YesPress Profile - Company Edition
Affinity.
A logo, a network, and one quietly subversive idea: the CRM should fill itself in.
A logo, a network, and one quietly subversive idea: the CRM should fill itself in.
It is a Tuesday morning in a Sand Hill Road conference room. A partner opens her laptop, types nothing, and somehow knows the entire history of the founder she is about to meet - who introduced them, how warm the connection is, which colleague last exchanged emails with the company, and what was said. The intelligence did not arrive by magic. It arrived by Affinity.
Affinity is a San Francisco software company that has spent the last decade trying to fix a problem nobody much enjoyed talking about: dealmakers hated their CRM. They hated typing into it. They hated forgetting to. They hated that the most valuable thing in their business - the relationship - was the one thing the software did the worst job tracking. So Affinity built a different kind of CRM, one that watches the inbox and the calendar and quietly does the typing for you.
Today the company employs about 290 people, counts more than three thousand organizations as customers across sixty-five countries, and runs the relationship graphs behind some of the best-known firms in venture capital, private equity, and investment banking. In March 2026 it brought in a new chief executive, Ken Fine, while co-founders Ray Zhou and Shubham Goel stepped onto the board. That is the snapshot. Now the story.
A pull quote, the way pull quotes used to look. Yellow because nothing about this idea is shy.
If you have ever sold software to a venture capital firm, you know the joke. Partners buy the seat. Analysts maintain the data. Nobody trusts the result. Salesforce, in its many tailored shapes, was built for transactional pipelines - leads in, customers out, predictable forecasts. Private capital does not work that way. A deal can be ten years in the making. The relationship that closes it might be a college roommate, a board referral, or a polite reply at a Burning Man camp. Try logging that.
Zhou and Goel, who had been computer science students at Stanford and roommates before they were co-founders, kept noticing the same gap in conversation after conversation. There was software for the transaction. There was not software for the journey. Bankers kept their relationships in their heads. Partners kept them in their assistants. The data, if it existed, lived in a thousand inboxes that nobody could legally read at once.
The least glamorous origin story in venture capital: two roommates, one missing feature.
The bet was simple to describe and unreasonable to execute. If you could capture every email, calendar invite, and meeting across an entire firm and you could do it cleanly, with permission, and you could turn the resulting fog of metadata into a structured graph of people, companies, and relationships, then you could give dealmakers something they had never had: an honest picture of their own network. You could tell them who at the firm knew the CEO they were chasing. You could score how warm that connection actually was. You could surface the introduction nobody remembered they could make.
None of this is easy. The graph is enormous. The privacy work is unglamorous. And the customer is famously skeptical of anyone selling them software. Affinity took the slow road, which turns out to be the fast one. It shipped, listened, shipped again. It hired sellers who had sat on the other side of the table. It made the platform good enough that customers started showing up to invest.
When a vendor's cap table starts looking like a customer list, something is working.
Affinity sells a CRM, but calling it a CRM is like calling a self-driving car a sedan. The platform structures and analyzes more than a billion data points across emails, calendars, and third-party data sources. From those points it builds a living map of who at your firm knows whom, how well, how recently, and through what mutual connections. The interface looks reassuringly like a pipeline. The intelligence underneath is the difference.
The relationship intelligence platform built for private capital. No typing required.
Drops the relationship graph into the Salesforce instance your firm already runs.
Benchmarks your deal flow against the rest of the market, before the partner meeting.
Automated enrichment that fills in the company and contact fields no one wanted to.
Launched May 2025. AI-driven deal discovery aimed at finding the company before your rivals do.
Five products, one premise. Stop asking humans to do what software can watch.
Ten years, four rounds, one CEO transition, and a lot of inbox parsing in between.
Here is what makes the story credible. The customers are not a parade of unknown logos. Bain Capital Ventures uses Affinity. Kleiner Perkins uses Affinity. SoftBank Group uses Affinity. So do Fidelity Investments, Woodside Capital Partners, Tishman Speyer, American Family Insurance, and, on the enterprise side, Nike, Qualcomm, and Twilio. When the firms with the most relationships entrust their relationship graph to a vendor, that is a useful market signal.
Bar charts are a love language. These ones speak fluent dealmaker.
Ray Zhou and Shubham Goel did not invent the CRM. They invented the part of it nobody types in.
Affinity describes itself as a relationship intelligence platform. The phrase is doing a lot of work. Strip the marketing and what is left is this: the company believes that the most consequential information at a deal-driven firm lives in the gaps between the org chart - the messy human edges where one person knows another well enough to vouch for them. Affinity's mission is to make those edges visible, durable, and usable by every member of a team, not just the partner with the longest contact list.
That mission has practical consequences. It is why the product captures activity automatically rather than asking users to log it. It is why the platform measures relationship strength as a score rather than a boolean. It is why the company invests as heavily in privacy and permissioning as it does in features. A relationship graph without trust is a liability. Affinity is trying to build one that is an asset.
The difference between a hierarchy and a human network, summarized.
The wider story here is that AI changed the calculus of what software is allowed to know. Five years ago, asking a vendor to read every email in a firm in order to score relationship strength would have been a difficult sale. Today, with large language models doing serious work on enterprise data and dealmakers under fresh pressure to find returns in a tighter market, it is increasingly the default. Affinity Sourcing, launched in 2025, points where the platform is heading: not just remembering who you know, but suggesting who you ought to know next.
The competitive field has noticed. Salesforce continues to push deeper into vertical CRM. DealCloud, 4Degrees, Attio, and a parade of startups are angling for adjacent slices of relationship intelligence. None of them, yet, has the customer roster Affinity has assembled. None of them has the data graph. The question for the next chapter is not whether relationship intelligence becomes a category. It already has. The question is who runs it.
Zhou and Goel made the list. The product made the case.
Set in 2021. The bar has likely only moved up.
The firm leading was also a customer first. Twice the conviction.
Relationship intelligence travels well. Borders, less so.
Four sticky notes for the founders' wall, free of charge.
The partner finishes her meeting. She steps out of the room into the long California afternoon, and on her laptop a new thread has already begun. A colleague three floors down, she now knows, sat next to the founder's chief of staff at a dinner last September. The intro will write itself. The deal will not, but the path is suddenly shorter by a week. That is the small, quiet, almost boring promise of the company - not that it will close the deal for you, but that it will stop you from missing the one you were always going to close, if only somebody had remembered.
Affinity remembered.