected.
Account Mapping
Employs intelligent mapping that considers all available data, including account names, websites, and contextual information from opportunities, to match accounts as they are recognized by sales reps.
Maps accounts based on string-matching legal entity names and country. This can lead to separate entries for entities like "GE Healthcare" and "GE Healthcare Ltd."
Customization & Flexibility
Works with you to customize business logic and handle unique situations, such as handling accounts used for transactional/billing purposes.
Hierarchies are largely fixed. Customizations require a paid service with a lead time of 4-6 weeks and are subject to API rate limits.
Support
Provides a dedicated Customer Success Manager, a shared Slack channel, and a 48-hour service level agreement for acting on feedback.
Customer support is limited.
Understanding Our Data Matching: Kernel vs. D&B
To ensure the highest quality data in your CRM, it's important to understand how we match and enrich your records. Our approach is fundamentally different from traditional models, such as Dun & Bradstreet's DUNSRight, resulting in more accurate and reliable data.
The Limitations of Traditional Matching: D&B's DUNSRight
Traditional data providers like Dun & Bradstreet often rely on a limited set of data points for matching companies. D&B's DUNSRight model, for example, primarily uses a company's name and country to find a match. This can lead to several issues:
Inaccurate Matches: With only a name and country, it's easy to confuse companies with similar names.
Lack of Transparency: D&B uses a "sliding scale" of confidence scores (e.g., a score of '6'). It's not clear what these scores mean or how they are determined, making it difficult to trust the match.
Lower Match Rates: If a company isn't in their existing database, they can't provide a match.
The Kernel Approach: A Modern, Data-Driven Model
Kernel takes a more comprehensive approach to data matching, which results in a higher degree of accuracy and a more reliable dataset for your team. Here's what makes our model different:
Richer data for better matches: We don't just rely on name and country. The company website is a key data point for us, along with any other available information in the account object. This allows us to create a more complete picture of the company and ensure we're matching the right one.
Transparent reasoning: We can provide a clear reasoning behind our matches, so you can have confidence in the data.
Higher match rate through proactive data collection: If a company isn't in our system, our work doesn't stop there. We proactively crawl all available data, especially from the company's website and other online sources, to gather the information needed to make an accurate match.
Feedback loop: We have a feedback field directly into your CRM. If you ever disagree with a match, you can let us know with a single click. Our team will then review and fix the match within 48 hours, ensuring our data is always improving.
Last updated