How Managers Restore Accountability Without Losing Trust
Accountability can slip when expectations are unclear, systems break down, or follow-up fades. Managers can restore performance without damaging trust by focusing on facts, shared solutions, and clear next steps. Insights from experts in the field show how to reset standards, address obstacles, and keep progress on track.
- Debug the System, Then Assign Fixes
- Ask Early, Document the Standard, Follow Up
- Set Time-Limited Daily Check-Ins
- Lead With Curiosity and Clarify Priorities
- Fix Measurement Before You Judge Performance
- Let Employees Own Action Items
- Rebuild the Process Side by Side
- Build Shared Checklists After Setbacks
- Invite Self-Assessment Before Expectations
- Open With What Changed, Then Set a Date
- Call Before You Comment
- Examine Conditions, Then Remove Friction
- Hear the Reason, Then Confirm the Bar
- Reset Expectations With Visible Near-Term Action
- Walk the Job, Then Diagnose the Miss
- Guide Self-Recognition and Acknowledge Progress
- Trace Workflow Gaps Through the Run of Show
- Center the Child Experience, Then Co-Create Recovery
- Name Obstacles and Redistribute Work
- Review Cases and Rebuild Standards
- Use Facts to Build a Two-Week Reset
- Define Proof and Track One Commitment
- Schedule Early Touchpoints to Prevent Breakdowns
- Use Data to Solve Problems Together
- Recall Successes and Set Three Goals
Debug the System, Then Assign Fixes
1. Tell them early and tell them plainly. I have a six week window to mention a problem to a trading performer before it hardens into a massive blob of frustration that I don't want to deliver to them in one excoriating blast. Thus I treat a slipping trading performer just as I would a degraded server - first I pull the logs, then we rip it apart to find the problem. Before I speak to anyone at TradingFXVPS about this, I'll first draw a timeline of what happened. As I noted earlier, one of our senior support engineers had his average time to close an escalation go from 4 hours to 11 over the past week or so. And interestingly enough, that slide began exactly when we migrated our last batch of broker clients to our new MT5 infrastructure.
2. We run post-mortems on our work (whether it fails or not) just as we do for outages, listing off the facts in the following order: what changed, when it changed, what happened as a result of said change, how one can fix said problems, and who is going to implement said fixes by when. As I described for the Senior Support Engineer who's had his performance drop to 9 in the last 11 hours, his problems are obvious when laid out in the correct order. The Senior was the sole contact point for the 100+ MTS servers for the 100 or so accounts that were migrated to new MT5 infrastructure for our broker clients and his queue had started to rot while he was the only one able to answer the slew of questions that the newly-migrated accounts were having with the MTS servers in the legacy stack that he was the only one to have fully worked with prior to this. The 'fix' here was mechanical in nature: document the legacy stack as the Senior had; distribute the questioning of newly-migrated accounts to two other team members; and give the Senior 14 days to bring his numbers back up to 18. The Senior was able to accomplish this in 9 days. The key to trust in a post-mortem is to frame it so that you're debugging the system to find the root cause(s) of a given problem and not putting the individual on trial for said problem(s). As long as the fixes identified have a named accountable to complete each fix by a set date, all will be well. A 'you seem off lately' type of comment, however, is not something that can be fixed by the individual to whom said comment was directed. Nobody can fix 'off'.
Ask Early, Document the Standard, Follow Up
When a capable employee starts slipping, my first assumption is that something changed — because capable people do not usually just stop caring. In a gym, it looks like this: a trainer who was always early starts arriving right at class time, follow-ups that used to happen the same day start sliding, and the energy members used to comment on is gone. The conversation I have is early, private, and specific. I name exactly what I have seen change, with dates, not a general 'your attitude has slipped.' Then I ask one question and stop talking: 'What changed?' Sometimes it is a schedule problem, sometimes something at home, sometimes they are quietly carrying a workload issue I created. Whatever it is, we agree on what good looks like again, in writing, and on a short check-in date — two weeks, not someday. Accountability is preserved because the standard never moved and the follow-up is real. Trust is preserved because the employee got a conversation before they got a reputation. Wait until you are frustrated, and you are no longer fixing a slip. You are building a case, and the employee can tell the difference.
Set Time-Limited Daily Check-Ins
When a capable person's work starts slipping, I set up a short, fixed window of check-ins instead of a big sit-down talk. I say, "Let's do 10 minutes every morning for two weeks, and on day 14 this stops." That end date carries the whole approach. Most check-ins feel like surveillance because nobody knows when they'll end or what earns an exit.
Here's how I run them.
The employee picks the time, and after the first two days they lead the call. We only cover three questions, which are what got done yesterday, what's planned today and what's in the way. I keep notes, but I don't add new tasks during the two weeks.
At the edtech company I co-founded, our strongest marketer had gone quiet, and his campaign launches kept slipping a few days each. On day six, he told me a vendor had stopped answering his emails and he'd been too embarrassed to say so. We swapped the vendor, the check-ins ended on day 14 and his launches went back on schedule. In my experience, an end date tells people you trust them to get back on track.
Lead With Curiosity and Clarify Priorities
When a capable employee's work begins slipping, I start with the Ted Lasso principle as I like to call it. Curiosity first, ahead of correction or anything else. In one case in particular that comes to mind, a previously reliable team member began missing details and deadlines, and it would have been easy to frame the conversation around performance alone. Instead, I opened with what I had observed, asked what had changed, and then separated the issues into capacity, clarity, and accountability. We found that priorities had multiplied without anyone explicitly removing older work and the employee was not comfortable asking for help because they felt it might undermine their credibility. We reset the list, clarified what success looked like, and agreed on a few concrete checkpoints. Performance improved because the employee understood both the support available and the standard they still owned.
Fix Measurement Before You Judge Performance
I cannot give you a conversation script with results behind it. What I can describe is the step before the conversation, which is where we went wrong first.
This summer we built a page for our project managers to use in one-to-ones. For each person it lists the tasks past their due date, sorted by how late they are, with a plain reading: overloaded, working through an old backlog, low activity, or under control. A bare count of late tasks hides the difference. Ten tasks that slipped this week usually mean a heavy week, and the conversation is about load. Ten tasks that have been late for a month mean a pile nobody cleared, and that conversation is about what to drop or hand over.
The first version got it wrong. It labelled one person as critical, meaning urgent work piling up right now. A reviewer who rebuilt every label from the raw data found that almost all of that person's urgent late tasks were more than two weeks old, and only one was recent. It was an old backlog, not a fresh emergency. A manager would have walked into a conversation about the wrong problem. The rule now needs most of the urgent tasks to be recent before it says critical.
A second page, on missing due dates, taught us the other half. Its first headline named one client as the worst and rated everyone on the team weak. Most of what it flagged was internal process work, including copies of our own process documents filed under that client, not client work. After we corrected that, the weak scores still sat across the whole team, which points at the standard we set. The page now presents it as a team signal, not a verdict on one person.
I have no measurement of whether the conversations that followed restored anything. The tool is only a few months old.
Let Employees Own Action Items
If I see an accomplished employee's performance drop, I engage in a discussion within a week without forming my story of why it has happened yet. The discussion starts with a question rather than with a conclusion: "This is what I see, and this is what has changed. What is hindering you?"
The technique that helped bring back accountability to the process was ending every discussion with the employee making a note of the action items and the date when we will check in on them.
In one department, an excellent analyst faced three missed deadlines in one month and became quite defensive. In our first discussion, we found out that two other departments have increased her workload with urgent tasks, which were responsible for most of the delays. We cut down her queue by 30% and agreed on a follow-up. The proportion of timely deliveries grew from 58% to 92% in six weeks, while her engagement level improved from 5.4 to 8.0.
The bottom line: curiosity finds the cause; ownership fixes it.
Rebuild the Process Side by Side
When audit documentation got sloppy, I didn't send a corrective email. I asked the team member to walk through their process alongside our Search GAP standards. We built a new template together in 20 minutes. That's what works with remote teams. Fix the problem side by side instead of handing down instructions. They know their stuff. Give them that credit and things course-correct fast.
Build Shared Checklists After Setbacks
When a batch went bad, I sat with our lead grower. Not to point fingers, but to figure out what warning signs we missed. We built a daily checklist together and she rolled it out to the whole team. Nobody got defensive because the focus was on catching problems earlier, not who dropped the ball. People speak up more when they know it won't be used against them.
Invite Self-Assessment Before Expectations
I ask the person to grade their own last month before I say anything. "If you looked back at the past four weeks, what would you want to have gone differently?"
Nearly everyone names it themselves. Capable people have a sharper read on their own work than managers give them credit for, and hearing it in their words changes the whole conversation, because now we're comparing notes instead of me delivering a verdict.
Speaking of that, I'd like to explain what happens when they don't name it.
That's useful information too. If their read is genuinely positive and mine isn't, the gap is about expectations, not effort, and I've been unclear somewhere. From what I've seen, that's the real cause more often than anyone admits.
The follow-up I use there is "here's what I was expecting, tell me where that got lost." Puts the breakdown between us instead of on them.
Either way we end the same. One change, one date, written down, so neither of us relies on memory a month from now.
Open With What Changed, Then Set a Date
When a capable person starts slipping, I've learned the slip is almost never about ability. It's usually something happening outside the work, or a part of the job that quietly stopped making sense to them. So I don't open with the missed deadline.
I run a small team of about 11, and I do it one on one, usually walking the production floor or grabbing coffee, never in front of anyone. My opening line is something like: "This isn't your usual work, and I'd rather understand what changed than guess." Then I stop talking. The silence does more than any speech. Most of the time they tell me exactly what's in the way, and it's fixable: a bad handoff, an unclear priority, too many things landing on one person.
Then we agree on one specific next step and a short check-in date, written down, so accountability is clear without feeling like a trial.
As a former NFL player, I saw that the best coaches pulled you aside early, told you the truth, and made it obvious they still believed in you. Trust survives hard conversations when the person knows you're on their side.
Call Before You Comment
An analyst on our research team told me she had stopped opening my Discord comments because they arrived in one heap every Friday. Her investor research had been some of our best, the groundwork for how we help client founders raise. Over about 2 months it got thinner. My notes may never have been the real problem.
On a mostly remote team you can miss a slide like that for weeks. We got on a video call and I asked what had changed, then waited. She had been quietly covering for a colleague who left. Call before comment is the team's rule now.
Examine Conditions, Then Remove Friction
The first question I ask is not "what is going on with your work?" It is "what has changed in how you are working?"
Performance rarely slips because someone suddenly became less capable. Something shifted. Their workload composition changed. A teammate they collaborated well with left. A process got introduced that fights against their natural rhythm. The work stayed the same but the conditions around it changed in a way that is draining them.
At Pigment, I had a contributor whose output dropped noticeably over about six weeks. My instinct was to address the work directly. Instead, I asked what her week looked like compared to three months ago. She described a situation where a reorganization had pulled her into twice as many cross-functional meetings, cutting her focused work time in half. The quality dropped because she was doing the same work in half the time, while spending the rest of her day in conversations that did not need her.
The fix was simple once we saw the real problem. We removed her from four recurring meetings and her output recovered within two weeks. No performance improvement plan. No difficult conversation about standards. Just removing the friction that was making it impossible for her to work the way she naturally operates.
Most performance conversations start with the symptom. The effective ones start with the conditions.
Hear the Reason, Then Confirm the Bar
I start the conversation the first time I notice a pattern, not after it's become a grievance. Frustration hardens when a manager sits on it for weeks and then unloads everything at once. By then the employee feels ambushed and you feel justified, and neither of you can hear the other.
So I do it privately and I open with what I've noticed, in plain terms. Something like, "The last few turnovers have needed fixes after the fact, and that isn't how you normally work. What's changed?" Then I stop talking. A capable person nearly always has a reason, and it's often one I'd never have guessed from the outside. Sometimes it's a schedule that stacked up, sometimes something at home, sometimes a task that was never explained well.
Accountability comes after that, and it works better because they've been heard first. We agree on one specific thing to fix and a date to look at it together again. I ask them to say it back in their own words, since a standard they've put into their own words is one they actually own.
In a business where cleaners work alone inside someone's home, trust is the whole arrangement. I'd much rather spend ten uncomfortable minutes early than let a good person quietly decide I've stopped noticing or stopped caring.
Reset Expectations With Visible Near-Term Action
I'm Co-Owner/COO at Nexa Roofing, so I deal with this in a trade where trust is built job by job. If our communication slips during a leak, emergency repair, or replacement, the customer feels it immediately.
My best approach is a "reset, not a reprimand" conversation: "This doesn't match the standard I've seen from you. What changed, and what do we need to fix this week?" That keeps their dignity intact while making it clear the standard still matters.
One example: a strong team member started delaying customer updates on emergency roofing jobs. We tied it back to our promise of transparency, clear timelines, and 24/7 reliability, then agreed that every active customer would get a clear next-step update before the day ended.
The key is to make accountability visible and near-term. Don't say "do better"; define the next behavior, the next deadline, and when you'll check back in.
Walk the Job, Then Diagnose the Miss
I run Mackey Services in the Greater Houston area, where one skipped detail can turn into a failed compressor, electrical hazard, or emergency callback. So when a good employee slips, I try to get to facts fast without making it personal.
My best opening is: "Walk me through the last job like I was standing next to you." On a rooftop HVAC call, that means hearing whether they checked belts, capacitors, coils, refrigerant, drains, and verified airflow before leaving.
If something was missed, I ask, "Was this a knowledge issue, a workload issue, or a care issue?" That question keeps trust intact because it gives them room to be honest, but it also makes clear that the standard is not optional.
Then I set one tight follow-up: the next few calls get a simple checklist review with a lead tech or manager. Accountability improves fastest when the employee knows exactly what "back on track" looks like.
Guide Self-Recognition and Acknowledge Progress
Running multiple brands -- closets, clothing, agency -- means I'm constantly managing people across very different creative and technical roles. When someone's output starts slipping, I've learned the worst thing you can do is wait, hoping it corrects itself.
The conversation I've found most effective starts with curiosity, not critique. I ask what's changed in their process, not what's wrong with them. Nine times out of ten, there's something real underneath -- a miscommunication about expectations, or they've quietly taken on more than they can handle.
At Closets by David Alan, our standard is white-glove at every touchpoint -- from the first consultation through final installation. When an installer or designer starts cutting corners on that, I'll walk them back through a specific project together, ask them what they'd want their client to experience, and let them identify the gap themselves. That self-recognition sticks far longer than me pointing it out.
The trust piece comes from following through consistently -- not just having the hard conversation once and walking away. Once someone recalibrates and delivers, I make sure they hear that too. Accountability without acknowledgment just builds resentment.
Trace Workflow Gaps Through the Run of Show
As owner of Blue Ridge Sound I run small crews through tight timelines on weddings and live events where one slip in setup or coverage can break the whole day. That experience has taught me to catch performance dips early by tying the conversation directly to the run of show we both already know.
I start by asking what part of the venue plan or equipment flow feels off right now instead of leading with the drop in quality. We then walk the coverage zones together and agree on one specific handoff or check they will own for the next event.
This keeps the focus on the shared system rather than personal shortfalls. The same method has carried over to my workflow automation work with STANDBY Local where small process gaps surface the same way.
Center the Child Experience, Then Co-Create Recovery
I run an early childhood education center where the stakes of a slipping staff member are immediate -- a distracted or disengaged teacher directly affects a room full of 3-year-olds the same day. That urgency forced me to develop a clear-eyed approach early on.
When I notice the drift, I anchor the conversation in the child's experience rather than the employee's behavior. Instead of "you've been inconsistent lately," it's "I noticed circle time has felt rushed this week -- what's making that hard right now?" That reframe shifts the conversation from defense to problem-solving almost immediately.
The most useful thing I've done is separate the observation conversation from the expectations conversation -- two different meetings. The first is purely listening. The second is where we co-create the recovery plan together, which means they own it rather than just tolerating it.
What I've learned after 17 years in this environment: capable people don't slip because they stopped caring. They slip because something changed and nobody asked. The question "what's making this hard right now?" has restored more accountability for me than any performance improvement framework ever has.
Name Obstacles and Redistribute Work
As a trial attorney who leads teams through complex civil rights and employment cases, I see performance slips early when preparing for jury verdicts against large institutions. Direct conversations prevent small gaps from turning into lost evidence or weak arguments.
I open by naming the exact change in deliverables, such as missed deadlines on discovery responses, and ask what obstacles stand in the way. This keeps the focus on the work rather than personal judgment.
In a recent whistleblower matter involving retaliation claims, that approach revealed a team member struggling with case volume after a family issue. We redistributed tasks and set clear checkpoints, which restored output without eroding loyalty.
The same method applies across my firm because every case demands consistent preparation to hold powerful parties accountable.
Review Cases and Rebuild Standards
I run a full-arch implant and in-house digital dental center, where a missed handoff can affect same-day teeth, medically complex patients, or oncology clearance timelines. So I try to address slipping work early, privately, and with curiosity before it turns into resentment.
My opener is usually: "You're capable, and this is not the level I'm used to seeing from you. What's changed?" Then I name the specific pattern, not their character, and ask what support, clarity, or obstacle we need to fix.
One example is our in-house lab workflow for All-on-X cases. If design notes, CBCT planning details, or healing-teeth handoffs start getting inconsistent, we don't lecture; we walk the case together, rebuild the checklist, and agree on what "done" means before the next patient is in the chair.
The trust comes from making the standard non-negotiable but the conversation respectful. My rule with staff is the same as with patients: treat the patient, not just the teeth -- and with employees, address the pattern without reducing the person to the mistake.
Use Facts to Build a Two-Week Reset
I address it in the first week with numbers on the table, before either of us has time to build a story about the other. In my NBN telecom contracting company we tracked every job in our own software, so when a good jointer's output or quality dropped, I saw it within days. I would sit down with that person privately, show them the last few weeks next to their normal standard, and ask one question: what changed? Often the cause was something fixable, a difficult site, the wrong crew pairing, or pressure at home, and people were relieved to be asked instead of accused.
The approach that restored accountability was agreeing on a short, specific reset together, usually two weeks with one clear target, and checking in every couple of days rather than waiting for a review. Because our pay was largely performance based, the numbers already mattered to them, so the conversation stayed about getting back to where they wanted to be. I never asked anyone to do work I would not do myself, and I think that is why those talks rarely turned into resentment. Catching the slip early, with facts and a question rather than a verdict, kept good people with us and kept trust intact on both sides.
Define Proof and Track One Commitment
Sometimes the best phrase is, "Let's define the next proof." It moves the conversation away from personality and toward one observable result. Name the missed standard, set a 14-day checkpoint and agree on the single proof that would show the work is back on track. Honestly, vague concern is where resentment grows. And don't turn one rough week into a verdict. Ask, "Which part stopped working first, the expectation, capacity, handoff or follow-through?"
If work that normally ships Friday has slipped to Monday twice in six weeks, the checkpoint stays specific, especially when Slack is noisy, priorities shifted twice, handoffs got fuzzy and everyone's calendar is packed at 4:30 p.m. (classic). Pick one visible commitment for the next 14 days. Keep it narrow. But write it down in plain language so nobody has to reconstruct the conversation later. I think trust survives when accountability is concrete and reversible. If the proof lands, confidence recovers; if it doesn't, the next conversation starts from facts.
Schedule Early Touchpoints to Prevent Breakdowns
In private households it starts small. The nanny who used to send a photo from the park stops sending anything. The housekeeper who anticipated everything starts waiting to be asked. By the time the family feels frustrated, the staff member usually does too, often for a reason nobody has said out loud.
That's why every placement we make has check-ins at two weeks, three months and six months. The three-month one matters most: the honeymoon is over, routines have set, and small irritations are still small.
The approach that works best is to open with curiosity, not the complaint. "What's changed for you lately?" opens a door that "You've been late three times" slams shut. Very often the answer is practical. The hours crept up, the role grew without anyone agreeing to it, or the parents give conflicting instructions. Once that's on the table, we agree on one or two specific expectations and a date to review them together.
The Spanish say "mas vale prevenir que curar," better to prevent than to cure. Most placements that fall apart could have been saved by one honest conversation a month earlier.
Use Data to Solve Problems Together
At CrewHR, I start performance talks with actual data. Missed shifts, flagged conflicts, whatever the platform shows. It keeps things about what happened, not how anyone feels. Then I skip blame and go straight to what needs fixing and how we can help, like tweaking their settings or adding notifications. People step up faster when they see you're solving problems with them, not just pointing fingers.
Recall Successes and Set Three Goals
When someone starts struggling, I remind them of specific deals where they came through, then address what's slipped. Recently we started weekly check-ins and narrowed focus to three goals. Energy and results shifted within the first week. Setting targets together makes accountability feel mutual, not forced.
Pursue The Passion · Career Education & Inspiration
Back on the road
