I made the call in the parking lot of my Oak Cliff location on a Thursday in October. I had driven there to handle a scheduling mess that my manager had escalated to me for the fourth week in a row. I sat in my car for twenty minutes before going in, and in those twenty minutes I wrote a text to my husband that said “I think I have to let Rachel go.” I had known for six months. The parking lot was just where I finally admitted it.
Rachel was the first manager I ever hired. She had opened location two with me, picked out the back-bar shelving, trained the first four stylists we put on that floor. I promoted her because she was there from day one and I thought loyalty would translate into leadership. It did not. She was warm and the team liked her, but she could not make a hard decision to save her life, and a second-location manager who cannot make hard decisions is not a manager. She is a very expensive friend.
This is the story of firing her. It took me eighteen months longer than it should have, cost me two good stylists and roughly $41,000, and it is the single most important thing I have learned about scaling.
When to fire a salon manager (and when you actually admit it)
Looking back, I can tell you the exact week I knew Rachel was not going to work out. Month seven. One of our colorists, Janae, came to me directly instead of to Rachel with a schedule conflict, and when I asked why, she said “because Rachel will just tell me to figure it out with the team.” That is not a manager. That is a person with a title.
But I did not act for another eleven months. Here is why, in the order the excuses came:
- She had opened the location with me. Firing her felt like erasing history.
- The team loved her personally. I was afraid they would quit in protest.
- I had never fired a manager before. I did not know how.
- I kept telling myself she could grow into it with more coaching.
- I did not have a replacement lined up, and the thought of managing Oak Cliff myself again terrified me.
None of these were business reasons. All of them were avoidance dressed up as strategy. A Harvard Business Review piece on firing decisions puts it plainly: most leaders who regret termination decisions regret waiting, not acting. I fit that description exactly.
⚠️ The month-seven rule
If you have a manager and by month seven your team is still routing decisions around them to get to you, you do not have a manager problem. You have a confirmation problem. You already know. You are waiting for permission.
The real cost of keeping an underperforming salon manager
I want to name the numbers because growth content loves to skip this part. The cost of keeping the wrong manager in place for eleven extra months was not just her salary. It was everything that leaked out around her.
Two stylists gave notice during those eleven months, and in their exit conversations both of them named the same thing: they did not feel like anyone was actually running the floor. Replacing a stylist runs about 150% of annual wages when you factor in recruiting, training, and the production gap, according to SHRM’s 2025 turnover data referenced across industry research. At our wage bands, that was roughly $28,000 across the two of them. Oak Cliff revenue dipped about 9% across Q2 and Q3 because we were understaffed and my best performers were covering for the schedule chaos. And I paid myself nothing extra for the eight weeks I spent personally running that location after Rachel left. That is the number I still flinch at.
The actual replacement cost, per SHRM guidance that runs mid-level replacements at six to nine months of salary, added another piece on top.
The version of the story where I fire Rachel at month eight? I keep both those stylists. Q2 revenue stays flat. I do not lose my summer.
The eighteen-month timeline, honestly
Month 1 (hire)
Promoted Rachel to manage Oak Cliff on opening day. No formal job description. Title was the onboarding.
Month 7 (first red flag)
Janae routes around her for a schedule decision. I notice. I do nothing.
Month 10 (pattern forms)
Three of seven stylists are now bringing issues directly to me. I tell myself I need to coach her better.
Month 14 (first departure)
A senior colorist gives two weeks. In the exit conversation she tells me Oak Cliff has no leadership. I cry on the drive home and still do nothing.
Month 16 (PIP I should have opened at month 8)
I finally write a performance improvement plan with three specific metrics. I give her ninety days. She does not hit any of them.
Month 18 (the parking lot)
Thursday in October. I sit in the car. I write the text. I go in and I schedule the conversation for the end of the next day.
Month 19 (the firing)
Twenty-minute conversation, documented, with a severance package. She cried. I cried. The team did not quit in protest. Two of them told me later they had been waiting for me to do it.
The part of that timeline that still bothers me is month sixteen. By then I was not coaching her. I was building a paper trail. A performance improvement plan is a legitimate tool when you believe someone can improve. When you are running one as documentation for a firing you have already decided on, you are running theater dressed as coaching, and the person on the other side can usually tell. Rachel told me later she knew by week two of the PIP. I had wasted another ninety days of her life and mine pretending.
What I do now before promoting a manager
The systems I built after Rachel are covered in a separate piece on promoting stylists into manager roles, but the short version is this: I now run a twelve-week shadow period before any manager promotion, with three specific tests built into it. Can the candidate deliver a hard piece of feedback to a peer in my presence without sugar-coating it. Can they build and defend a weekly schedule that covers a callout scenario I hand them. Can they have a money conversation. Rachel would have failed all three. I just never asked her to try.
I also changed how I run one-on-ones with my managers. Every two weeks, thirty minutes, same agenda: team, clients, money, you. Same four prompts, every time. If a manager cannot give me something concrete under each prompt, I now know in sixty days, not sixteen months.
The lesson I resisted hardest
The story I had told myself about scaling was that a good owner stays loyal to the people who helped her build the thing. Rachel and I opened Oak Cliff together. She took a cut to move into management. She believed in the vision. All of that was true, and none of it made her the right person to run a location at month eighteen.
Loyalty is owed to people. Titles are owed to performance. I had conflated the two, and the cost of that confusion was a location that underperformed for a year, two stylists who deserved better, and a friend I could have kept longer if I had treated the management role with the seriousness it required. I fired her late, and because I fired her late, I lost her as a friend too. She came back six months later to do her own hair. Not as a stylist at my salon. As a client paying full price. That is the version of the story I keep coming back to.
What I would tell you if you are sitting on this decision
If you have been rereading this article because you have a Rachel, here is the short form. You know. Write down what you know. Give it a specific timeline with specific metrics. If you are already twelve weeks into a PIP that is not working, you are not coaching anymore. You are avoiding. The longer you wait, the more the cost moves from your manager to the team around her, and that is the cost that does not come back.
Rachel is fine now. She went back to cutting hair, which she loved and was great at, at a salon across town. Oak Cliff runs better than it did when she was there. And I still drive past that parking lot on the way to my Uptown location and think about the twenty minutes I spent in that car. It took me eighteen months to have a conversation that lasted twenty minutes. That is the math of letting someone go too late.
