This is the first edition of a map we will publish every year. It takes the four shapes of the AI dividend and applies them to every industry we serve: twenty-five sectors, each scored against the three customer gates, located by the shape at its center of gravity, with the split inside it, the force most likely to bank its dividend if it stands still, and which line is moving. Read it twice: once for your market, and once for the mix inside your own walls.
If you have not read the framework, here is the short version. The AI dividend is the gap between what work used to cost and what it costs now that machines carry part of the load. It is real everywhere; who keeps it depends on three questions about your customers. Where buyers cannot fully specify the work, the dividend takes a Rising Bar: AI becomes table stakes and the gains flow into customer expectations. Where they can specify it but still choose on trust, it takes a Slow Melt: margins hold until the buyers start shopping. Where they specify it and shop it but the market cannot concentrate, it takes a Rollup: consolidation capital takes most. Where all three gates stand open, the Floor Falls: efficiency compounds until one or two players remain. Three forces off the map can decide the dividend regardless of your shape: the layer above, the regulator, and the capital. The framework article explains each, and what we would do in your seat under every one of them.
2026 Key Findings
- The headlines are inverted. Weighted by payroll, roughly one in three American jobs sits under a Rising Bar, where AI is already table stakes. Fewer than one in twenty-five sits on a true Falling Floor. Adopt-or-die is correct for the smallest shape on the map and wrong for the majority.
- The most exposed sector we serve is Logistics. Load-to-capacity matching is specifiable, rate-selected, and network-driven, which is the exact recipe for a Falling Floor. Brokerage and matching sit nearest the edge.
- The swing state is Retail. It is the largest block of jobs that could change shape in a single edition, and the trigger is not a new model. It is the customer’s agent doing the shopping.
- The slowest line is Aerospace. Certification freezes the competitive boundary more firmly than in any other sector, which is precisely why the supplier tiers beneath the primes are quietly rolling up.
- The layer above, not a rival, is the dominant threat in roughly a third of the map. Tourism, food service, retail, media, education, airline distribution, healthcare’s revenue cycle, and biotech’s discovery platforms all face a Falling Floor operating directly above a Rising Bar. This is the risk the Rising Bar majority is actually carrying.
- Two exceptions prove the gates. Pharmaceuticals has been winner-take-all for a century by patent, and AI changes only who reaches the patent first. Utilities have no selection gate at all, so the regulator, not the market, decides who keeps the dividend.
- The fastest-moving line in 2026 is the buying line. Agentic procurement threatens every trust-sheltered market at once: banking, credit unions, residential real estate, commodity manufacturing, and the audit and legal work that corporate buyers are learning to purchase by the spec.
- The sectors with the widest window get the least urgency. The four Slow Melts on the map, banking, credit unions, residential real estate, and education, have the most protected years left and the quietest AI conversation. That mismatch is the one this map exists to correct.
The Weight of the Map
Now the arithmetic behind the first finding.
Weight every sector by its workforce, using Bureau of Labor Statistics payroll data, and the AI economy inverts the headlines. Roughly one in three American jobs sits in industries whose center of gravity is the Rising Bar: healthcare’s eighteen and a half million workers, manufacturing’s nearly twelve million, food service’s twelve million, plus media, energy, legal work, and the rest of the economy where judgment, licenses, and physical presence hold the gates shut. The industries centered on the Falling Floor employ fewer than one American worker in twenty-five, roughly four percent.
One caution about what that number can and cannot prove. Employment weight maps jobs, not profit pools, and a sector can keep nearly all of its people while a layer above it takes an outsized share of its margin. The map says so, sector by sector, wherever it applies.
The urgency industry writes for the smallest shape on the map. Adopt-or-die is factually correct for a sliver of the economy and factually wrong for the majority, and the majority can tell.
Read that asymmetry twice, because the entire AI conversation is built against it. Capable executives checked the rhetoric against their window and tuned out, and the window was the wrong instrument: it shows neither the ratchet in the largest shape, nor the shelters expiring, nor the layer overhead already pricing the ceiling.
The Map at a Glance
Sectors are grouped by the shape at their center of gravity. “Who banks it” names the party positioned to keep the dividend if the sector stands still. “Line moving” names which boundary is shifting: the working line (capability, slow) or the buying line (the customer’s agent, fast).
| Sector | Shape at center of gravity | The split inside | Who banks it if you stand still | Line moving |
|---|---|---|---|---|
| Falling Floor | ||||
| Logistics & Supply Chain | Falling Floor, at the edge | Brokerage and matching nearest the floor; warehousing and 3PL in Rollup; network design holds a Rising Bar | The most efficient matcher; capital in 3PL | Both |
| PropTech | Falling Floor | Features are no longer a moat; proprietary data and workflow lock-in are | The category winner, not necessarily you | Working, fast (software-native) |
| Rollup | ||||
| Insurance | Rollup | Personal lines near the floor under aggregators; specialty and complex commercial hold a Rising Bar | Consolidation capital; aggregators in personal lines | Both |
| Slow Melt | ||||
| Banking | Slow Melt | Deposit stickiness shelters the franchise; operations are fully specifiable | You, until the customer’s agent shops rates | Buying, fast; the clock has started |
| Credit Unions | Slow Melt | Same as banking; the cooperative structure returns a captured dividend to members | Members, if captured in time; otherwise the agents and the megabanks | Buying, fast |
| Residential Real Estate | Slow Melt | Title and escrow melting already; judgment and advocacy hold the bar | You, until commission pressure and agentic buying open the gate | Buying, fast |
| Education & EdTech | Slow Melt | Credential sheltered; the teaching layer beneath it is being captured by AI tutors | The tutoring layer above the institution | Working, slow on the credential; the layer is moving fast |
| Rising Bar | ||||
| Legal | Rising Bar | Research, review, and drafting are commodity; volume practices are Rollup targets where regulation permits | Customers, as expectations; capital in volume practices | Buying, fast: legal ops buying matters by spec |
| Family Offices | Rising Bar, most sheltered | Reporting sophistication is the bar | Principals, as expectations | Working, slow |
| Private Equity | Rising Bar for itself; the capital for everyone else | Screening is table stakes; edge lives in proprietary flow and the operating playbook | The funds whose playbook lands in EBITDA | Working, slow |
| Commercial Real Estate | Rising Bar over melting layers | Underwriting commoditizing; appraisal in Slow Melt; property management in Rollup | Customers on deals; capital in property management | Working, slow; the rollup is moving now |
| Healthcare | Rising Bar in care, three shapes at once | Revenue cycle industrializing; practices consolidating | Vendors on the revenue cycle; capital in practices | Working, slow on care; capital moving fast |
| Biotech | Rising Bar, floor upstream | Discovery velocity is table stakes; the Falling Floor lives in the discovery platforms | The discovery platforms | Working, slow |
| Manufacturing | Rising Bar, commodity exposed | Differentiated and custom work sheltered; commodity lines become price-takers under agentic procurement | Customers’ procurement agents, on commodity lines | Buying, fast for commodity lines |
| Aerospace | Rising Bar, slowest line | Certification freezes the line; supplier tiers rolling up | Capital, in the tiers beneath the primes | Frozen; the rollup is moving |
| Oil & Gas | Rising Bar, ratchets each cycle | Oilfield services are bid-selected and in Rollup; the back office is pure capture | Cycle survivors; capital in services | Working, slow; ratchets with every downturn |
| Airlines | Rising Bar, oligopoly by structure | Revenue management is a machine arms race; distribution is the fight | Whoever the traveler’s agent books through | Buying, fast on distribution |
| Food Service | Rising Bar under platforms | Delivery margin already with the platforms; app rankings select for you; multi-unit back office in Rollup | The delivery layer; capital in multi-unit | Layer already captured; capital moving |
| Media & Entertainment | Rising Bar | Production collapsing to the price of tooling; IP and audience own the dividend; distribution captured upstream | IP and audience owners; the distribution layer | Working, slow |
| Tourism | Rising Bar under platforms | The experience is local and felt; distribution went upstream twenty years ago | The booking layer | Buying, fast: agentic booking |
| Split and swing | ||||
| Professional Services | Every shape at once | Bookkeeping, tax, and compliance in Rollup; audit in Slow Melt; advisory holds a Rising Bar | Capital on the routine half; customers on the judgment half | Both |
| Financial Services | Two speeds | The product layer is repricing toward the floor; advice holds the bar | The cheapest product operator; advisors keep theirs | Working, slow; the index-fund precedent |
| Retail | The swing state | Commodity SKUs to price competition under the marketplace layer; brand, curation, and experience hold the bar | The marketplace layer, on commodity | Buying, fast: the customer’s agent |
| Exceptions | ||||
| Pharmaceuticals | The patent exception | Winner-take-all by patent for a century; AI changes who reaches it first | Whoever reaches the patent first | Working, slow: development speed |
| Utilities | The regulated exception | No selection gate; data-center load reshaping the demand curve | The regulator decides, in rate design | Neither; argued in rate cases |
The Map, Sector by Sector
What follows is every industry we serve, located by its center of gravity, with the split that matters most inside it. Sectors are grouped the way we group our practice, so you can find yours the way you would find it in our menu.
Business and Professional Services
Professional Services · every shape at once. This is the framework’s home terrain, because every shape lives here at once. Bookkeeping, tax preparation, and compliance work are specifiable and fragmented, which is why consolidation capital is buying accounting firms at a pace the profession has never seen. Audit is the canonical Slow Melt, specifiable work still sheltered by committee selection and brand. Advisory and judgment work holds the Rising Bar. The firms that thrive will industrialize the routine half of the book before someone else prices it for them, and elevate the judgment half where durable margin lives.
Legal · Rising Bar. No model will argue your client’s case, and accountability keeps counsel firmly under the Rising Bar. But the layers that once subsidized the practice, research, review, and routine drafting, are already commodity, and volume practices in personal injury, immigration, and estates are Rollup targets wherever regulation permits. The quiet threat is corporate legal operations learning to buy specifiable matters by the spec, which opens the Shop it gate one matter type at a time.
Finance
Financial Services · two speeds. The product layer, standard lending, asset allocation, transaction processing, is repricing toward whoever runs cheapest, while advice keeps its margin behind trust. That gap between two speeds is the strategy window. Index funds already demonstrated what happens when a financial product becomes fully specifiable, and AI extends that logic to more of the product shelf every year.
Banking · Slow Melt. Deposit stickiness has protected pre-AI economics longer than most industries get, which makes banking a deep Slow Melt. The shelter holds until the customer’s agent starts shopping rates, and that clock has started. The banks that use the protected years to industrialize behind the trust wall will meet algorithmic selection on their own terms. The ones that spend the window congratulating themselves on retention will meet it on someone else’s.
Credit Unions · Slow Melt. Everything true of banking is truer here, with one addition worth saying plainly: the cooperative structure means the dividend, once captured, returns to the very members the megabanks would take it from. That is a real strategic asset in the era ahead. The question is whether shared-scale AI, through CUSOs and consortia, arrives before agentic rate shopping does.
Insurance · Rollup. Consolidation capital considers this sector its favorite meal: specifiable work, fragmented agencies and MGAs, operations ready for AI. Personal lines already live near the Falling Floor, selected by aggregators on price. Specialty and complex commercial lines hold the Rising Bar behind broker judgment. For agency owners the practical fact is this: your AI posture is being priced into your multiple right now, whether or not you plan to sell.
Family Offices · Rising Bar, sheltered. The most sheltered corner of finance on the map. Discretion, bespoke scope, and trust close the first two gates almost completely. The bar still rises: reporting sophistication that impressed principals two years ago is now the baseline, and the office that cannot answer a question in an afternoon that peers answer in a minute will feel it.
Private Equity · Rising Bar, and the capital. Sponsors sit under the Rising Bar for themselves, and play the capital for everyone else. Every fund now screens everything with AI, so screening is no longer edge. The edge has moved to proprietary deal flow and to the operating playbook that makes the dividend show up in portfolio-company EBITDA rather than in a board slide. This map doubles as a hunting guide; the Rollup entries on it are the sectors where consolidation economics are improving fastest.
Real Estate
Commercial Real Estate · Rising Bar over melting layers. Deals still close on relationships and capital trust, which keeps deal-making under the Rising Bar. Underneath it, underwriting tools are commoditizing, appraisal sits in a Slow Melt of credential-sheltered specifiable work, and property management is consolidating under Rollup economics. The judgment layer must carry more of the margin every year.
Residential Real Estate · Slow Melt. Most of the transaction is now specifiable work, still selected by referral, which is the textbook Slow Melt. Commission pressure and agentic buying are prying at the Shop it gate simultaneously, and title and escrow are melting already. The agents who survive will be the ones selling judgment and advocacy rather than access to a process a model can run.
PropTech · Falling Floor. The one software-native sector we serve, and software was always winner-take-most. What changed is that AI made software easy to build, which means features stopped being a moat. Defensibility now lives in proprietary data and workflow lock-in, and PropTech firms without one of those are discovering that their category’s winner-take-all prize can be taken by someone else.
Healthcare and Life Sciences
Healthcare · three shapes at once. No one delegates care to a model, and licensure guarantees it, which anchors care delivery under the Rising Bar. But two other stories are running underneath. The revenue cycle beneath care is industrializing fast, and the dividend from it is leaking upstream to vendors. And the practices themselves are consolidating under Rollup capital. Health systems and practice owners are playing three shapes at once, and most are only watching one.
Biotech · Rising Bar, floor upstream. AI gives everyone more shots on goal, which quietly raises the pipeline bar for all. Discovery velocity is now table stakes; the Falling Floor economics live upstream, in the discovery platforms. For clinical-stage companies the strategic question is which capabilities to own versus rent, because renting the same intelligence as every competitor confers exactly no advantage.
Pharmaceuticals · the patent exception. Here is the map’s most interesting exception: pharma has been winner-take-all for a century, one molecule at a time, because a patent is a state-granted monopoly. AI does not change the prize. It changes who reaches it first. Development speed, trial design, and evidence generation are the races; the patent office holds the trophy.
Industrial
Manufacturing · Rising Bar, commodity exposed. Capacity and capital keep any one player from taking it all, so no manufacturer needs to fear a monopolist. What commodity manufacturers should fear is their customers. When procurement agents source on spec and price alone, commodity lines become price-takers overnight, and the Shop it gate opens from the demand side. Differentiated and custom manufacturing stays sheltered under the Rising Bar, where operational AI is now simply the cost of quoting competitively.
Aerospace · Rising Bar, slowest line. Certification freezes the competitive line more firmly here than in any sector we serve, which makes aerospace the slowest-moving territory on the map. That is real protection. It is also why the supplier tiers beneath the primes are quietly consolidating, because Rollup capital loves a market where the customers cannot easily switch and the targets cannot easily scale.
Logistics and Supply Chain · Falling Floor edge. The most exposed sector on this map, and we say that with respect for how hard the work is. Matching loads to capacity is specifiable, rate-selected, and network-driven, which is the exact recipe for a Falling Floor. Brokerage and matching live nearest the edge; physical warehousing and 3PL fragment into Rollup territory; network design and complex orchestration hold the Rising Bar. If you are in this sector, the strategy conversation is not about whether to move. It is about which of three moves to make, and the window for choosing is real but not long.
Oil and Gas · Rising Bar, Rollup services. Nobody takes all in a commodity market, but efficiency decides who survives the next cycle, and the AI bar ratchets up with every downturn without ever coming back down. Oilfield services, with bid-selected scopes of work, run on Rollup economics. The back office is pure dividend capture waiting to happen.
Utilities · the regulated exception. The exception that proves the entire framework. Your customers cannot leave, so there is no selection gate at all, and the regulator, not the market, decides who keeps the AI dividend. That turns AI investment into a rate-case argument as much as a technology decision: prudency, reliability gains that survive commission scrutiny, recognition in rate design. Meanwhile data-center load growth is remaking your demand curve from the outside. Different playbook entirely, and mostly not the one vendors are selling you.
Consumer Sectors
Retail · the swing state. The swing state of the whole map, and the largest block of jobs that could change shape. The moment your customer’s agent does the shopping, loyalty stops being a feeling and becomes a filter setting. Commodity SKUs drift toward pure price competition, where the marketplace layer above you takes the margin. Brand, curation, and experience hold the Rising Bar. The work ahead is a portfolio exercise, SKU by SKU: industrialize the commodity side, and invest the savings where differentiation actually lives.
Airlines · oligopoly by structure. Slots, fleets, and capital close the Sole-source it gate: they keep the market from concentrating past an oligopoly, and no algorithm changes that. Revenue management is the purest table-stakes arms race in the economy, run by machines on all sides and won by none. The fight that decides margin is distribution, and the next booking channel will not be a person. Whoever the traveler’s agent books through takes the toll.
Food Service · Rising Bar under platforms. Nobody automates the reason guests walk in, which keeps hospitality itself under the Rising Bar. But the platforms already took the delivery margin, and the app rankings that decide your visibility are a selection algorithm you do not control. Multi-unit operators face a second front: back-office industrialization is what Rollup consolidators bring with them, so it had better be what independents build for themselves.
Media and Entertainment · Rising Bar. Production tasks are collapsing to the price of tooling while hits stay stubbornly unpredictable, which is the Rising Bar in its purest creative form. The dividend flows to whoever owns the IP and the audience, not whoever renders the frames. Distribution was captured upstream years ago, and the power law of attention does its own winner-take-most work on the demand side regardless of what any studio does.
Tourism · Rising Bar under platforms. Distribution went to the platforms twenty years ago, and agentic booking will deepen that capture; this is the industry whose history proves the whole framework. The experience is yours, physical, local, and felt, and it holds the Rising Bar. The margin on reaching the traveler is the fight, and it is fought upstream of your front desk.
Education
Education and EdTech · Slow Melt. Accreditation and credential selection shelter the institutions, which makes education a deep Slow Melt, slowed further by public funding and politics. But the starkest upstream capture on the map is happening here: AI tutors are going straight for the teaching layer itself, beneath the credential. The moat holds. The castle it protects is being rebuilt upstream, and the institutions that thrive will be the ones that put the new teaching layer inside their own walls instead of watching students rent it elsewhere.
How We Score It
Each sector is scored on the three customer gates as dials, not switches: how fully buyers can specify the work, how algorithmically they select, and how far the market can concentrate. The score describes the sector’s center of gravity, meaning where most of its revenue and employment sits, and every entry names the activities that sit elsewhere, because no company lives in one shape. Where a sector’s dividend is decided by something outside its market, we say which force: the layer above, the regulator, or the capital.
The weighting uses Bureau of Labor Statistics payroll employment, which counts jobs, not profit pools. The line-speed call names which boundary is moving in each sector, the working line or the buying line, because that is what determines how much time a sector has. In the language of the underlying economics these are regime shifts; we call them line moves because that is what they look like from inside a company.
The scoring is ours, informed by what our bench of several hundred sector-matched former CIOs, CTOs, and CISOs reports from inside these industries, and it will be argued with. Good. Argue with us. The exploratory conversation is where the map gets tested against a real company, and that testing is how next year’s edition gets better.
The Map Will Move
The annual refresh is a commitment, not a flourish. Each year we will rescore every sector against the three gates, publish which industries migrated toward a boundary, reweight the map against fresh federal employment data, and fold in what our bench reports. The rescoring tracks the working line. The buying line can flip a sector between editions, and we will call those flips when we see them. Some of what this first edition says will be wrong by the third. That is what maps of moving weather are for.
For the 2027 edition, five sectors sit closest to a boundary, and all five sit on the buying line: Retail, where the customer’s agent decides whether loyalty is still a feeling; Banking and Credit Unions, where rate shopping by agent ends the melt; Residential Real Estate, where commission pressure and agentic buying are prying at the same gate; and commodity Manufacturing, where procurement agents source on spec. Watch those five for the first evidence that the buying line has moved.
Until then, find yourself on the map twice, once for your market and once for the mix inside your walls, and then read what we would do in your seat under each shape. If you would rather do that reading out loud, with someone who has run a company in your shape, that conversation is where every engagement of ours begins.
