The average cost of losing an associate who joined as an articling student is somewhere between $300,000 and $500,000. They tend to leave right around year four or five, exactly when they start becoming profitable. A significant driver of that turnover is management behavior that many firms have not figured out how to measure or reward.

That was the thread running through the session Sandra Bekhor of Bekhor Management and I hosted on June 25. See the recording here. Sandra has spent more than 20 years coaching lawyers on leadership and team management, with a particular focus on delegation and feedback. I came at it from the profitability side. What kept coming back is how directly management habits shape firm economics, and how rarely firms treat the two as connected.

Why Managing a Legal Team Is So Hard

We started with what most managing partners already know but rarely name directly: lawyers are not trained to manage people. Law school trains the critical, analytical mind. It rewards the lawyer who finds the flaw in an argument and who assesses risk before anyone else in the room does. Those skills are essential for legal work. In a management conversation, the instinct to find the flaw is usually the thing that gets in the way.

Sandra made the point that good management requires a different part of the mind: empathy and the ability to listen without immediately judging. Lawyers have those capacities. They are just not the ones that get reinforced in legal training or firm culture.

So when lawyers move into senior roles, they are expected to lead and develop their people, without much preparation for either. And without strong role models in the firm to learn from, most are expected to figure it out on their own, which usually means they do not.

The Cost of Withheld Feedback

One of the patterns Sandra sees consistently: litigators who will argue fearlessly in court will not give a direct piece of critical feedback to an associate.

Her explanation made sense to me. In court, everything is rule-bound. The norms for how to behave are clear, and following them is not seen as aggressive. In a feedback conversation, there is no rulebook. The associate is someone you work beside every day. So the conversation gets deferred.

The cost compounds. An associate who makes a mistake in January and does not hear about it will repeat it in February, March, and April. By year-end review, the associate is hearing critical feedback for the first time, on twelve months of work. That is not just a development failure. It is a profitability problem. You have a person billing on your behalf all year at a fraction of what they could be producing.

One point Sandra raised that I do not think gets enough attention: your strongest performers want feedback just as much as anyone who is struggling, if not more. High performers are driven to grow. If your firm is not part of that growth, they will find one that is. They tend to leave right around year four or five. That is exactly when your investment in them starts to pay off.

Why Delegation Does Not Happen

This is where the financial case becomes very direct. Work that goes to a $350 associate should not be sitting with a partner billing at $800 or $900 an hour. The billing math here is clear, even if the behavior rarely reflects it.

Sandra identified fear of losing control as the primary obstacle. That fear is legitimate and takes different forms. Some partners worry about their reputation if a file goes wrong. Some fear client complaints. Some are simply convinced the work will come back needing to be redone anyway.

I added the structural piece. In many firms, partner compensation is tied primarily to billable hours. A partner who delegates aggressively is reducing her own hours. Her compensation reflects that. So she keeps the work. That is a rational response to the system in front of her, but it works directly against your firm’s long-term profitability.

Sandra’s reframe on this: partners who invest in training their team lose some billable hours in the short term. But if the team develops and stays, profitability rises over time. In my experience, the partners who invest in their people tend to build the more profitable practices.

What to Do About It

The practical suggestions we landed on were straightforward.

At the start of each year, each associate identifies two or three personal goals in conversation with the supervising partner. Specific goals tied to that person’s career trajectory, not pulled from a firm-wide checklist. Then a check-in every 90 days. This gives the partner a natural framework for feedback that does not feel adversarial, and it signals to the associate that someone is paying attention. That signal matters more than most firms realize.

Sandra and I both referenced the Balanced Scorecard as a structural tool worth considering. The idea is to measure partner performance beyond billing metrics. Financial results matter, but so does how well partners are developing their teams and building capacity for the long term. For a managing partner compensated partly against a balanced scorecard, the non-billable work of developing people gets counted. That changes behavior.

The harder question, which an audience member put directly, is how you make management changes stick when partners are still paid primarily on billable hours. The honest answer is that you probably cannot, not at scale and not over time. Some portion of partner compensation needs to reflect how well partners are managing and delegating. Even five or ten percent signals that the firm takes it seriously. Setting those goals at the start of the year and reviewing them at the 90-day mark gives it structure.

The AI Question

Near the end of the session, an audience member raised a specific question about AI and delegation: if junior work is being replaced by AI, what do you do with your junior associates and paralegals?

One view: you retrain them as validators. Have AI take the first pass. Have the junior associate review the output and produce a corrected version, then send that to a senior associate or partner for a final pass. The skills shift from producing the work to interrogating it. That is a legitimate developmental path, and some firms are already building it into how they count billable hours.

Sandra’s point: whatever structure you build around AI tools, the management principles for developing your people hold.

None of what Sandra and I described is complicated. What makes it hard is that many small and midsize firm partner compensation systems actively reward the opposite behavior. If you want to think through what changing that looks like in your firm, reach out.