Confirm Launches Auto-Calibration for Employee Performance Reviews

Confirm Saving Enterprises Thousands of Hours and Making Employee Ratings Fairer using Organizational Network Analysis and Auto-Calibration Feature


Confirm, the first platform to inject science into performance reviews to ensure job advancement is based on data, announced today that it has launched a new feature which is the first and only solution to make performance review calibration processes fairer and dramatically faster.

Traditionally companies spend thousands of hours in performance calibration sessions so managers can discuss, compare, and reconcile their employee ratings with one another. The theoretical goal of calibration is to create consistency across the organization, but often these sessions, which can take weeks or months to complete, become political with the loudest voice or most senior manager determining outcomes of employees’ ratings vs. the employees’ actual impact. This process, ironically, introduces bias into performance reviews and can paralyze teams because of the lengthy process of discussing each individual employee.

To cut down the painstaking process, Confirm has built a new feature, “Auto-calibration”, which reduces calibrations to just a week, or even a few days, saving roughly 50 -75 percent of the time calibration can take. After employees submit their reviews, Confirm Auto-calibration runs a series of checks on each manager rating and points out the subset of employees the company should focus calibration discussions around. These checks are a combination of a rule set defined by the customer and what Confirm learns about the employee’s performance through Organizational Network Analysis (ONA).

This allows companies to focus only on the minority of employees whose ratings need attention.

Confirm’s goal is to make performance reviews fairer, inclusive, and data driven. Businesses using Confirm Auto-calibration can import the rule set they typically apply in calibrations, such as checking for reviewer biases, into the tool. Combined with what Confirm knows about the employee’s performance, the software automatically highlights the handful of employees that should be discussed in calibration.

For example, a manager might have given all high scores to their entire team. Or a manager gave the employee a low rating, however, the employee is viewed as highly influential and impactful by their network. In another example, a newly promoted employee received a top manager rating in their newly promoted role. This might be well deserved, but the data says it’s an abnormally high rating for newly promoted workers at this company. Confirm now automatically highlights these handful of cases so calibrators can ensure these ratings are appropriate versus spending time on every manager rating, the majority of which often do not change.

Confirm’s first-of-kind ONA approach to employee performance reviews considers the network model of work where all employees can weigh in on who at their company they turn to for advice and help, and who they believe is doing outstanding work, so analysis can spot correlations and trends and identify the company’s true top performers. Confirm’s ONA-based reviews create a fairer playing field by ensuring performance reviews are based on larger sets of data rather than only the opinions of a single manager.

“Confirm’s mission is to rehaul the broken and biased performance review process that is the norm at most companies,” said Josh Merrill, Co-founder and CEO of Confirm. “We want to make the entire process easier and less time consuming for managers; software and automation now makes that possible. With Confirm Auto-calibration, we are introducing a first for the industry that will save companies thousands of hours. Most importantly, this feature will help quell the loudest voice in the room phenomenon, which means a fairer performance review process for all.”

Discover the full potential of your Hrtech strategy with our comprehensive Hrtech News and Hrtech Interviews.


Please enter your comment!
Please enter your name here