Performica builds a social graph of a company's workforce by monitoring communication patterns, then predicts attrition risk: when one employee leaves, their closest colleagues are six times more.
ENTRY ANGLES
Build attrition-prediction tool as entry point, expanding to engagement and organizational design from social graph foundation · Internal tool approach starting within large companies with attrition problems to gain paying customers and live datasets · Social graph infrastructure as technical moat supporting multiple product layers
VERTICALS
CAPABILITIES
Social graph infrastructure development and modeling, Attrition prediction and workforce analytics algorithms, Longitudinal data collection and analysis
When one employee quits, the ripple effect inside a company is far larger than most HR teams account for. Performica's founding insight is that a departure makes the quitter's closest colleagues six times more likely to consider leaving – and that probability cascades outward through the organization in waves. One exit, left unmanaged, can trigger a chain reaction that empties entire teams.
Performica works by connecting to a company's existing communication and HR infrastructure – email, Slack, internal tools – and building a continuously updated social graph: who works closely with whom, measured by actual interaction frequency rather than org-chart proximity. When the HR system logs a resignation, Performica automatically identifies the colleagues most likely to be destabilized by the news and surfaces a recommended action list for managers and HR – ranging from compensation adjustments to targeted conversations, drawn from a library of best-practice retention playbooks that an AI layer tailors to each individual's role and tenure.
The platform claims an average 15% reduction in attrition – which translates to a 525% ROI given what turnover actually costs. For a 2,000-person company losing roughly $24M annually to attrition, preventing just six additional departures covers the platform's annual fee. Pricing works out to about $15 per employee per month at that scale.
Performica was originally built as an internal tool at genomics company Invitae before spinning out as an independent startup in 2022. It has since raised $3M in its first outside round.
The HR software market has split cleanly into two camps: tools for finding and hiring talent, and tools for keeping the talent you already have. Performica belongs firmly to the second category – which is the faster-growing one as the cost of replacing skilled employees increasingly outpaces the cost of retaining them.
The retention space has attracted a diverse set of approaches. TeamRaderie and similar platforms focus on social bonding and corporate events. LearnLux and Origin address financial stress, working on the thesis that an employee with money problems is a distracted employee. Keep Financial goes further, offering employer-funded loans repaid through tenure. WorkHound captures real-time sentiment to catch conflicts before they escalate.
What distinguishes Performica is the depth of signal and the specificity of intervention. Most retention tools operate on surveys or self-reported data. A social graph built from actual communication patterns is harder to game and more predictive – it doesn't rely on employees accurately describing their engagement level, it observes it.
The social graph is also a platform with legs beyond its initial use case. Isolated employees – those with few connections in the graph – show elevated attrition risk even without a triggering event. Strong cross-departmental connections can identify candidates for lateral moves or cross-functional teams. The data that makes retention prediction possible is the same data that could inform career pathing, knowledge transfer planning, and succession management.
The startup's origin story as an internal product is also instructive. The standard test for whether an internal tool is worth productizing is simple: does our problem exist at other companies, and do we have a meaningfully better solution than what's on the market? Performica passed that test, which is why it became a standalone business rather than a cost center.
Building for employee retention is now a credible market, not a theoretical one. Companies have demonstrated willingness to pay for tools that demonstrably reduce attrition, and the talent shortage that made retention economics so compelling hasn't resolved.
Starting with a social graph is a legitimate technical foundation because the same infrastructure supports multiple product layers. The graph itself is the moat – it takes time to build, reflects actual behavior rather than stated preferences, and becomes more accurate with longitudinal data. A new entrant could begin with an attrition-prediction use case and expand into engagement, organizational design, or internal mobility from the same data substrate.
The analogy to how Performica itself was built is the most actionable entry point: if you're at a company large enough to have a real attrition problem, the internal tool path gives you a paying first customer, a live dataset, and proof of value before you write a single sales deck.