ESG Advisory Services Case Studies
Professional5 Case StudiesUpdated Sep 2026
ESG advisory work sits between strategy, operations, and assurance, and most of its difficulty shows up in the gap between what a company intends and what it can prove. The five cases below come from engagements where clients had reason to act, whether a customer requirement, a lender covenant, a community complaint, or a board question no one could answer. Each one moves from the situation to the constraints, the analysis, the actions, and the result. Names and figures are fictional. Read them as worked examples, because the right answer depends on the evidence in hand.
Case Study 1
Cutting energy and water use at a bottling plant
Hallsworth Bottling Group, a regional beverage producer with three plants and two distribution sites, employing about 1,400 people.
The situation
Hallsworth had committed to lower site energy and water intensity, and its largest retail customer wanted a verified baseline before the supply agreement came up for renewal. Two of the three plants had metered utilities, but the third had only monthly bills, and its water use was estimated from pump run hours. Leadership wanted a reduction figure it could publish without interrupting bottling during peak summer, when lines ran close to capacity. The sustainability team had a pilot budget of 180,000 dollars and no one with metering expertise. Two proposals were already on the table: replace the chillers with high-efficiency units, or install variable speed drives on the wash line.
The challenge
The problem was that neither proposal could be compared fairly. Chiller replacement looked attractive on paper, but its cooling load baseline was assumed rather than measured. The drive retrofit cost less to install but paid back only if the wash line ran at reduced speed for a known share of hours, and nobody had logged those hours. Any reduction the company reported would need to survive a customer audit, which meant the measurement method mattered as much as the projected savings.
Analysis
Teams grounded in ESG Advisory Services usually start by separating the objective from the instrument. Hallsworth's real objective was a reduction figure the customer's procurement team would accept, which set a higher evidence bar than a general efficiency gain. Two plants had metered utilities and could support a baseline directly. The third needed temporary submetering across at least one full production cycle before any projection counted as a baseline. Operational limits also mattered: any change touching the bottling lines in peak season risked lost output, and the plant manager had made clear that a missed shipment would end the pilot. Floor operators would feel the wash line change most.
The comparison came down to what each option would let the company prove. Chiller replacement had the larger projected saving, but its baseline depended on cooling load estimates that nobody had measured at the plant. The drive retrofit had a smaller capital cost and a shorter payback, provided the wash line ran at reduced speed for a known share of hours. After six weeks of logged run hours, the drive case held up and the chiller case did not. The verification method used the existing meters, reconciled monthly against bills and adjusted for production volume, so the reduction could be restated on the same basis each quarter.
- Opening weeksSubmetering installed and baseline logging begins on the third plant.
- After six weeksRun-hour data shows the drive case holds and the chiller case does not.
- Following monthDrive pilot runs on one wash line with rollback threshold in place.
- End of the quarterMonthly reconciliation restates the reduction for the customer review.
- Next capital cycleChiller replacement considered only after measured cooling profile confirms savings.
Actions taken
- Install temporary submetering on the third plant's chiller and wash line circuits for one full production cycle, and record run hours alongside electrical load so the baseline reflects actual operating patterns rather than nameplate ratings.
- Pilot the variable speed drive on one wash line during a low-volume week, with the plant manager holding a rollback procedure and a fixed threshold for lost output.
- Adopt the monthly reconciliation method in writing, including the production adjustment formula, and share it with the customer's procurement team before any reduction figure is published.
- Defer the chiller replacement to the next capital cycle and attach a condition: a measured cooling load profile that confirms the projected saving within a stated tolerance.
Results
The verified reduction across the two metered plants and the submetered third plant came to about 11 percent of energy per case and 8 percent of water per case, restated quarterly. The customer accepted the verification method at its next review and renewed the agreement. The chiller replacement stayed deferred. Floor staff raised two alarm setting issues during the pilot, and both were resolved in the second week. The company now has a verification routine it can reuse at its other two sites, which was the outcome the sustainability team wanted most.
Questions for discussion
- If the third plant had only been able to supply utility bills, which of the two interventions would have been easier to defend to the customer, and why?
- How should a team decide how much submetering is enough before a baseline counts as verifiable?
- What would you expect floor staff to resist first in a retrofit like this, and how would you involve them early?
Case Study 2
Reconciling emissions figures across four production sites
Carrowmore Fabrication, a steel components manufacturer with four production sites and a holding company that reports consolidated figures to its lender.
The situation
The holding company published an annual sustainability report with a greenhouse gas total, an energy figure, and an injury rate. A lender covenant tied a small margin adjustment to the energy number. An internal audit then flagged that two of the four sites had reported energy from estimated fuel purchases, and that one site had changed its boundary partway through the period, dropping a leased warehouse. No single person owned the data from end to end. Site managers entered numbers into spreadsheets, a finance analyst consolidated them, and the sustainability lead wrote the narrative without seeing source files. The board wanted the report reissued before the lender's review meeting.
The challenge
The consolidated figures looked consistent, which was part of the problem. Underneath, the evidence behind each indicator varied widely, and the group could not say which numbers were measured, which were estimated, and which had been adjusted by hand. Repairing the report meant agreeing the boundary, naming owners, and building controls within a few weeks. The work had to fit a quarter close that was already full, and two sites had no environmental staff at all.
Analysis
Carrowmore's lender used the energy figure to test a covenant, so the company needed a total it could defend, not a perfect inventory. That objective set the evidence standard. Purchase records and meter reads were acceptable primary sources, while spreadsheet estimates needed a documented method and a stated uncertainty range. The boundary question was harder. Excluding the leased warehouse, a choice made mid-period, had lowered the total without a record of why. The smaller sites could not absorb a full rebuild, so the work had to be narrow and sequenced. Stakeholder impact fell on the finance team, who would be asked to certify figures they had never originated.
Ownership was the gap that caused most of the trouble. Site managers keyed in numbers, finance consolidated them, and the sustainability lead wrote the narrative from a summary tab. The fix was a register: each indicator got a named owner, a source document type, a method note, and a reviewer who did not enter the figure. Monthly tie-outs compared utility invoices with the entered totals, and any estimate above a set share of the site total needed a second signature. Disclosure then stated which figures were measured, which were estimated, and which excluded the warehouse, with the restatement logged in an appendix. Dull work, but it holds up under review.
Actions taken
- Write a boundary memo that lists every site, lease, and joint venture, states the consolidation approach, and records the warehouse decision with the reason and the figure it changes.
- Assign a named owner and reviewer for each of the twelve indicators in the report, and record the source document type for every entry in a shared register that finance can audit.
- Replace estimated fuel figures at the two sites with invoice-based totals for the full period, and flag any remaining estimate with a stated uncertainty range and a second signature.
- Run a monthly tie-out between utility invoices and entered totals for six months, and report variances above the agreed threshold to the audit committee before the lender review, not after.
Results
Within the quarter, the tie-outs found two invoice totals keyed into the wrong site column and one estimate overstated by a meaningful share. Corrections brought the energy total inside the uncertainty range of the invoice-based figure. The reissued report disclosed the boundary decision and labeled each indicator as measured or estimated. The lender accepted the covenant calculation without a further request. Two indicators remain estimates, and the report says so. The register now sits inside the finance close, and the sustainability lead no longer drafts narrative from summary tabs.
Questions for discussion
- Which indicator in your own reporting would be hardest to trace back to a primary record, and who currently owns it?
- When a boundary change lowers a headline figure, what should trigger a disclosure to the people relying on it?
- How much estimation is acceptable in a first-year report before it undermines the credibility of the whole document?
Case Study 3
Easing neighbor concern over a quarry expansion
Northfield Aggregate, a family-owned quarry operator with one hard-rock site and a crushing plant that supplies road contractors across the district.
The situation
Northfield planned to extend its quarry by a modest area and build a new haul road to reach it. Residents in two neighborhoods near the eastern boundary had raised concerns at a public meeting: dust on laundry and cars, blasting vibration, and truck traffic past a primary school. The expansion needed a permit amendment, and the planning office asked for a community impact assessment before it would decide. The company had a community liaison who answered calls but kept no structured log, no dust monitoring data, and no agreed hours for blasting. Management wanted the amendment approved without a long delay, because the current cell was close to exhausted.
The challenge
The concerns were a mix of measured problems and perception. Some complaints matched real dust events in dry weather, while others reflected worry about property values and about a quarry most residents had never visited. The company could not dismiss either group, and any mitigation had to be proportionate. Expensive measures without evidence would waste budget, while too little would confirm the sense that the company was not listening.
Analysis
The company wanted an approved amendment on a schedule that kept its contracts intact, but the permit office would only accept a credible impact picture. The liaison's call log showed scattered complaints with no pattern, so the advisory team matched complaint dates against wind records from the site weather station and placed two temporary dust monitors at the nearest homes for one dry season. Blasting was tied to the crusher's maintenance cycle, so moving it would cost production days. The impact fell unevenly. Households closest to the pit bore most of the dust and vibration, while residents farther off were mostly worried about precedent and property values.
A stakeholder map drawn in the first week separated three groups: households within the dust plume, families along the haul route past the school, and the wider district that would use the expanded supply. Each group had different exposure and different trust in the company. A drop-in session at the library, a site walk for three affected households, and a written summary sent to the school board followed from that map. Mitigation was kept proportionate. Water carts on the haul road and a revised truck schedule outside school arrival times were cheap and visible. Enclosed conveyors were considered but deferred until the monitors showed whether dust was a sustained problem.
- Measure Monitors and vibration logs record conditions at the nearest homes.
- Listen Drop-in sessions and site walks capture concerns in residents' own terms.
- Adjust Haul schedule and water carts change when readings or complaints rise.
- Report Monthly figures go to the community group and the permit office.
Actions taken
- Install and maintain two dust monitors and one vibration logger at the nearest homes for a full dry season, with readings shared monthly with the residents who agreed to host them.
- Hold a drop-in session at the library and a separate site walk for three affected households, recording every concern and marking which ones the company will act on and which it cannot.
- Adopt a revised haul schedule that keeps trucks off the school route during arrival and departure windows, and add water carts on the access road during dry weeks.
- Set a written trigger for enclosed conveyors: if monitor readings exceed the agreed dust level on more than a stated number of days, fund the conveyor upgrade in the next budget cycle.
Results
The permit office approved the amendment after reviewing the monitoring summary and the consultation record. Over the dry season, dust readings exceeded the agreed level on four days, all tied to specific haul-road events that the schedule change and water carts were meant to address. No further exceedances followed in the next period. Two households asked for their monitors to stay in place longer, which the company agreed to. The school route complaints stopped after the revised schedule. The conveyor upgrade remains in the budget plan with its trigger, and the liaison now reports complaints as a monthly figure to the community group.
Questions for discussion
- Where in your own projects does a complaint log stand in for evidence, and what would a better record look like?
- How should a company decide which resident concerns it can act on and which it must explain without promising change?
- What trigger would tell you that a proportionate measure has stopped being enough?
Case Study 4
Prioritizing labor checks across a packaging supply chain
Corvane Plastics, a manufacturer of molded food containers that buys resin and finished parts from about 60 suppliers across three regions.
The situation
Corvane's largest retail customer had added a supplier code covering working hours, wages, and worker documents. A review of 60 suppliers produced 14 questionnaire responses that were incomplete or contradicted site records. Three suppliers had passed earlier audits, but a trade union report and a regional news story raised concerns about unpaid overtime and recruitment fees at one of them. Corvane could not audit every supplier within the customer's deadline, and its procurement team had no one trained in labor interviews. Two suppliers were up for contract renewal within the quarter, so any decision had to fit the commercial calendar.
The challenge
Where to spend limited audit days was the real question. Supplier risk was uneven: some were small plants with informal payroll, others were large contract manufacturers with strong systems on paper. Ranking by spend alone would miss the small plants where labor practices were hardest to see. The team needed a method that combined risk signals with spend exposure, and a clear definition of what would count as corrective-action evidence.
Analysis
Corvane's commercial aim was to keep the customer contract and avoid a supply interruption, but the customer's requirement was about verifiable conditions. That made the evidence bar higher than a signed policy. Questionnaires had limited value where answers contradicted payroll records. Document sampling gave the strongest evidence: time records, payslips, recruitment fee receipts, and contracts in workers' languages. Audit days were limited, and the two suppliers facing renewal had the largest volumes. Pulling them from supply would have hurt the customer's schedule. Stakeholder impact reached past the company: workers in those plants would bear the cost of any abrupt change, and a sudden cutoff could push practices underground.
Prioritization combined two signals. Spend exposure showed where a failure would hurt Corvane most, and a risk rating drawn from country context, sector, and recruitment model showed where a failure was most likely. The team scored each supplier on both, then placed them on a grid. Smaller plants with informal payroll and agency-heavy hiring ranked high on risk despite low spend, and they received sampling visits before the larger contract manufacturers with certified systems. Corrective action needed its own evidence standard. A written plan was not enough. Closure required reimbursed recruitment fees traced to receipts, corrected pay records covering the affected period, and worker interviews conducted without management present.
Actions taken
- Run document-based sampling at the four highest-risk suppliers, pulling three months of time records and payslips and comparing them against the hours stated in the questionnaire responses.
- Hire an external labor interview team with interpreters in the workers' languages, and conduct interviews off site or outside shift hours so that workers can speak without management present.
- Require a corrective action plan from the flagged suppliers with dated steps, and accept closure only when repaid fees are traced to receipts and corrected pay records cover the affected period.
- Agree a phased sourcing plan with the two renewing suppliers that keeps volume flowing, shifts new orders only after verified progress, and gives Corvane a documented exit if repairs stall.
Results
Document sampling at the four highest-risk plants found unpaid overtime at two of them and recruitment fees charged to workers at one. Interviews confirmed the pattern at the same plant and surfaced a contract language problem at a second. Both suppliers agreed corrective plans. The renewal went ahead for one of them on a phased basis, with new orders held back until repayments were verified. The other supplier moved to a shorter contract with a quarterly review. The customer accepted the evidence standard, and Corvane's procurement team now carries a sampling schedule as part of every renewal.
Questions for discussion
- When a supplier's paperwork is complete but its payroll records look thin, which evidence would you trust more, and why?
- How would you decide whether a repaid recruitment fee closes a corrective action, or whether the underlying hiring model also has to change?
- What should happen to a supplier's volume if its corrective plan stalls for a second review period?
Case Study 5
Planning a staged transition for cold storage facilities
Pemberly Ridge Cold Storage, a refrigerated warehouse operator with four facilities that leases space to produce importers and dairy distributors.
The situation
Pemberly Ridge leased refrigerated space to produce importers and dairy distributors. Its main facility drew power from a grid connection that had been only partly upgraded, and two of the four sites used ammonia refrigeration systems facing tighter rules on leak detection and refrigerant replacement. Customers had started asking about backup power and energy sourcing, and the largest client had signaled it would move volume if the company could not show a plan for heat-event outages. The board wanted a plan for the next capital cycle, but the policy direction on carbon pricing and refrigerant rules remained unsettled. Capital available for the first stage was limited to a modest share of annual earnings.
The challenge
The main variables could pull in different directions. A stricter refrigerant rule would favor one replacement path, a sharp rise in grid prices another, and tighter client demand a third. Committing to one path too early risked stranded spending, while waiting for policy clarity risked losing the client. The company needed a plan that could hold under several futures, with stages that could be paused or accelerated based on signposts it could actually observe.
Analysis
The business objective was to keep the largest client and avoid an outage during heat events, ahead of any abstract climate goal. Evidence on future costs was weak. Policy timing was unclear, and vendor quotes for alternative refrigerants varied widely. Instead the team built four scenarios from two uncertain drivers, tightening refrigerant rules and sharp grid price rises, and tested each against outage risk, cost of ownership, and capacity to serve the client. Constraints were capital, the need to keep cold chains intact during any retrofit, and a workforce partly trained on new systems. The client's planning teams and the drivers who moved product between sites would feel each stage.
A staged plan turns the scenarios into decisions with thresholds. Stage one spends only on work that pays under most scenarios: leak detection on the ammonia systems, controls that cut peak demand, and a backup generator sized for the largest site's critical loads. Stage two waits on signposts it can watch, such as a published refrigerant rule or a grid tariff change beyond a set percentage. Measures the board can check include outage hours during heat events, kilowatt hours per pallet position, leak incidents, and the share of capacity with verified backup. Equipment chosen for stage two must install without closing a cold room beyond the spoilage window.
Actions taken
- Install continuous leak detection on both ammonia systems and set alarm thresholds with the maintenance contractor, reporting any alarm to the operations manager within the same shift.
- Size and order a backup generator for the largest site's critical loads, and test a full transfer under load twice in the first season before any client is told the site has backup.
- Define the stage two triggers in writing: a published refrigerant rule, a tariff change beyond the agreed percentage, or two heat-event outages in one season, each with a named decision owner.
- Share the staged plan and its measures with the largest client's planning team, agreeing in advance which metrics will be reported each quarter and what volume shift would follow a missed threshold.
Results
After the first season, leak detection caught two small seepages at the ammonia systems, both repaired within the shift. The generator transfer test failed once on a control fault, then passed on the second attempt. Outage hours during heat events fell to zero at the main site, though the smaller facility still had a gap in backup coverage. The client accepted the staged plan and kept its volume on a shorter review cycle. The second stage remains on hold, since the refrigerant rule has not been published. Its triggers are written, and the operations manager reviews them each quarter against the tracked measures.
Questions for discussion
- Which of the four scenarios would you weight most heavily if the client's contract renewal came before the policy was clear, and what would you change?
- What makes a signpost measurable enough to trigger a decision, rather than a general concern the board can keep discussing?
- How would you keep a staged plan credible to customers when the second stage is deliberately not yet committed?
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.