Cursor’s head of talent says companies get one crucial thing wrong with work trials
Cursor's head of talent, Adam Ward, said work trials, while expensive, offer technical signals and show how candidates collaborate and fit.
Adam Ward, Cursor's head of talent, contends that companies often err in their approach to work trials. According to Ward, a common mistake is making the trial overly focused on extracting information from the candidate. He explained that this mindset leads to an "us-versus-them" dynamic, where the company is perceived as assessing and extracting information, while the candidate receives little in return.
Ward advocates for a two-way approach, where candidates work alongside employees on a project. This method should feel more collaborative, with careful consideration of who the candidate interacts with, from the initial greeting to who they sit next to during lunch. By controlling these aspects, Ward believes it ensures the candidate has the necessary information to make an informed decision and provides valuable feedback to staff on the candidate.
Cursor, acquired by SpaceX for $60 billion, maintains a commitment to on-sites despite their expense and the significant ask they place on candidates. Ward maintains that on-sites provide a more comprehensive view of a potential hire compared to traditional interviews, sending a strong message to both the candidate and the company.
The emphasis on work trials comes at a time when competition for top AI talent is intense. Ward emphasized that recruiting at Cursor is a company-wide effort, focusing on identifying a small pool of exceptional candidates and nurturing them over time, rather than filtering a large number of applicants through a standard interview process.
Cursor's CEO, Michael Truell, corroborates this approach, stating that the company goes to great lengths to recruit top talent, even reaching out to engineers who had previously declined and employing creative methods to revive conversations.
Written by urgent.news from Business Insider's reporting — not their text. Machine-written — it may contain errors, so check the original before relying on it.