Leadership and Team Management Case Studies
Professional5 Case StudiesUpdated Sep 2026
Most leadership problems show up first as missed numbers, slow handoffs, or a stakeholder who stops returning calls. The five cases below follow working teams through those moments, from a support group with blurred ownership to a regional leadership group choosing between projects it cannot fully price. Each case gives the situation, the constraint that made it hard, two analysis passes, four actions the leader actually took, and what changed afterward. The figures are internal to each organization and are meant to show scale, not to serve as benchmarks. Read them as working notes.
Case Study 1
A support team misses its targets while roles blur
Harbor Lane Benefits Desk, a fourteen-person internal team answering employee questions for a logistics company with 4,200 staff.
The situation
Harbor Lane's benefits desk handles roughly 2,300 tickets a month from employees, split across three intake channels: a shared inbox, a web form, and phone transfers from the payroll line. Growth had added people faster than it added structure. Tickets were tagged by whoever touched them first, so the same retirement eligibility question could be counted as a policy query, a payroll error, or a complaint. The target was to close 85 percent of tickets within two business days. The team closed 61 percent last month. Three senior agents each believed they owned escalations, and the payroll liaison believed escalations belonged to payroll.
The challenge
Yuki, the desk manager, had been asked by the director of total rewards to fix the backlog before open enrollment. She could add headcount only after the backlog stopped growing, and the payroll director wanted the phone transfers removed entirely. Yuki needed to find which problems were about roles, which were about service levels that nobody had agreed to, and which were simply gaps in the workflow data. She had three weeks and no clean report that separated those three.
Analysis
The first pass was about ownership rather than effort. Yuki mapped each ticket type to the person who had last closed it, and escalations passed through five people with no one accountable for the final answer. Practitioners trained in Leadership and Team Management would read that pattern as a roles problem before a capacity problem, because adding agents to an unclear queue only multiplies handoffs. The workflow data showed that 22 percent of tickets were reopened after a reply, which pointed to answers that were incomplete rather than slow. No service level had been written for escalations, so the two-day target covered only about half the work.
The second pass concerned what the other teams expected. Payroll wanted transfers gone because they generated errors that landed on its desk, but payroll had never reported those errors in a form the benefits team could check. The director of total rewards had set the 85 percent target without asking what counted as closed, so some agents closed tickets by sending a holding reply. Yuki's recommendation was a short written agreement: define the tickets that count, set separate targets for routine questions and escalations, and name one owner for each escalation category. She also proposed a weekly review that read reopened tickets alongside closures.
Actions taken
- Yuki wrote a one-page ownership map that named a single accountable agent for each ticket category, including escalations, and sent it to payroll and total rewards for sign-off before the next Monday.
- She set two service levels: four business hours for acknowledgment on escalations, and two days for resolution on routine questions, with the reopened-ticket rate reported beside each figure to expose quick closures.
- She moved routine benefit questions off the payroll phone line to the web form, and kept a same-day callback path for pay errors that payroll agreed to log.
- The team ran a weekly fifteen-minute review of the ten oldest open escalations, with the owner speaking first and the reopened rate posted on the wall screen, and Yuki tracked attendance.
Results
Within six weeks the closure rate on routine questions rose from 61 percent to 84 percent, and the reopened rate fell from 22 percent to 9 percent. Escalation acknowledgment moved from a full day to under five business hours. Payroll logged 40 pay errors in the first month, more than it had shared before, so the desk could see them. The 85 percent target was not met in every category during the first cycle, and two agents held escalations too long. The review made that visible early instead of letting it drift. Headcount stayed flat through open enrollment.
Questions for discussion
- Which of the three problems Yuki identified would have been hardest to see from the monthly dashboard alone, and why?
- If payroll had refused to log pay errors, what would the benefits desk have needed to change in its own targets?
- How would you decide whether a service level should cover all tickets or only the categories with a clear owner?
Case Study 2
A regional director loses confidence in a supply forecast
Brightwater Industrial Supply, a distributor of pumps and fittings with six regional warehouses and a twelve-person planning group.
The situation
Brightwater's planning group produces a weekly stock forecast for six regional warehouses, and its largest account relies on it to set buying schedules. Over the previous two months, three of the forecasts had understated demand for the same fittings, and two orders had been partially filled. Dana, the regional operations director for that account, had escalated twice in writing and then stopped replying to the planning group's emails. Her stated expectation was a forecast that would hold for at least six weeks and a named person she could call when a number moved. Planning reported one company-wide accuracy figure, which looked healthy and said nothing about her region.
The challenge
Ruth, the account manager, had two weeks to get Dana back on a call before the contractor network renewed its volume commitment. Planning could not promise accuracy it had not measured, and sales wanted a discount offered as a quick fix. Ruth had to separate what could be fixed immediately, such as the open orders, from what needed a process change, and then make commitments that planning could keep. She had no regional accuracy history and no record of who had been told what in the earlier emails.
Analysis
The first analysis looked at who owned the relationship, because the escalation had grown out of a gap between two roles. Planning treated Dana's emails as requests for information and answered them with the forecast file. Sales treated them as complaints about price and answered them with discounts. Neither group had a service level for response time on stakeholder messages, and the workflow data showed that the median time to a substantive reply was nine business days, although the messages were marked routine. Ruth found that the forecast had been revised four times in the same month without anyone telling the account.
The second analysis focused on what Dana had actually asked for, which was narrower than the escalation tone suggested. Experienced managers often separate the stated complaint from the operating need, and here the need was predictability: a forecast that would not move without notice, and a person who would explain why. The accuracy figure planning reported was a company-wide average that hid a regional error rate of roughly 18 percent. A 90 percent target at four weeks out had never been tested, so Ruth proposed one based on the last ten weeks and a notice rule for revisions above 10 percent.
- Acknowledge Confirm the affected orders and the specific harm before proposing fixes.
- Commit Name a measurable target and the owner who will report against it.
- Report Publish the figure on a fixed cadence, including the months it misses.
- Adjust Revise targets from observed data and tell the account before changing.
Actions taken
- Ruth called Dana the same day, read back every open order tied to the understated forecasts, and confirmed which could still be filled from the nearest warehouse.
- Planning agreed to a named forecast owner for the account, who would send a written note whenever a weekly figure changed by more than 10 percent, with the reason stated in plain terms.
- The team published a regional accuracy figure at the four-week horizon, calculated from the last ten weeks of actual demand, with an initial target of 88 percent reviewed each month.
- Ruth set a standing call every second Thursday for the account, with a planning analyst attending, and agreed that the sales lead would stop offering discounts as a response to forecast questions.
Results
Dana accepted the call within two days and the open orders were filled within the week, with one partial shipment that she agreed to accept. The first two monthly reviews showed regional accuracy at 86 percent and then 89 percent at the four-week horizon, short of target but moving in the right direction. Planning sent eleven revision notices in the first month, and Dana replied to four of them with questions rather than complaints. The contractor network renewed its volume commitment at the end of the cycle. Ruth still reports the figure monthly.
Questions for discussion
- Which of the commitments Ruth made to Dana would have been easiest for planning to break, and what would you have changed first?
- Why might a company-wide accuracy figure have delayed this escalation rather than prevented it?
- What evidence would you need before agreeing to a regional target that the planning group had never measured?
Case Study 3
Drawing sets return for rework after every review cycle
Ashgrove Structural, an engineering consultancy of about 85 people that prepares construction drawings for commercial projects.
The situation
Ashgrove's drawing packages move through a fixed sequence: a drafter prepares the sheets, a senior engineer checks calculations, a coordinator assembles the set, and a principal signs off before issue to the client. Over the last several projects, the set had been returned for rework in roughly 40 percent of review cycles, and the average package passed through the sequence three times before issue. Rework was not evenly spread. Most returns traced back to the handoff between drafting and checking, where drawings were sent as they were finished rather than as complete sheets. The service target was issue within ten business days of the internal freeze date, and the average was closer to twenty-two.
The challenge
Teodor, the operations lead, was asked to cut the cycle time by half without adding staff or changing the client's review schedule. The principals did not want the checking step shortened, and the drafters felt the returns were caused by vague comments from checkers. Teodor needed to map where each return originated, compare the fixes available, and show which one would reduce rework without weakening the review. The workflow data was incomplete, since comments were recorded in email and not in the project tracker.
Analysis
The mapping started with the handoffs rather than the people, because each side described the same events from opposite ends. Teodor traced 55 returns across the last 18 projects and recorded the handoff where each defect entered the sequence. Of these, 41 entered between drafting and checking, 9 between checking and coordination, and 5 at principal review, where the cause was usually a scope change rather than an error. The checkers' service level was defined as turnaround time, not comment quality, so a checker who returned a sheet with six vague comments still met the target. That definition explained much of the delay.
The second analysis compared the options against what each stakeholder would accept. Experienced managers tend to score options on the constraint that is hardest to move, which here was the principals' review step. Option one tightened the drafting checklist and required complete sheets before handoff, which the drafters could accept but which would add about two days to drafting. Option two added a shared comment log in the tracker with structured categories, which costs little but depends on checkers changing habits. Option three moved the checkers' service level to comment resolution rather than turnaround, which targets the cause directly and is the hardest to negotiate.
Actions taken
- Teodor adopted option two first, because it could start the same week, and required every checker comment to carry one of four categories: scope, standard, calculation, or clarity.
- The drafting checklist was changed so that no sheet left the drafter until it passed a two-minute self-check against the title block, the sheet index, and the revision history.
- The checkers' service level was rewritten to measure how many comments were resolved on the first return, with turnaround kept as a secondary measure for the coordinator's planning.
- Teodor moved principal review to a single scheduled window per package, with scope changes logged as formal change requests so they stopped arriving as informal comments during checking.
Results
Within the next four projects, the share of review cycles returned for rework dropped from roughly 40 percent to 19 percent. Average time from internal freeze to issue fell from 22 business days to 13, still above the ten-day target. Comments resolved on first return rose to 81 percent. The principals initially resisted the scheduled review window, but two of the four principals reported that fewer late scope changes made their time more predictable. The drafters noted that the self-check added minutes but removed a day of waiting.
Questions for discussion
- Which of the three options would you have piloted first if the principals had refused any change to their review window?
- How would you weigh a service level that rewards faster returns against one that rewards fewer comments per sheet?
- What would you need to see in the tracker before deciding that the comment log was working as intended?
Case Study 4
Attorneys resist a new matter tracking system during rollout
Pellham Legal Aid, a nonprofit office of 46 staff, including 19 attorneys, handling housing and benefits matters across three counties.
The situation
Pellham had tracked client matters in a mix of spreadsheets, shared folders, and paper intake forms for years. The board approved a case management system with a fixed go-live date, and the managing attorney, Imani, was responsible for adoption. The system required intake data at the time a matter opened, and it generated the deadline reports that funders now expected. Of the 19 attorneys, 6 had entered at least one matter in the first two weeks, and 4 of those had asked for their spreadsheets to be kept alive. The policy stated that all open matters must be entered within three business days of intake, a service level the old process had never measured.
The challenge
Imani had to bring the attorneys on board before the funder's first reporting deadline, which fell about six weeks after go-live. The attorneys said the system slowed intake, and some had a point: the required fields had been copied from a template that nobody had revised. Imani needed to separate the concerns that were about the software from those about the policy, and to decide which adoption steps to sequence first without letting the deadline slip. The supervising partner wanted the spreadsheets retired at once.
Analysis
The concerns sorted into three groups when Imani listed them against the intake log. The first was about fields: 11 of the 14 required items were not used in the attorney's own case notes, so entering them felt like paperwork for someone else. The second was about timing, because attorneys did intake on the phone and the system required a matter number before the call ended. The third was about trust, since two attorneys had lost a deadline during an earlier migration. Only the first group could be fixed by changing the form, which made it the logical starting point.
The sequencing decision mattered more than the training schedule. Imani put the field change first, which could be made in the template within a week and would answer the concern most attorneys had voiced in person. The second step was a phone intake mode that let attorneys open a matter with a temporary number and complete the record within the three-day window. The third was a shared tracker view for deadlines, which was the feature the two skeptical attorneys had asked about directly. Adoption was measured as the share of open matters entered within three business days, reported weekly, with the spreadsheet count tracked alongside.
Actions taken
- Imani asked the two most skeptical attorneys to review the revised intake template before it went live, and she took their edits on the required fields without argument.
- Required intake fields dropped from 14 to 9, and the other five moved to a screen completed after the call, which ended the complaint that phone intake stalled.
- The partners agreed that the deadline view would be the first screen attorneys saw each morning, and that the old spreadsheets would be read-only for 60 days rather than retired on the first day.
- Imani ran a fifteen-minute review each Friday with the intake coordinator, reading the weekly share of matters entered on time and naming the two attorneys below target for a one-to-one call.
Results
By the end of the adoption window, 17 of the 19 attorneys were entering most open matters within three business days, compared with 6 who had entered any matter in the first two weeks. The on-time share reached 88 percent, against a target of 90, and the spreadsheet count fell from 11 to 2. One of the two skeptical attorneys became the informal trainer for the deadline view, which helped more than the scheduled sessions. The funder's first deadline report was filed on time, with all required fields complete. Imani still reviews the weekly figure.
Questions for discussion
- Which of the three adoption concerns would you have tested first, and what evidence would have changed your order?
- What risks come with reading the spreadsheets as read-only for a fixed period rather than retiring them at once?
- How would you judge whether the two skeptical attorneys' concerns were resolved or only set aside?
Case Study 5
Choosing two of three initiatives on incomplete data
Corrigan Freight Systems, a software company of 210 employees that builds routing and billing tools for regional carriers.
The situation
The leadership group at Corrigan had budget for two of three initiatives in the coming cycle. The first was a rebuild of the customer portal, requested by support, which handled about 3,400 tickets a month and had an average handling time that was rising. The second was an automated quoting engine, championed by sales, which had a pilot with two carriers but no revenue data yet. The third was an integration with a fleet telematics partner, which the partner wanted signed within the quarter. Each sponsor had a business case, and each case rested on estimates that the group could not check in the time available.
The challenge
Clement, the chief operating officer, had to recommend a sequence to the board within three weeks. The sales lead argued that the quoting engine would pay back fastest, support argued the portal could not wait, and the partner's deadline was real but its penalty terms were unclear. Clement needed criteria that every sponsor would accept, a way to weigh risk when the estimates disagreed, and a decision that kept at least one option open. He did not have clean cost data for any of the three.
Analysis
The first step was to agree on criteria before anyone reopened the business cases. Clement proposed four: revenue or cost effect, the share of the estimate that could be verified within 30 days, the cost of being wrong, and whether the choice closed off the others. Scoring the cases against these showed that the sales case had the largest claimed upside but the weakest verification, since its revenue assumptions came from two carriers. The support case was less ambitious, but its handling time data came from the ticketing system and could be checked directly. Roles mattered too, since each sponsor defended numbers they did not own.
The second analysis asked what each option would cost to reverse. The partnership was the least reversible, since the signed term would bind the group for its full length, and its penalty clause was unclear, so the group could not size the exposure. The quoting engine was the most reversible, since the pilot could expand or stop without contract changes. The portal sat in between, since deferring it would raise a handling time already on record. A staged decision fit this pattern. The group would fund the portal, extend the quoting pilot, and ask the partner for a short extension with penalty terms in writing.
- 1Score on criteriaEach sponsor scores their own case against four shared criteria first.
- 2Verify key claimsCheck handling data and pilot revenue assumptions with carriers.
- 3Fund the reversibleFund the portal now, and cap the quoting pilot at three carriers.
- 4Negotiate the bindingSeek a short extension with penalty terms stated in writing.
- 5Review at set datesRevisit each option when its named check date arrives, not earlier.
Actions taken
- Clement set the four criteria in writing and asked each sponsor to score their own case before the group met, so that the scores could be compared rather than debated.
- Support's handling time figures were checked against three weeks of raw ticket data, and the sales revenue assumptions were tested with the two pilot carriers and a third carrier that had not yet joined.
- Clement funded the portal rebuild, capped the quoting pilot at a third carrier with a fixed review date, and asked the partner for a 60-day extension in exchange for a written penalty schedule.
- The board paper stated what would change the decision: a verified revenue figure from the pilot, or a penalty schedule the partnership could not accept, with a date for each check.
Results
The board accepted the staged plan in the cycle's first review. The portal rebuild shipped on schedule, and average handling time fell by a little over a fifth in the first full month, though support noted the gain came mostly from fewer repeat contacts. The quoting pilot grew to three carriers and produced verified revenue for two of them, enough to justify a wider rollout in the next review but not the broader figure sales had proposed. The partner granted the 60-day extension, and the penalty schedule showed exposure smaller than feared. Two of three checks reported on time, and the third moved with board agreement.
Questions for discussion
- Which criterion would you have weighted most heavily, and would the ranking change if verification carried double the weight?
- When a sponsor's estimate is the only one available, how much should a staged decision rely on it?
- What would have to be true for the group to choose the partnership first despite its lower reversibility?
Learning Resources
Professional CASE STUDIES
China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.