The Reconciliation Page: Where the Growth Story Meets the Evidence

Sellers commission a quality of earnings report to defend the past. The growth case has no comparable convention, so buyers rebuild it themselves and price what they cannot test. A single reconciliation page, tracing every line of the growth bridge from the narrative through its operating assumptions to the evidence, turns a forecast to be believed into a set of conditions to be tested. Its value lies in the gaps it shows, not in the agreement it claims. Sellers defend the past with evidence and the future with narrative The past arrives in a sale process already audited. A sell-side quality of earnings report has become close to standard in sponsor-backed deals: a Forvis Mazars partner estimates that at least 90% of the PE-backed transactions they see carry one, against about half of founder-led lower-middle-market businesses by the same estimate. Buyers appear to pay for it. GF Data’s analysis of 360 transactions completed since the third quarter of 2024 found that sellers who ran a sell-side QoE achieved an average of 7.4x TEV/EBITDA, against 7.0x for those who did not. That is an association rather than proof of cause, but its direction is consistent with what a buyer would expect to pay for evidence. The growth case is treated differently. It arrives as a narrative in the CIM and a bridge in the model: organic growth, new geographies, acquisitions, margin improvement, each with a number attached. The buyer is asked to accept the narrative and test the bridge, but nothing in the document connects one to the other. So the buyer’s team does what it usually does. It rebuilds the case from its own sources, finds the places where the pieces do not add up, and prices them. That is where processes break. Asked for the single biggest reason deals failed to close in 2025, respondents to Axial’s March 2026 survey of 107 lower-middle-market participants, from both the buy and the sell side, put valuation expectations first, at 28.3%, and diligence findings second, at 24.5%. Between them, the two describe the ground a growth case occupies: what the seller believes the business is worth, and what the buyer can verify. The past arrives audited. The future arrives as a story. A buyer prices the second at the confidence of its weakest assumption. A growth case is assembled from sources nobody reconciled A typical growth case draws on five sources: a market study, customer and contract data, management’s forecast, an acquisition pipeline, and the operating plan. Each is usually prepared by a different person, for a different purpose, at a different time. The market study sizes demand across a region. The customer data shows who buys today. The forecast states what management intends to achieve. The pipeline lists companies someone has spoken to. The operating plan assumes the trucks, routes and processing capacity will be there. Each document can be internally consistent without ever having been checked against the others. The questions a buyer asks in week four are the ones that sit between them. Can the company reach the customers the plan assumes, from the facilities it actually owns? Is there enough route or processing capacity to serve the volume the forecast adds? Which of the acquisitions in the bridge are named, engaged businesses, and which are a number standing in for a company nobody has identified yet? A CIM seldom answers them directly, because each sits at a seam between two sources and belongs to neither. The page reconciles three things, line by line The reconciliation page is one page in the evidence book that connects the investment narrative, the operating assumptions and the financial projections. It does so for every line of the growth bridge, in the same order the model presents them. Four kinds of growth claim appear in almost every environmental services bridge. Each rests on a different operating condition, and each can be tested against a different body of evidence. The tier column does most of the work. Organic growth resolves largely to the company’s own trading history, which is why buyers tend to leave it nearly intact. Network expansion resolves to the denominator described in the network share piece of this series: demand within economic reach of the company’s fixed nodes, not demand in the region. Inorganic growth and margin improvement rest increasingly on things that have not happened yet, and the page should say so rather than present them in the same typeface as the contracted revenue above them. The page is worth most where it disagrees The instinct in preparation is to make the reconciliation agree with management’s plan. That is the one outcome that makes the page worthless. A bottom-up build that arrives at exactly the forecast tells a buyer that it was either reverse-engineered or never independent, as the earlier piece on CIM claims argued. The value of the page is that it can show a gap, and that the gap is specific. In the illustration, the gap is $10 million of a $36 million growth case. Presented this way, it is not a weakness to bury. It is the part of the case the seller can still work on before launch, priced in the open, rather than the part a buyer discounts silently at a number the seller never sees. More useful still, the gap is located. The buyer knows that the pricing line is close, that the network line turns on one disposal contract, and that the inorganic line needs named companies behind its last $5 million. Those are diligence questions with an owner and a date. A reconciliation that agrees with management proves nothing. A reconciliation that shows a $10 million gap, line by line, proves the build is independent. Each party in the process uses the page differently The same page does different work for each party, and at different points in the process. For the banker, the page turns a defensive conversation into a prepared one: the questions that would otherwise arrive as valuation
Network Share: Why Market Share Is the Wrong Denominator

Market share divides a company’s revenue by every dollar of spend in its sector, including the dollars it could never economically serve. In route-based businesses the denominator that matters is set by fixed nodes, drive time and permits. Measured against it, two platforms with identical market share can be opposite businesses. Network share is the first measure in this series that cannot be computed without the graph. Market share divides by demand the company could never serve The previous piece set out where record-based evidence stops. This one begins the case for what sits inside that boundary, starting with a measure that cannot be built any other way. Almost every growth case in environmental services opens with the same ratio: company revenue over total market spend. A platform earning $60 million in a region with $2 billion of recorded spend holds 3%, and the other 97% becomes the runway. In a route business both terms are wrong. The denominator counts demand hundreds of miles from any facility the company owns that could accept it. The numerator treats an account held because no one else can reach it the same as one held against three competitors. The ratio is precise, and it describes nothing a buyer can underwrite. Network share replaces both terms. It is the share of economically serviceable demand that the company actually captures, where serviceable means within economic reach of the company’s own fixed nodes, acceptable under their permits and inside their capacity. The denominator is a set of named sites, not a sector total. That is why it can only be computed from the graph described in Piece 1: the demand layer resolved to sites and streams, the network layer to who serves whom, the assets layer to which outlets can take what. The scarce asset is the fixed node, not the truck A truck can be bought, leased or subcontracted in weeks. A permitted treatment plant, transfer station or disposal outlet takes years to site and permit, and often faces local opposition at every step. Routes radiate from nodes; nodes do not follow routes. Whoever controls the node in a geography sets the cost floor that everyone else there competes against. That makes the node’s footprint, not the company, the unit of competitive position. A company with modest national share can be dominant around a single treatment facility, while a larger competitor with no outlet of its own in that region pays a rival’s gate fee on every load and wins only on price. Revenue cannot tell the two apart. Network share can, because its denominator is drawn around the assets. Exhibit 1 sets the two ratios side by side. Trucks follow demand. Demand follows the node. A market share figure that does not know where the nodes are does not know where the business competes. The radius is not a circle A node’s economic reach is not a distance. It ends where the round-trip cost of serving a site, plus the cost of processing at the node, exceeds the price the site will pay or the delivered cost from the nearest competing outlet. That boundary follows drive time rather than miles. It stretches along interstates, stops at a toll crossing or a truck-restricted corridor, and contracts where congestion is predictable. It also differs by stream. A high-value, low-volume stream such as lab-pack hazardous waste justifies hours on the road; a bulk, low-margin stream justifies far less. The same facility therefore has several footprints at once, one for each stream it is permitted to accept, and none at all for a stream it cannot accept, however close the generator. The circle drawn on a CIM map blends every footprint into one, and the blend is always more generous than any of its parts. Identical market share can conceal opposite businesses Neither number is a verdict. Platform A’s 40% signals density and a structural cost advantage, but also a ceiling: from here, growth depends on extending reach. Platform B’s 8% signals a long runway inside its footprint, but one it has not converted, while every load it hauls pays a competitor’s gate fee. The two need different value creation plans, different acquisition screens and different diligence. Market share recommends the same plan for both. Captive and contested demand are different assets Inside the denominator, demand divides again. Captive demand sits within economic reach of one node only, the company’s. Contested demand sits within reach of two or more. Captive volume supports pricing and is the hardest revenue for a buyer to lose; contested volume is where share is won and lost and where price increases leak. A platform whose captured demand is mostly captive is a different credit from one whose captured demand is mostly contested, at the same headline share. The construct is not specific to waste Any business that moves material between fixed nodes and dispersed customers is measured against the wrong denominator by market share. Ready-mix concrete must generally be discharged within about ninety minutes of batching, so each plant’s market is a drive-time polygon and national share is close to meaningless. Aggregates are heavy and cheap relative to haul cost, which makes a quarry’s position a local question. Industrial laundry and uniform rental run routes from a plant whose capacity and drive time bound the territory. In each case the useful question is identical: what share of the demand reachable from this node does the company hold, and how much of it could no one else serve? Antitrust review already works this way. Remedies in waste consolidations have been defined local market by local market, around collection areas and disposal outlets rather than national totals. A deal team that leads with national market share is using a denominator the regulator has already set aside. Network share is a claim, so it carries a tier The denominator is the harder half to defend. The numerator, named accounts the company serves, is largely Observed. The denominator rests on drive-time assumptions, pricing, permit interpretation and
PFAS Is Redrawing the Biosolids Outlet Map

Washington has stepped back from setting a biosolids standard but left the liability in place. States, landfills, farmers and courts are now closing outlets on their own timetables — and their deadlines converge in 2028. The question is no longer how much material a platform handles, but how many places it can still send it. THE OUTLET MAP, IN FOUR NUMBERS IN BRIEF 1. Washington has withdrawn the rulebook but left the liability EPA’s July 2026 draft guidance on PFOA and PFOS in biosolids sets no concentration limit, carries no force of law, and replaces the January 2025 draft risk assessment the agency now calls flawed. Its comment period was extended to October 5, 2026. In August, EPA also rescinded its 2022 memo on addressing PFAS in wastewater permits. Liability moved the other way. On August 18, 2026, the D.C. Circuit upheld EPA’s designation of PFOA and PFOS as CERCLA hazardous substances, preserving strict, joint and several, retroactive liability. EPA’s protection for passive receivers such as utilities remains a policy, not a statute, and utility groups are still asking Congress for a fix. Industry petitioners had until October 2 to seek rehearing; for now, the designation stands nationwide. A federal number would have told landfills, farmers and neighboring states what acceptable material looks like. Without one, and with liability intact, each sets that line privately. The effect is not deregulation. It is regulation by the receiving gate. 2. Outlets rarely close by ban — they close by acceptance Statutes are the most visible way an outlet narrows and the most forgiving: the three enacted this year give 15 to 30 months of notice. The other mechanisms give far less, and all are operating now. The landfill mechanism is circular. Landfills that take sludge generate PFAS-bearing leachate, which treatment plants receive and send back out in their own sludge. As plants refuse or condition leachate, the landfill outlet gets costlier and its appetite for wet sludge falls. Maine has now made PFAS leachate treatment a condition of expanding its largest landfill, and EPA has signaled that federal leachate effluent guidelines could follow in 2027. An outlet’s durability depends on the outlet’s own outlets — and diligence that stops at the landfill gate misses half the chain. 3. The binding constraint is often in a neighbor’s statute book Maryland generates about 600,000 wet tons of sludge a year and hauls roughly 56% out of state, mainly to Virginia and Pennsylvania. Virginia’s 2026 law applies its monthly testing and 50 ppb prohibition to out-of-state suppliers. For a Maryland utility spreading in Virginia, the operative deadline is therefore July 2027, not Maryland’s October 2028. Seen from the other bank of the Potomac, a Potomac Riverkeeper Network study found that 35% of the sludge land-applied in Virginia came from out of state. The receiving state’s rule bites first because it is enforced at the point of placement. The same pattern runs elsewhere: South Florida’s Class B sludge has long gone to North Florida pastures that close to it in 2028, and Sanford, Maine already hauls its biosolids more than 140 miles. A generator’s outlet-risk register must therefore track every legislature its material reaches, not just its own. Harmonization cuts both ways: Maryland and Virginia aligned at 50 ppb, which eases compliance but compresses the cross-border routing arbitrage we described in August. That value now survives mainly at the seams between corridors. 4. 2028 is the convergence year Florida, which produces about 340,000 dry tons a year, restricts covered land-application sites to Class AA material from July 2028. Maryland’s prohibition starts that October. Juniper Ridge fills around 2028 unless its expansion — approved twice and challenged again by the Penobscot Nation — proceeds. Much of the replacement capacity — Sedron’s Indiantown facility in Florida, York Sewer District’s supercritical water oxidation pilot in Maine — is scheduled for the same year. Taken one at a time, each deadline looks manageable; plotted together, they cluster. creates procurement crowding, as utilities in several states seek capacity at once; contract exposure, as agreements priced on 2026 costs cross the window without change-in-law pass-through; and concentrated execution risk, since one slipped project leaves little slack elsewhere. A five- or ten-year municipal agreement signed this year will live through all three. 5. Losing an outlet multiplies cost — but the curve can bend back When an outlet closes, tonnage reroutes and the remaining outlet sets the price. A Brunswick, Maine district saw landfill costs rise from about $80 a ton to $120–170. Some Michigan utilities went from about $20 to $120 per wet ton after the state’s PFAS strategy took hold. Across utilities surveyed by CDM Smith with NACWA, WEF and NEBRA, PFAS policy raised average management costs by about 37%, with some facilities seeing costs double or more. The Brunswick district’s general manager’s advice to other states: “Expect a ban to be coming one way or another.” Pima County, Arizona shows the reverse. A 2020 moratorium raised its annual cost from $1.3 million to $3.3 million; site testing with the University of Arizona found low PFAS levels, and the moratorium was rescinded. Maine’s own 2024 waste plan contemplates a cautious return to land application if PFAS-destruction technology matures. Site-level evidence is an option on reopening the cheapest outlet — which makes source tracking a strategic asset rather than a compliance cost. 6. Replacement capacity is late, lumpy and contested Capital is arriving — Ara Partners has committed up to $500 million to Sedron — but delivery is uneven. Maine’s Crossroads dryer slipped from 2025 into 2026. The Aries gasifier proposed for Sanford is in contested review, and the company’s New Jersey plant was idled in 2025. Utilities are responding by building their own: Portland Water District is studying digestion, drying and pyrolysis, and Orlando is commissioning a plant-scale dryer. On destruction, a 2025 UK Environment Agency review found high-temperature incineration is still the only proven large-scale method, and new incinerators are hard to permit. During the gap between closure and
The Questions AI Cannot Answer in Diligence

Transaction value turns disproportionately on what was agreed, what is intended and what people are capable of — and none of it is written down anywhere a model can read. The risk is not that AI misses these questions. It is that it answers them anyway, with no signal that it has run out of evidence. This is where our evidence stops, why, and what should take over. The dangerous answer is the fluent one Every firm selling analytical capability into M&A — ours included — has an incentive to describe what its system can do and say nothing about what it cannot. The category is saturated with the first kind of claim. This piece is the second kind: a precise account of where records, and the models built on them, stop being evidence. The failure mode is not the one people expect. AI rarely fails in diligence by returning nothing. It fails by returning something — a renewal probability, a switching risk, a read on management — assembled from adjacent data and delivered in the same tone as a verified fact. McKinsey’s comparison of LLM-generated research against expert-interview reports found the generated output more optimistic in seven of ten industries, and roughly 40% of the important data points surfaced by experts absent from the LLM answers and unrecoverable by further prompting. The missing items were not obscure. They included contract structures, unit economics and regulatory hurdles: the variables a deal is priced on. AI does not fail in diligence by answering badly. It fails by answering at all — fluently, completely, and in the same voice it uses when it is right. Records show what happened. Value turns on what was agreed, intended and believed The industry graph at the center of this series is built from records: permits, manifests, filings, transactions, facility data. Records are behavior, observed after the fact. They are very good at establishing what exists, where, at what scale, and who has served whom. They are structurally silent on three things: what parties agreed privately, what they intend to do next, and what the people running a business are actually capable of. Consider a chemical plant that has shipped spent solvent to the same receiving facility, through the same transporter, for six years. The graph can show all of it: volumes, frequency, routing, the competing facilities within haul radius. It cannot show that the plant’s new owner has put the contract out to tender, or that its environmental manager is costing on-site recovery. Either fact takes the account’s value to zero within two years. Neither appears in any record until after it has happened. No better model closes that gap. A model reasoning over records inherits the boundary of the records; a model reasoning over the open web inherits a worse one, plus everything the web says that is not true. Exhibit 1 sets out six questions that routinely decide value in environmental services transactions, and exactly where the evidence stops on each. EXHIBIT 1: Six questions that move value and the point at which records stop answering them. Source: Espalier. Records include permits, regulatory manifests, facility filings, and transaction data. Inference is most dangerous where it is most plausible Inside the hard boundary sits a softer one: places where the data invites a conclusion it cannot support. These are the failure modes we design against, and each of them looks like rigor from the outside. EXHIBIT 2: Four ways a record-based build overreaches and the control that stops each one. Source: Espalier; US EPA, RCRA biennial reporting requirements (40 CFR 262.41). The common thread is that each error converts a property of the reporting regime into a claim about the market. A model will make that conversion every time unless the system forbids it, because the converted claim is always the more complete-looking answer. A missing record tells you about the reporting threshold, not about the market. Treat it as zero and the build looks most complete exactly where it is weakest. Coverage is incomplete by design. Publish where, and by how much No record-based build covers a market completely, and any build that implies otherwise should be discounted on that basis alone. Reporting thresholds, state-by-state variation in what is filed, and private-pay streams with no regulatory footprint all leave holes that are structural, not accidental. The response is not to paper over them. It is to make coverage a stated property of every claim. Each assertion carries a confidence tier — Observed where the record set supports it directly, Modeled where it is extended under published assumptions, Strategic where it rests on judgment — and the share of the denominator that sits in each. A network share figure that is 80% Observed is a different claim from one that is 80% Modeled, even when the headline is identical. A buyer is entitled to know which one they are being shown. Where a gap is material, the build also names what would close it: a state records request, a structured sample of customer calls, the seller’s own invoice data released in confirmatory diligence. A gap with a named remedy is a diligence item with a cost and a date. A gap without one is a discount, applied by the buyer at a number the seller never sees. The questions in Exhibit 1 need a status of their own Confidence tiering covers claims the evidence can carry at some level of confidence. The questions in Exhibit 1 cannot be carried at any. We mark them Outside the record: material to value, unanswerable from data, and routed by name to the instrument that can answer them. That status completes the division of labor. EXHIBIT 3: Each instrument answers one kind of question and fails in a predictable way when asked another. Source: Espalier. Tier definitions as introduced in Piece 3 of this series. The rule the exhibit enforces is simple: no question moves to a cheaper instrument than the one that can answer it. AI
The Economic Feedstock: Why a Ton of Food Waste Is Not a Ton

Regulation creates the ton. Feedstock decides what it is worth. As mandates reach smaller, more dispersed generators, advantage will go to those who underwrite feedstock cohorts — not tonnage. Our last piece showed that regulated tons and economically addressable tons are not the same number. This one goes a level deeper. Even once a ton is addressable, its value depends on what it is. A brewery’s wash water, a pallet of expired packaged yogurt, a restaurant’s plate scrapings and a grease-trap pump-out all enter the market as “food waste.” They need different trucks, different preprocessing and different digesters. They command different gate fees and produce different outputs. On a scale, they are identical. On a P&L, they are different businesses. The ton is an accounting unit. The feedstock is the economic unit. Capital allocated on the first — site selection, capacity sizing, acquisition valuation — is routinely mispriced on the second. Headline tonnage conceals the mix that drives value EPA’s final anaerobic digestion survey makes the point with its own data. Respondents processed 16.0 million tons of food waste in 2022 — but 14.7 million tons of it was beverage-processing waste, much of it at digesters co-located with breweries. Strip it out and the national total falls to 1.36 million tons. In 2023, reported volume fell to 7.7 million tons while the non-beverage figure barely moved, at 1.43 million. The headline halved; the rest of the market was essentially flat. Part of the distortion is physics. EPA converts beverage waste at the density of water, 8.34 pounds per gallon, against 3.8 pounds for source-separated organics. The same gallon becomes 2.2 times the tonnage. Any market sizing, capacity plan or valuation built on undifferentiated tons inherits that error. Regulation is adding the hardest tons last The next wave of mandates will not add more of the same ton. New York’s law, which today covers generators of two tons per week within 25 miles of an organics recycler, drops to one ton and a 50-mile radius in January 2027, and to half a ton in 2029. Each step pulls in smaller, more dispersed generators — lower stop density, more packaging, more contamination, longer hauls. Massachusetts shows why incremental regulation does not guarantee incremental tons. Diversion reached 380,000 tons in 2023, up from roughly 100,000 before the 2014 ban. But when the threshold fell to half a ton in 2022, processors saw less growth than anticipated — many newly covered generators were already diverting. MassDEP estimates the current ban alone gets the state to roughly 400,000 tons against a 2030 goal of 780,000. The remainder must come from smaller commercial and residential sources: precisely the feedstock with the weakest unit economics. Tellingly, New York’s regulator already reasons at the feedstock-and-capacity level. A generator’s obligation applies only to the extent a recycler within range has capacity to accept its scraps, based on DEC’s annual capacity estimate. When the agency models feedstock against capacity, an investor underwriting on tonnage is working with less information than the regulator. Food waste is a portfolio of feedstocks, each with its own economics Physical characteristics set the route. High-moisture beverage and process liquids suit wet digesters. Fats, oils and grease are energy-dense co-digestion feedstocks that demand dedicated receiving. Packaged retail material is worth little until it is depackaged, and the rejects cost money to landfill. Source-separated scraps go to compost or digestion depending on contamination and local outlets. Given the chance, the market re-routes tons toward the best-paying outlet. In Massachusetts, food waste sent to digestion rose from about 48,000 tons in 2016 to 239,000 in 2023, while composting fell from roughly 80,000 to 26,000. Same regulation, same state — the feedstock followed the economics. There is no price per ton — only net value per delivered ton Food waste does not clear at a single market price. Of 60 digesters in EPA’s survey, 80% collect tipping fees; among the 18 disclosing them, the 2023 median was $22.23 per ton and the mean $32.27. The spread is more telling than the average. Facilities price by feedstock — higher when material needs depackaging, charging grease haulers but not source-separated food-waste haulers, or charging for food waste but not manure. The gate fee is only the first line. As the first article in this series described, the full revenue stack runs through RNG, environmental credits and digestate — and every layer is feedstock-dependent. The metric that matters is net value per delivered ton: gate fee plus output value, less haul, preprocessing, rejects and conversion cost. Two tons at the same gate fee can sit on opposite sides of break-even. Nameplate capacity is not feedstock capacity The same logic applies on the supply side. EPA found that 55% of 44 responding digesters had spare capacity for off-site feedstock, but cautioned that usable capacity depends on proximity, preprocessing, co-feedstock availability and the optimal blend. A digester with idle nameplate can still be unable to take the next available ton profitably — too far, too dry, too packaged or wrong for the blend. Capacity is a property of a facility and a feedstock together, not of a facility alone. Decision intelligence turns the feedstock into an underwritable asset If value lives at the feedstock level, decisions must be made there — and that requires a system, not a spreadsheet of state tonnage estimates. A decision intelligence layer for organics does three things. It resolves the market to named generators, classified into feedstock cohorts from public regulatory data — designated-generator lists, permits, facility filings and capacity estimates. It prices each cohort by route, computing net value per delivered ton from haul distance, preprocessing burden, gate economics and outlet value. It refreshes as thresholds ratchet, so New York’s 2027 and 2029 tranches are modeled generator by generator before they arrive. The payoff is sharpest in underwriting. 50,000 tons of clean industrial liquid a short haul from a wet digester can be worth more than 75,000 tons of fragmented packaged retail requiring long hauls and heavy
Diligence should stress-test the build, not re-size the market

When a seller brings a genuine evidence layer, the buyer’s diligence scope does not shrink. It narrows onto the assumptions that actually move value—and gets harder in a more useful way. PE deal teams, corporate development and commercial due diligence providers still spend much of a process rebuilding the market the seller already had. The underwriting question is not how large the market is. It is how much of it this platform can convert into a business—and that is a question a seller can put on the table on day one. Buyer diligence still starts by rebuilding the seller’s market A growth story opens with market size: a fragmented sector, thousands of potential targets, a long consolidation runway. The buyer’s response is just as predictable. The deal team commissions commercial diligence, the CDD provider runs its interview program, corporate development rebuilds the target list from its own CRM, and each arrives at a market estimate that is directionally similar and irreconcilable in the detail. That work is not wasted, but most of it answers the wrong question. A market estimate describes what exists. An investment case depends on what can be captured—how many businesses fit the strategy, can be reached, can be acquired at the assumed price and can be operated profitably inside the network. The gap between the two is where underwriting actually happens, and it gets the least time because the budget went to re-establishing the denominator. “A market estimate can include every business in a sector. A build can include only the ones that fit, can be reached, can be acquired and can be run at a profit.” The build is a conversion ladder, and every rung carries its own evidence The useful unit of analysis is not the market but the conversion from market to build. In route-based sectors such as waste, that conversion has identifiable rungs, and each resolves to a different layer of the market graph. A target list can be long without being actionable: a business may fit the sector but not the service line, sit outside economic haul radius of the platform’s outlets, need capital that breaks the acquisition math, or have an owner with no reason to sell. An identified target is not an engaged target, an engaged target is not an actionable opportunity, and an actionable opportunity is not a closed acquisition. A credible build keeps those states separate instead of letting every name contribute equally to the forecast. Its revenue then comes from an explicit equation rather than a single bridge bar: Each term is a separate diligence question. How many transactions can this team close a year? What revenue profile do the reachable targets carry? How long before acquired revenue arrives at platform margins? Geography decides what acquired revenue is worth The same dollar of acquired revenue can be worth very different amounts depending on where it sits. A tuck-in that densifies existing routes and fills idle processing capacity creates value beyond its standalone P&L. One that needs a new yard, new trucks and a long haul to the nearest permitted outlet can dilute margin while adding revenue. That is why network share—the share of economically serviceable demand within haul radius—is the right lens for a roll-up, not national market share. A footprint that looks attractive at national scale can demand a different operating model at county level. The build therefore carries every acquisition through to service mix, EBITDA margin, integration cost, capex, working capital, transport and processing cost, and capacity, with each synergy stated alongside its evidence. Revenue is the output of the build. The operating model decides what it is worth. An evidence layer narrows buyer diligence to three jobs When the seller arrives with the conversion ladder already built and every rung traceable to records, the buyer’s scope changes shape. Reconstruction falls away and three harder jobs remain: verify the build, test the assumptions that move the answer, and form an independent view on what the seller cannot evidence—management quality, customer intent, renewal risk and integration reality. Each buyer experiences the shift differently. For the CDD provider in particular, the shift is from evidence assembly to interpretation—the part of the engagement that was always worth paying for. Stress-testing finds the assumptions the value rests on A stress test is often mistaken for a haircut. Its purpose is to locate dependence: which assumptions the build rests on, and how far the outcome moves when they move. What if deal pace is half the plan? If entry multiples rise a turn? If conversion from actionable to closed falls below the historical record? If acquired sites ramp slower, or processing capacity binds first? Run against the ladder, the answers sort the growth case into growth supported by existing evidence, growth dependent on testable assumptions, and growth that requires future execution or a strategic trigger. The reconciliation page then shows where the bottom-up build diverges from management’s plan and which driver accounts for the gap. A plan does not become weaker because its assumptions are visible. It becomes harder to misunderstand. Seller-selected evidence should be audited, not trusted The legitimate objection is that seller-produced evidence is seller-selected evidence. It is. The answer is not to ask buyers for trust but to hand them the tests that would catch a weak build. An evidence layer that fails any of these should be discounted accordingly. Buyers should welcome diligence that gets harder in a more useful way A genuine evidence layer does not remove diligence; it relocates it. Buyers spend fewer weeks re-establishing facts the seller already held and more on the handful of assumptions that change price. Disagreements arrive early and specific—forty targets outside haul radius, a capacity constraint read differently—rather than late and diffuse. The seller gains as well: a build the buyer can audit is a build the buyer can credit. ESPALIER VIEW The question in a growth-led process is not whether the market is big enough to support the story. It is whether the build is
Reshoring Is Redrawing the Hazardous-Waste Generator Map

The U.S. manufacturing build-out is not just adding waste volume — it is moving where technically complex waste is generated. That signal shows up years before the waste does, in project pipelines and process chemistry. Reading it early is a decision-intelligence problem. Sources: Reshoring Initiative 2024 Annual Report (244,000 U.S. jobs announced; advanced-manufacturing categories roughly two-thirds of the total); U.S. EPA National Biennial RCRA Hazardous Waste Report (~35M tons generated; large-quantity-generator c For hazardous-waste operators, the reshoring story is easy to oversimplify: more manufacturing should mean more industrial activity, which should mean more waste. Directionally true — strategically incomplete. The capacity being added is not evenly distributed. Much of it is concentrated in advanced manufacturing — semiconductors, batteries, electronics — which the Reshoring Initiative reports made up roughly two-thirds of 2024 reshoring and foreign-direct-investment job announcements. Those facilities don’t just add volume. They introduce different process chemistries, different waste streams, and different treatment requirements into markets whose environmental-services infrastructure was built around an older industrial base. The opportunity is not simply more hazardous waste. It is a different hazardous-waste generator map — and the most valuable signal appears before the waste does. The generator map follows the manufacturing map Hazardous waste is a derivative market: demand for collection, characterization, transportation, treatment, and disposal begins with activity at the generator. Historically that built recognizable industrial ecosystems — chemicals, refining, and metals developed in corridors, and environmental-services infrastructure grew up around them: denser routes, consolidating transfer stations, specialized treatment, long-standing generator relationships. The current investment cycle can create the same effect in a new set of markets, because one large facility rarely arrives alone. Exhibit 1 — One manufacturing investment can seed a broader waste ecosystem. The addressable market can extend well beyond the anchor facility. Actual streams and classifications depend on the specific process. Not every manufacturing dollar creates the same waste opportunity A dollar invested in vehicle assembly does not create the same environmental-services opportunity as a dollar in semiconductor fabrication, battery materials, or chemical manufacturing. What matters is not announced capex — it is process chemistry: what the facility makes, which chemicals and processes it uses, which residual streams result, which require specialized handling, what can be recovered, and where the nearest permitted outlet sits. Those questions determine waste-addressable value far better than the headline investment number. Exhibit 2 — Process chemistry, not capex, sets the waste-addressable value. Illustrative; actual waste classification depends on the specific process and facility. The point is directional: the same capex can carry very different waste value. The federal data helps — but only so far. Under RCRA, Large Quantity Generators (broadly, ≥1,000 kg/month of non-acute hazardous waste) file biennial reports, giving EPA facility-level detail on the nature, quantity, and disposition of waste. It is an unusually useful dataset, and it is inherently historical. A facility under construction today will not appear as a mature generator; by the time its volumes are fully visible in the data, part of the commercial opportunity may already be captured. The constraint may be capability before it is capacity It is tempting to turn reshoring into a capacity-shortage thesis. The more immediate question is whether the right capability exists in the right market. Hazardous waste is not interchangeable — different streams require different treatment technologies, permits, recovery pathways, and outlets. A market can have providers and still lack an economical local solution for a particular stream. When the appropriate treatment is distant, transportation, transfer, consolidation, and technical characterization all matter more, and operators with established downstream relationships gain an edge. The opportunity sits across the chain, not only at the final asset: generator → onsite service → collection → transfer / consolidation → transportation → treatment / recovery → disposal. A regional operator need not own an incinerator or landfill to benefit — control of generator relationships, permitted transfer capacity, technical expertise, and reliable downstream access can each be decisive. Geography starts to matter differently — and the signal is readable early Picture two regional hazardous-waste businesses with similar revenue, margins, and history. One operates in a mature industrial market with a stable generator base and settled competition. The other sits within trucking distance of several advanced-manufacturing projects moving from construction toward production. On trailing financials they look alike. Their forward strategic value does not. The second can establish generator relationships before volumes mature, add technical capability around the new customer base, and build density as suppliers arrive. The difference is invisible in historical financials — and visible in the pipeline. Building the forward generator map: a decision-intelligence problem There is a timing problem at the center of this thesis. Historical generator data tells you where waste has been concentrated; it is backward-looking and biennial. A factory announcement, on the other hand, is not revenue — projects are delayed, resized, or never reach the announced production level. Announced capex alone is a weak predictor. The value sits in the gap between the two maps, and closing that gap is a data problem: fusing where waste is today with where the generator base is forming, and translating an industrial project into a waste-stream forecast. The inputs are public but scattered — generator reports, facility permits, project and FDI pipelines, supplier filings — and, uniquely here, they require a chemistry-translation layer that turns “semiconductor fab under construction” into “fluoride and solvent streams needing a permitted outlet 140 miles away.” Access is not the barrier; assembly and interpretation are. A decision system connects five layers and tracks each project through its stages — announcement → construction → equipment → commissioning → production → supplier formation — so relevance rises as a project nears production. Exhibit 3 — What the decision system fuses to build the forward generator map Overlay the five layers and one output falls out: the build-ahead-of-demand pockets — markets where new industrial activity meets a technically demanding stream and an infrastructure gap. That is where an operator should position sales coverage, technical capability, transfer capacity, and downstream
Every Claim in a CIM Needs an Address

A buyer discounts what cannot be checked, and checking requires that each assertion resolve to something specific — a confidence tier, an evidence section, a query, a set of records. That address space is the decision system. Without it, a CIM is a hundred and twenty claims priced on trust. The discount is a mechanism, not a verdict It is tempting to read a buyer’s markdown as skepticism about the business. Usually, it is nothing of the kind. The buyer has read the same CIM, met the same management team and reached broadly the same view of the company. What they lack is any means of testing the parts of the case that matter most. A claim that can be tested gets argued about. A claim that cannot be tested gets priced. That is the whole mechanism, and it explains the feature of the process sellers find most frustrating: the discount is never raised in a meeting. There is nothing to raise. An assertion the buyer cannot verify does not generate a question — it generates a number in a model that the seller learns about only as a revised letter. Buyers are not skeptical. They are under-instrumented. Give them an instrument other than price and they will use it. Uncertainty does not arrive as one discount. It lands in four places Sellers tend to think of the gap as a single number applied to the headline. Buyers do not work that way. Different kinds of doubt attach to different instruments, applied by different people at different points in the process. EXHIBIT 1: Four instruments, four kinds of doubt — and only one of them is ever visible to the seller: Earnouts appeared in 24% of deals in SRS Acquiom’s 2026 study of more than 2,300 private-target transactions. Structure is where unverifiable growth goes when it cannot be priced. Each adjustment is separately defensible. Together they compound, and the compounding happens without a single fatal flaw ever being identified. The unit of underwriting is the claim, not the document A CIM is usually read as a narrative. It is underwritten as a list. Somewhere between a hundred and a hundred and fifty discrete assertions sit inside a typical document — about market size, share, retention, pricing headroom, contract duration, capacity, capex, win rates — and a buyer’s analyst works through them one at a time, sorting each into what can be tested and what has to be taken on trust. That sorting is the whole of diligence, and the seller has no visibility into it. Which means the question in sell-side preparation is not whether the document is persuasive. It is whether each claim inside it resolves to anything. A claim resolves when it carries an address: a confidence tier, a section of the evidence layer, a query against the market graph, and the record set that query returns. Four things, attached to the assertion itself rather than held in a separate appendix nobody opens. A claim with that address can be interrogated, disputed and corrected. A claim without one can only be believed or discounted, and buyers are not paid to believe. EXHIBIT 2: What an addressed claim looks like — one sentence from a CIM, resolved end to end: Structure is real; figures are illustrative. The point is that seven lines replace a sentence the buyer would otherwise have had to take on trust — and that the last line is the one that changes the price. Note what the last row does. It does not defend the number. It hands the buyer the means to attack it, which is the opposite of what sell-side instinct suggests and considerably more valuable. A denominator that survives the buyer’s own filtering is worth more than one they were asked to accept, because they have priced it themselves. Most claims are addressable, and the addresses sit in nameable layers The categories buyers discount are not diffuse. Each corresponds to a specific question, and most of those questions have an observable answer in a specific layer of the graph — demand, supply, commercial, network, assets, ownership, technology, signals. EXHIBIT 3: Most of the discount maps to evidence that exists and was not assembled: Two rows are deliberately negative. A system that claims to address every claim is being described as a product rather than a build. The same logic explains the pattern in how buyers treat a growth bridge. They rarely reject a driver outright; they cut each line in inverse proportion to how much of it is addressable today. Existing customer growth survives nearly intact because it resolves to the company’s own trading history. Pricing and cross-sell are cut harder. New wins are cut hardest, because nothing in the document establishes that the demand exists, let alone that it is reachable. The buyer is not discounting the driver. They are discounting the tier. Where the address space ends Two categories resist this treatment, and a seller is better served by hearing it now than by discovering it in week six. Add-backs are a financial-records question. Whether a restructuring cost is genuinely non-recurring, or simply this year’s version of a cost that recurs in another form, is settled in the accounts and in the buyer’s judgement about the operating model. No market evidence touches it. The narrower discipline that helps: distinguish adjustments that are economically credible from those that are merely technically defensible and drop the second category before the buyer has to. Management dependency is half observable. Whether contracts sit with the institution or the individual is a matter of record. Whether the relationship survives the individual leaving is a matter of judgement, and it belongs in a conversation with the customer rather than in a data set. The obvious objection: this opens the evidence layer to competitors Any banker reading this reaches the same place. Half the bidders in a process are strategics, and the serviceable demand build is a map of their customers as much as the seller’s.
Why National Waste Procurement Isn’t Actually National

The contract can be national; the economics stay local. The best programs are nationally governed and locally engineered — and a decision system is what makes each market’s real supply curve visible before the bid, not after. Waste is sold as a service but delivered through physical infrastructure. Every pickup begins on a local route and ends at a local or regional transfer station, recycling facility, organics processor, landfill, or specialized outlet. The customer may sign one master agreement; the material still has to move through a market-specific network operating under different capacity, permitting, labor, and competitive conditions. As large operators put it in their own filings, waste is “largely a local business.” The trouble starts when procurement treats national waste like a standardized indirect-spend category — aggregate the locations, compare unit rates, pick one supplier. That improves control, but it misreads the cost base. The invoice can be centralized; the service cannot be detached from local geography. A national contract standardizes governance. It cannot standardize the supply curve. The invoice is centralized; the service never leaves local geography — and that gap is exactly where a national program creates value or quietly destroys it. A national contract sits on top of local operating systems Five local variables determine the delivered cost of a site. A national bid can compare them; it cannot make them disappear. The national price is an allocation — not a market price A supplier can present one national rate card, but the price is still assembled from local economics. Some markets have dense routes and nearby owned disposal; others require long hauls, third-party tipping, or subcontracted collection. A blended rate redistributes those differences — it does not eliminate them. That creates a familiar trap: the buyer celebrates a low national average while high-cost sites are underpriced, low-cost sites subsidize the portfolio, and exceptions surface after implementation. The savings were visible in the bid model but not durable in the operating model. The variation is material. EREF’s 2024 analysis of 351 landfills reported a national average tip fee of $62.28 per ton — but the report exists precisely to show how that number differs by region, facility size, and ownership. Disposal is only one component of price; adding route and service variation widens the gap further. Scale helps — but only where it changes the local economics National scale is real value: it standardizes contract language, consolidates invoices, sharpens reporting and escalation discipline, and gives the customer one governance structure. But it is most valuable when it is backed by local integration. One national operator reported that roughly 67% of the solid waste it collected in 2025 was disposed at landfills it owned or operated; the advantage comes from controlling more of the local material flow, not from a national logo. Where the prime supplier lacks local routes or outlets, coverage is created by subcontracting — which can be the right answer, but changes the economics: another operator performs the service, the prime adds governance and margin, and visibility can weaken. That structure should be priced consciously, not treated as equivalent self-performance. How a decision system builds this view This is where a national-accounts provider wins or loses. The asset-light orchestrators, whose edge is coordinating local networks rather than owning trucks and landfills — do not compete by claiming universal self-performance. They compete by knowing, market by market, where the network is genuinely strong, where partners are required, and how to price and govern both honestly. That knowledge is a data problem — and it is exactly what a decision system is built to solve. The inputs are public, but they do not sit in one place. Franchise ordinances, hauler and transfer-station permits, landfill ownership and tip fees, diversion rules and material bans, and municipal contract terms and expirations are scattered across thousands of county and municipal systems, in inconsistent formats and on different cadences. Access is not the barrier; assembly is. A decision system connects them into a single, continuously current view of each local market — turning the abstract claim that “economics are local” into a specific supply curve a provider can price against. With that view, the national-accounts model inverts. Instead of quoting a blended national rate and discovering the exceptions after implementation, the provider prices each market from its true supply curve, bids to win where it has real operating advantage, governs the subcontracted markets transparently, and pre-positions backup capacity where a single hauler or a franchise renewal is the risk. The differentiated capability is not the broadest map — it is the best-understood one. What the view has to answer What this means for investors and operators The same logic that complicates procurement explains where strategic value sits. Route density lowers collection cost; transfer stations extend the economic radius of disposal; permitted capacity protects the outlet; local relationships stabilize volume — and a platform is worth more when these reinforce one another within a market. For consolidators, that means geographic adjacency can matter more than adding another distant dot to the map. For national-accounts businesses, the differentiated capability is not claiming universal self-performance — it is knowing where the network is strong, where partners are required, and how to price and govern both. That is precisely the view a decision system produces. THE ESPALIER VIEW The best national waste programs are nationally governed and locally engineered. They run on one data model, one performance framework, and one commercial architecture — while letting the provider mix and the price mechanism reflect the operating reality of each market. That is only possible if the local reality is legible. A decision system that assembles the scattered public record — franchises, haulers, permits, tip fees, and contracts — into a live, market-by-market supply curve is what lets a national-accounts provider price honestly, govern transparently, and win where it is genuinely strong. The map is national; the advantage is built market by market. THE BOTTOM LINE National procurement creates value when it makes local economics
PFAS Is Repricing Landfill Leachate

PFAS is changing who can accept leachate, what treatment it needs, and where cost and liability sit. That makes leachate an outlet problem, not a treatment-technology problem — and the winner is whoever controls the network of outlets the stream can still reach. For decades, leachate worked as a quiet secondary stream: collect it, connect to a sewer or haul it to a publicly owned treatment works, and rely on the wastewater system’s existing infrastructure. The economics were compelling because liquid is expensive to move and dedicated onsite treatment demands capital, operators, and residual expertise. The outlet absorbed the complexity — and it was cheap. PFAS breaks that model. Conventional biological wastewater treatment was never designed to destroy persistent fluorinated compounds, so PFAS passes through into effluent or partitions into sludge — linking the landfill’s incoming waste to the utility’s pretreatment program, to drinking-water protection downstream, and to the fate of biosolids. A low-cost outlet becomes a compliance and reputational exposure for whoever receives it. The consequence is not a new piece of treatment equipment. It is a reprice: the outlet that was cheap and reliable is becoming scarce, conditional, and expensive. This is not a treatment-technology story — it is an outlet story. PFAS doesn’t require every landfill to build a plant; it changes who will still accept the stream, on what terms, and at what price. Value accrues to whoever controls the outlets, not whoever owns the equipment. The chain PFAS forces on every stream Once PFAS is in scope, every gallon of leachate has to run the same five-step chain. Each step is a distinct service, and a distinct place value can be captured — and the platform that spans the chain, rather than selling one box in it, is the one that holds the customer. Exhibit 1 — Characterize → separate → move → destroy or dispose → prove closure Regulation is closing the outlet from several directions There is no single PFAS-leachate rule. The commercial signal is strongest when the regulatory actions are read together: each touches a different point in the chain, but collectively they raise the value of characterization, source control, pretreatment, and defensible final disposition. Exhibit 2 — Five regulatory routes, one direction of travel Regulatory status as of September 2026. Proposed actions are not final requirements. Federal rulemaking sets direction; state programs decide where spending starts. Maine now requires landfills that manage leachate to sample and report PFAS (beginning 2026) and, separately, PFAS characterization of treatment sludge before disposal or reuse (sampling by January 2027). Michigan sampled 32 active MSW landfills and found most were not major contributors to their wastewater plants — but a small number of high-load sites had to implement treatment through industrial pretreatment. The lesson: the market will not develop uniformly. High-load sites, sensitive receiving utilities, and constrained-outlet regions move first. The Espalier framework: map the constrained outlets, not the landfills Because the market is regional and site-specific, the right unit of analysis is not the number of landfills or the installed base of treatment equipment. It is the regional outlet network. The first attractive markets emerge where high PFAS mass load intersects constrained outlet access — the one quadrant where urgency and willingness to pay arrive together. Exhibit 3 — The leachate constraint map: where the first markets form. Position on this map is set by four variables: PFAS mass load (not concentration alone) · outlet scarcity · the receiving utility’s POTW and biosolids constraints. The economics reinforce the point. Minnesota estimated that removing and destroying PFAS from wastewater and biosolids could cost $14–28 billion over 20 years, with per-pound costs materially higher at smaller facilities and short-chain PFAS up to 70% more expensive to treat. That is not a leachate forecast, but it shows why dilution at a municipal plant will not become a universal solution — and why concentrated-stream treatment, tied to a secured outlet, is where the economics concentrate. Technology without an outlet is incomplete; an outlet without characterization is exposed. Where the value pools are — and how mature The most investable near-term opportunities sit upstream of the uncertain numeric limits. A customer can defer a full-scale destruction plant; it cannot defer understanding its PFAS profile, retaining an outlet, or documenting where its residuals went. Exhibit 4 — Value pools: near-term service and recurring vs. longer-term infrastructure THE INVESTMENT THESIS The winning business does not need to own every asset in the network. It needs to own the data, the decision rights, the customer relationship, and residual accountability — the ability to keep a landfill compliant and operating when the outlet requirements change, and to prove where the PFAS ultimately went. Assets can be rented; the customer outcome cannot. Six underwriting lenses Four moves operators should make now 1. Characterize now, on a mass-load basis — baseline your leachate with Method 1633 — total PFAS mass, not just concentration — before numeric limits arrive. You can defer a plant; you cannot defer knowing your profile. 2. Treat outlet access as a strategic asset — lock acceptance contracts, map alternates and hauling radius, and price the risk of losing your primary outlet before the utility reprices it for you. 3. Own the residual and its paper trail — map every spent medium, concentrate, and sludge stream to a defensible final disposition, with chain-of-custody data that survives a CERCLA look-back. 4. Buy optionality, not stranded capital — deploy modular pretreatment where mass load and outlet risk are highest; wait on full-scale destruction until performance and permitting are validated. THE ESPALIER VIEW The PFAS-leachate opportunity is not a market of landfills, and it is not a market of treatment equipment. It is a network of constrained outlets — and it can be mapped. PFAS mass load, permitted acceptance capacity, POTW and biosolids limits, hauling radius, and residual-path cost are all knowable, region by region. Map it that way and the first markets, the durable value pools, and the acquisition targets all resolve out of