Manufacturing
This sector comprises establishments primarily engaged in the chemical, mechanical or physical transformation of materials or substances into new products. These products may be finished, in the sense that they are ready to be used or consumed, or semi-finished, in the sense of becoming a raw material for an establishment to use in further manufacturing. Related activities, such as the assembly of the component parts of manufactured goods; the blending of materials; and the finishing of manufactured products by dyeing, heat-treating, plating and similar operations are also treated as manufacturing activities. Manufacturing establishments are known by a variety of trade designations, such as plants, factories or mills. — Statistics Canada, NAICS 2022A
- Establishments · CanadaA
- 51,934
- Under 10 employeesA
- 54%
- Establishments · USA
- 285,500
- Employment · USA
- 12,188,330
- Payroll · USA
- $851.3B
Size and shape
How many businesses there are and how small they are. Fragmentation is the first thing an entrant — or anyone selling software into this industry — needs to know, and it is one of the few things that is actually measured.
Canadian establishments by number of employeesA
Of 51,934 Canadian establishments with employees, 54% have fewer than ten — mostly small operators.
Where they areA
Largest four provinces by establishment count. Establishments with employees only — sole operators with no payroll are not in this table, so in trades and personal services the true number of businesses is higher.
Measured, not forecast: the share of US establishments opening in one year that were still active later. It counts good operators and bad ones together, which is exactly why it is the honest answer to “what are the odds”. It is for the whole sector rather than this market, and the ten-year figure comes from an older cohort because no younger one has reached ten years.
Market screens and studies
Market-entry records filed along this branch of the hierarchy. A record at or beneath this code is about this industry; one above it is about something wider that contains it.
Shelf space is bought, not won. A new CPG brand pays slotting fees, funds trade promotions, carries the retailer's payment terms and absorbs returns before it learns whether the product sells, and co-manufacturing minimums put the first production run in the tens of thousands of units. The direct-to-consumer route that was supposed to bypass all of that now costs more in paid acquisition than the margin on a grocery-priced item supports. The product content software this industry buys is screened separately at the same code.
The pre-screen dismissed this as feed milling and plant capital. The better half of the code deserves a harder look. Freshpet, the listed pure play in refrigerated pet food, reported 2025 net sales of $1,102.0M, up 13.0%, at a 40.8% gross margin [A]. That is the attraction. The cut is what the growth cost. Freshpet's capital expenditure was $239.1M in 2023, $187.1M in 2024 and $148.2M in 2025 — $574M in three years — and 2025 was its first year of positive free cash flow, $12.4M [A]; it guides to roughly $150M of capital again in 2026 [A]. Over those three years it generated $390.8M from operations [A] and put half as much again into plant. A company with a billion dollars of sales and a differentiated product was still spending more on kitchens than the business threw off. Pet food is made on extrusion, retort or chilled lines in facilities inspected to a standard retailers and export markets will accept, and the product has to be on the shelf, in volume, every week. The Canadian shape says the same thing from the other side: the largest single size band among the 434 establishments is 20–49 employees (31%) [A] — these are plants, not workshops. The way around the plant is a co-packed treat or kibble brand, which is no longer manufacturing; it is the Consumer Packaged Goods Brand Operation already screened at 311, with that record's shelf-access problem intact. Livestock feed, the other half of the code, was not researched.
This is a growing industry, which is not what the pre-screen's "commodity margins" suggests. Statistics Canada reports that Canadian processors crushed a record 11.6 million tonnes of canola in 2025, up 1.6% and the third consecutive annual record, and that domestic crushers took 50.5% of all canola disposition in the 2024/25 crop year against 41.5% exported as seed [A]. More of the crop is now processed where it is grown than leaves the country as seed. The cut is the size of the ticket to take part. When Cargill announced its Regina crush plant it was reported as a $350M facility with one million tonnes of annual capacity [B] — at least $350 of plant for every tonne of yearly throughput, before a tonne of seed is bought. And the plant, once built, sets no price: seed is bought against a futures market, oil and meal are sold against theirs, and the crusher earns the spread the board leaves. That rewards the lowest-cost, best-located, largest plant with its own rail and origination, which is why the builders are global grain houses. Flour milling, wet corn milling and breakfast cereal share the shape: US establishments in this group average 63 employees [A]. The 84 Canadian establishments with fewer than ten employees [A] are stone mills and cold-press oil makers. What they sell is a premium bag of flour or bottle of oil, and that is a packaged-goods brand — screened at 311 — that happens to own a mill.
The pre-screen saw a craft-scale entry with the shape of the bakery record at 3118, and the counts support that there is one: 169 of the 330 Canadian establishments have fewer than ten employees [A]. The sugar end of the code is simply closed — Rogers Sugar is spending $280–300M to add 100,000 tonnes of refining capacity to a business that sold 781,454 tonnes in fiscal 2025 [A], and nobody enters refining beside that. Chocolate and candy are where an entrant can start, and the test is what making them earns. Rocky Mountain Chocolate Factory is a long-established listed brand with its own factory and a captive franchise network to sell through. In fiscal 2026 it reported total revenue of $27.5M, down from $29.6M, and product and retail gross profit of $0.7M — after $0.1M the year before — with a net loss from continuing operations of $4.6M [A]. With the brand, the plant and the channel already in hand, manufacturing confectionery earned it almost nothing. The mechanism differs from bread. Confectionery is bought a few times a year as a gift, against a shelf price set by multinational brands that buy cocoa and sugar at scale, while the small maker buys the same world-priced inputs in pallets. The premium a customer will pay is real but seasonal and capped, and it is paid at the counter — so what works is a shop with a kitchen behind it, which is specialty food retail rather than manufacturing, and is won or lost on the lease.
The pre-screen pointed at small-batch preserving reached through co-packers, and the small end exists: 163 of 376 Canadian establishments have fewer than ten employees [A]. But the group is unusually top-heavy — 60 establishments, 16%, employ a hundred or more [A], and the average US plant has 83 staff [A] — because freezing and canning are harvest-timed, high-throughput operations contracted to growers a season ahead. The question is whether the category rewards anyone for being in it, and the cleanest recent answer is Green Giant. B&G Foods reports a Frozen & Vegetables segment that is primarily that brand. In fiscal 2025 the segment's net sales fell 9.4% to $358.6M and its adjusted EBITDA was –$0.3M, down from +$9.5M [A], on lower volumes, more trade promotion and higher raw-material cost. B&G took $34.8M of impairments on Green Giant in the fourth quarter alone, sold Le Sueur, moved Green Giant Canada to assets held for sale, and closed the sale of the Green Giant US frozen line on 2 March 2026 [A]. One of the best-known names in the freezer aisle earned nothing on a third of a billion dollars of sales, and its owner's response was to leave. When the category is shrinking under private label, a new processor is not competing for growth. It is competing to be the lowest-cost supplier of a retailer's own brand, against plants that are already depreciated. The small-batch jam, pickle or sauce maker sits outside that fight, but it is a brand selling through shelf and market stall, and is screened as Consumer Packaged Goods Brand Operation at 311.
Dairy processing in Canada is a good business to already be in. Saputo's Canada sector reported fiscal 2026 revenues of $5.423B, up 5.0% on higher volumes, with adjusted EBITDA of $697M — a 12.9% margin [A]. The same company's consolidated margin across all its sectors, most of them outside supply management, was 9.5% [A]: its regulated home market is where it earns the most. That is how the system is built. Under supply management a processor does not negotiate for milk. It buys from a provincial marketing board at an administered price, by end-use class, in a volume the board allocates, and imports that would undercut the result are held out by tariff-rate quotas. Everyone pays the same for the input and no one can be undersold from abroad, so margin goes to whoever has the plant scale, the allocation history and the retail listings. For an entrant that stability is the wall. A new plant needs a licence, an inspected facility and a milk allocation before it sells a litre, and the volume available to a newcomer is what the board's new-entrant or artisan programme releases, not what the business plan calls for. The small end is real — 153 of 434 establishments have fewer than ten employees, and Quebec alone has 150 plants [A], most of them fromageries — but those are farm-linked cheesemakers selling a regional premium, often processing their own quota milk. Reaching that position means entering dairy farming, where the quota is the capital. The pre-screen said supply-managed and plant-heavy; the first half is the one that binds.
There is a real and repeatedly documented shortage of provincially inspected slaughter capacity, particularly in British Columbia, and that shortage persists precisely because the barrier is high: inspected facility construction, effluent handling, HACCP plans and inspector availability all precede revenue. The demand signal is strong and the capital and compliance ramp is what keeps clearing the field.
The pre-screen paired small-scale processing with the fishing records, and the pairing is the problem. Commercial Fishing Licence & Quota (114) found that the right to catch has been capitalised into an asset its owners rent out. The plant is the next link along, and it owns neither end of the chain. Upstream, fish comes from licence and quota holders, many tied to a buyer by financing, or from integrated companies that land their own — when Premium Brands and a Mi'kmaq coalition took Clearwater private in 2021 at $8.25 a share, about $1B including debt [A], they were buying a company built on its own offshore licences. Downstream sit grocery and foodservice buyers. High Liner Foods, the listed Canadian processor with the strongest brand on that side, shows what the middle earns. In fiscal 2025 sales rose 7.1% to US$1,026.9M while volume rose 0.9% — the growth was price and mix, passed through from raw material and tariffs — and gross margin fell from 22.7% to 20.7%, adjusted EBITDA margin from 10.8% to 8.9% [A]. If a national brand cannot hold its margin when input costs move, a custom processor with no brand and no quota will not either. It is paid a processing fee set by whoever controls the fish. Nor is this a small-shop industry. Only 29% of the 372 Canadian plants have fewer than ten employees, while 40% have fifty or more [A] — seasonal lines with large crews, concentrated in Nova Scotia and Newfoundland and Labrador, where a processing licence is itself a provincial decision. The value in seafood sits in the quota and on the shelf, and the plant in between rents its margin from both.
Retail price resistance on a staple sets the ceiling and ingredient and labour costs set the floor, and the gap has narrowed every year. Wholesale volume looks like the answer and brings a grocery buyer who dictates price and delivery windows. The successful version is a café with an oven, which is a different business with hospitality's cost structure.
This is where small food makers actually file: 1,717 Canadian establishments, 61% of them with fewer than ten employees [A], across snack foods, coffee and tea, flavouring syrups, seasonings and dressings, and an "all other" remainder. It is a residual, and this screen examined one niche properly — coffee roasting, the most commonly entered — and did not examine the rest. Roasting is easy to start and carries a good margin. Farmer Bros., a US roaster founded in 1912, reported a fiscal 2025 gross margin of 43.5% on net sales of $342.3M [A]. It also reported operating expenses of $150.4M against gross profit of $148.9M, and a net loss of $14.5M [A], and in March 2026 agreed to be bought by Royal Cup for $1.29 a share in cash [A], a deal reported at about $32M in total [C] — under a tenth of one year's sales. The roasting made money. The expense was the route: branches, trucks, drivers and brewing equipment placed in thousands of restaurants and offices, which is how wholesale coffee is sold and serviced. A roaster with a century's head start on that network could not make it pay, and sold for a fraction of revenue to a competitor seeking route density. The entrant's version is the same problem smaller. Beans, a roaster and a bag are not scarce. The account list is, and it is built one café, one grocer and one office at a time against whoever already delivers there. The 311 brand record cuts on the cost of shelf access, and this one cuts a step earlier, on getting the product to a customer at all.
Category volumes are falling, the shakeout among small producers is well under way, and the unit economics only work through a taproom — which makes this a hospitality business with a brewery attached, carrying both cost structures. Excise, packaging and distribution access all favour the incumbents that already own shelf space.
The pre-screen said excise and licensing close this to new entrants. That is right, though the licence is not the part that does it. There are 16 tobacco manufacturing establishments in Canada [A]. Statistics Canada's own monthly table shows what they sell into: cigarette sales fell from 22.9 billion in 2010 to 12.3 billion in 2025, down 46% in fifteen years [A], and the decline has not levelled off — June 2024 sales were 16.8% below June 2023 [A]. The three large incumbents had creditor-protection plans approved by the court in 2025, built on a $32.5B global settlement, payable over time out of their operating profits [B]. An entrant carries none of that liability, which on paper is the only interesting opening in the industry: the same excise, the same retail price, and no settlement levy on the margin. It does not survive contact with the law. A federal excise manufacturing licence, stamping and bonding are the entry cost, and they are manageable. What cannot be overcome is that Canadian law requires plain, standardised packaging and prohibits nearly all promotion, so there is no lawful way to tell a smoker that a new brand exists or why to choose it. Brands that predate the rules keep their customers by habit, and a brand created after them cannot acquire any. The ten establishments with fewer than ten employees [A] are not evidence of an open door; this screen did not establish what they are, and at that size none of them is a challenger to the three incumbents. With sales down by nearly half in fifteen years there is also no growth to come in on.
81 more sit deeper in this branch — open a row under “Inside this industry” to reach them.
Software serving this industry
The vertical software markets filed along the same branch — who sells to these businesses and who they would have to displace — and then the generic categories every business buys whatever it does.
The product is a retailer integration map, and it takes years and a partner team to build. Salsify raised about $200M at a $2B valuation on the strength of syndicating content to more than 1,000 retailers for brands including P&G, Coca-Cola, Mars and Kraft Heinz; Syndigo holds the other half of the same network. A new entrant must rebuild those connections one retailer at a time while each retailer keeps changing its content requirements — the same moat the EDI record at 419 describes, in a different format.
FSMA 204 gives the category a dated regulatory trigger, which is the strongest demand signal in this batch — but the buyer is reached through GFSI certifying bodies and third-party auditors, not through search or self-serve. That channel is owned by the auditors and by the incumbents who sponsor them, and an entrant with no audit relationship has no route in. Nothing is disclosed here. Every vendor named on this record is private, or sits inside a parent that does not break the line out, so no revenue floor can be built and the market size is genuinely unknown rather than estimated.
Food and beverage ERP is the system of record a food maker runs the plant and the books on: formulas and recipe scaling, lot genealogy from ingredient receipt to finished case, catch weight, shelf life and expiry, allergen and nutrition data for the label, co-packer runs, retailer EDI and the traceability records regulators ask for. The anchor buyer is NAICS 3119 (snack foods, coffee and tea, seasonings, dressings and the rest of 'other food'), but the same software is sold across 3111–3118 (animal food, grain and oilseed, sugar and confectionery, fruit and vegetable preserving, dairy, meat, seafood, bakery) and 3121 (beverages). The market is held by industry-specialist roll-ups and by the horizontal giants' food editions, and both are well funded. Aptean is the clearest specialist consolidator: it bought IndustryBuilt of Mississauga, Ontario — maker of JustFood, a food and beverage ERP on Microsoft Dynamics — in March 2018, with more than 130 staff [A, Aptean release], and now sells a food and beverage ERP built on Dynamics 365 Business Central alongside its process-manufacturing lines [C]. Its owners are TA Associates (lead shareholder since October 2022), Insight Partners, Charlesbank and, since July 2024, Clearlake Capital; it says it serves 10,000+ customers in 80+ countries across all its verticals [C]. ECI Software Solutions bought Deacom (Chesterbrook, Pennsylvania), an ERP for batch and process manufacturers, in September 2021 [A]; ECI is co-controlled by Leonard Green and Apax, with a new GIC investment, since February 2025 [A]. Infor sells CloudSuite Food & Beverage and has been wholly owned by Koch Industries since April 2020 [A]; Rockwell Automation bought Plex for $2.22B in cash (about $2.15B after adjustments) in 2021 [A] and markets it to food and beverage plants for traceability and compliance [C]; Acumatica agreed in May 2025 to be acquired by Vista Equity Partners [B]. SAP, Oracle NetSuite, Sage X3 and Microsoft Dynamics (through partners such as Aptean, Inecta, Vicinity and BatchMaster) are present everywhere [C]. Below them sit long-tenured specialists — BatchMaster (Irvine, California; '40+ years' of process-manufacturing practice, sold standalone or bolted onto QuickBooks, Sage, SAP Business One or Dynamics) [C], SYSPRO [C], Inecta (New York, with an office in Hamilton, Ontario; founded 2001; Business Central-based food ERP) [C] and Vicinity Software (Marietta, Georgia; extends Dynamics for food, chemical and brewing formula manufacturers) [C] — and at the small end, general manufacturing tools such as Katana ($35M Series B led by Northzone, 2022) [B] and Wherefour [C]. Regulation is the obvious opening, and it has moved away. FDA's Food Traceability Rule (FSMA 204) required compliance by 20 January 2026; FDA proposed a 30-month extension to 20 July 2028, and the Continuing Appropriations … Extensions Act of 2026 then directed FDA not to enforce the rule before that date [A, fda.gov]. The forcing date is now almost two years out. In Canada, the Safe Food for Canadians Regulations already require one-step-forward, one-step-back traceability records, a lot code on the label and two years of record retention [A, inspection.canada.ca] — a requirement incumbents already ship for, not a new one. Lock-in is the decisive fact. A food maker's ERP holds every formula, costed bill of materials, allergen matrix, lot history and retailer EDI map; changing it means re-validating labels and recall procedures in a plant that ships daily. The independent Canadian specialist that existed — IndustryBuilt — is already inside Aptean. Incumbent vulnerability decides it: there is no unserved segment, the regulatory trigger slipped, and the specialists are backed by TA, Clearlake, Leonard Green, Apax, Koch and Rockwell. How this differs from its neighbours: 311-food-safety-and-traceability-software is the compliance layer that sits beside the ERP (Trustwell, TraceGains, SafetyChain — supplier documents, audits, HACCP records), not the inventory, costing and production system of record; 3241-process-manufacturing-and-refining-software is continuous-process modelling for refineries and chemicals (AspenTech, AVEVA); 541514-erp-enterprise-resource-planning is the horizontal ERP market, whose vendors appear here only in their food editions; 418-wholesale-distribution-erp-and-route-accounting is distributors' ERP (rebates, route settlement), where Aptean also appears. There is no brewery, winery and distillery software record; 3121-craft-beverage-production is the operator record. Tecsys (Montreal) was checked as a possible Canadian vendor and is a warehouse and supply-chain software house serving healthcare and distribution, not a food ERP. Nulogy (Toronto) is named because it is the Canadian software that co-packers and contract manufacturers for CPG brands run on, but it sits beside the ERP rather than replacing it.
The buyer pool is small, shrinking at the brewery end, and has already been sold to three consolidators — so an entrant would be fighting owners with payment and distribution networks for a few thousand customers. TTB's own count of brewery permits that filed an operational report fell from 8,199 (2022) to 7,888 (2024) and 7,228 (2025, before late filings); in 2025, 4,781 of them made 1,000 barrels or less and 1,040 made nothing at all [A]. Distilled-spirits permits are still rising (4,036 at the start of 2022, 5,069 at the end of 2024) [A], but they are the smallest pool. In Canada the whole beverage-manufacturing group is 1,862 establishments, 859 of them with fewer than ten employees (per the 3121 operating record) [A]. The vendors have consolidated around those buyers. Brewery and distillery production belongs to Ekos — bought by Next Glass (Untappd, Untappd for Business, Ollie; backed by PSG) in October 2025, with 1,200+ active producers and a two-year price freeze for existing customers [B]. Winery direct-to-consumer (ecommerce, POS, wine clubs, reservations) belongs to Commerce7, a British Columbia corporation owned by Fullsteam, which bought WineDirect's SaaS division in January 2025, VineSpring in 2022 and WinePulse in 2025, and says 3,500+ wineries use it [B/C]. Winery production is split between InnoVint and vintrace — and vintrace now sits inside Encompass Technologies, which VIP (the beverage-distributor software house Warburg Pincus agreed to buy for about $1B in February 2025) acquired on 14 September 2026 [B]; OrchestratedBEER's old domains now redirect to Encompass too [C]. The compliance work is real but it is a feature, not a moat an entrant can own: US producers file the Brewer's Report of Operations (TTB F 5130.9, or quarterly 5130.26), the Report of Wine Premises Operations (5120.17), the distilled-spirits monthly production, storage and processing reports (5110.40, 5110.11, 5110.28) and the excise return (5000.24) [A]; in Canada the CRA levies beer duty per hectolitre under the Excise Act and spirits and wine duty under the Excise Act, 2001, with the reduced rates on a domestic brewer's first 15,000 hL cut in half and the annual CPI adjustment capped at 2% for two more years by Bill C-30 (royal assent 18 June 2026) [A]. Every incumbent already generates these. How this differs from its neighbours: 3121 Craft Beverage Production is the operating business (taps, shelf listings, excise as a cost); this record is the software that business buys. 722511 Restaurant Operations Software is Toast's market — it meets this one only at the taproom POS, where Arryved sells against Toast and Square. 3241 Process Manufacturing is refinery-scale simulation; the national brewers in Canada run enterprise ERP and are not buyers of anything here. Market size decides it: a few thousand mostly sub-1,000-barrel producers, already captured by owners who earn on payments (Fullsteam, Next Glass PaySmart) and distribution data (VIP).
Metrc's position is a state procurement contract, not a product win — a licensee in a Metrc state uses Metrc because the regulator says so, and the commercial software sold around it is a connector. Winning here means winning state RFPs against an incumbent with a decade of regulator relationships, which is a government-sales business rather than a SaaS business. Nothing is disclosed here. Every vendor named on this record is private, or sits inside a parent that does not break the line out, so no revenue floor can be built and the market size is genuinely unknown rather than estimated.
Centric has consolidated apparel PLM and now sits behind Dassault's balance sheet, with the brand-side buyer standardised and the factory side served by whatever the brand mandates. The remaining independents are being bought rather than beaten. Nothing is disclosed here. Every vendor named on this record is private, or sits inside a parent that does not break the line out, so no revenue floor can be built and the market size is genuinely unknown rather than estimated.
The installed base is shrinking. Commercial print volumes have declined structurally for two decades and every seat gained is taken from a shop that will close or merge, so vendor revenue compounds negatively even with a better product. Ricoh and EFI hold the category as an attachment to press sales. Nothing is disclosed here. Every vendor named on this record is private, or sits inside a parent that does not break the line out, so no revenue floor can be built and the market size is genuinely unknown rather than estimated.
The models encode decades of process engineering that a new entrant cannot reproduce, and every deployment ties to the plant's existing control system. Buyers are a few hundred refineries and chemical complexes globally, each with safety-case review before any software touches operations. Sourced update: the two reference points in this market have both gone dark — AVEVA into Schneider, and AspenTech fully into Emerson for about $7.2B in March 2025. An entrant now competes against balance sheets it cannot see.
Veeva is the proof case every vertical SaaS pitch cites, and it now occupies its category completely. Validated-system requirements (GxP, 21 CFR Part 11) impose a compliance floor measured in years, not months.
The system of record for every adverse event a drug company hears about. Pharmacovigilance software takes in suspected adverse reactions from call centres, literature, partners, trials and social media; codes them in MedDRA and WHODrug; runs medical review; builds each Individual Case Safety Report (ICSR) in ICH E2B format and sends it to the regulators on the clock; detects signals across the case base; and assembles periodic reports (PSUR/PBRER, DSUR). How it differs from the three neighbouring 3254/5417 records: regulatory information management (RIM) tracks registrations and publishes the eCTD dossier — the application; clinical-trial data management (EDC, CTMS — Medidata, Veeva Clinical) runs the trial; life-sciences commercial software (Veeva CRM) runs the sales force. Safety runs for the whole life of the product after the first patient is dosed, and its output is a legally required report with a 15-day clock. The vendors overlap (ArisGlobal, Veeva, Ennov and EXTEDO sell both RIM and safety), so their funding is not re-told here. The regulators set the floor and just raised it. FDA's adverse-event system, renamed AEMS (formerly FAERS), requires E2B(R3) for premarket IND safety reports from 1 April 2026 and for postmarketing ICSRs sent through ESG NextGen from 1 October 2026, with E2B(R2) accepted only through 30 September 2026 [A, FDA]. EMA's EudraVigilance has required the ISO ICSR standard as implemented by ICH E2B(R3) since 30 June 2022, and electronic reporting is obligatory for marketing authorisation holders and trial sponsors [A, EMA]. Health Canada's Canada Vigilance guidance (effective 2018, clarified by a February 2026 notice) still aligns with E2B(R2), prefers electronic reporting through its Trading Partner Management Office, and sets 15 calendar days for serious domestic reactions [A, canada.ca]. Every safety database must be validated, connected to each authority's gateway and ready for inspection before it processes a single live case. Oracle Argus is the installed incumbent. Oracle's own documentation calls Argus Safety the core of Argus Standard and Enterprise editions, used to take in cases, code them, assess them and generate expedited and periodic reports; it now sells it as Safety One Argus with Safety One Intake and Empirica Signal [A, docs.oracle.com]. Oracle does not report the line, and oracle.com refuses automated access, so the installed base is UNVERIFIED — the evidence is indirect: Cognizant, which processes over 4 million case versions a year for clients, says it runs case processing on 'robust safety databases like Argus' and upgrades clients' Argus installs [C, Cognizant]. The one scale figure in the market is the one being sold. ArisGlobal (Waltham, Massachusetts; founded 1989; LifeSphere Safety) serves more than 200 enterprise customers, processes over 12 million safety cases a year and expects about $175M of 2026 revenue; Nordic Capital, owner since 2019, agreed on 23 July 2026 to sell it to Dassault Systèmes for about $1.8B in cash plus up to $200M tied to AI revenue milestones, closing expected in the second half of 2026 [A, Dassault and Nordic Capital releases]. That is roughly 10× expected revenue, and it puts ArisGlobal beside Medidata inside Dassault. Veeva Safety (intake, processing and submission, with SafetyDocs, Safety Signal and Safety Workbench) lists itself at 51–100 customers [C, veeva.com] inside Veeva's $1,748.4M R&D and Quality line (FY2026, of $3,195.3M total revenue) [A, 10-K]; Veeva's 10-K names IQVIA, Dassault Systèmes, OpenText and Oracle as its competitors in development applications. IQVIA sells the Vigilance Platform as SaaS alongside its Lifecycle Safety outsourcing [C, iqvia.com]. The second tier is private-equity owned and consolidating. Ennov (Paris; 650+ life-sciences customers across all its modules, including PV Case Management and Signal Detection) took a growth investment led by Bregal Sagemount with Ardian in May 2026, terms undisclosed [B, Goodwin]. Qinecsa (London; 1,300+ employees) was built from Commonwealth Informatics, Bioclinica's drug-safety business and MyMeds&Me and bought Denmark's Insife (HALOPV) in April 2024; it sells Reportum intake, HaloPV and Vigilance Workbench signal software alongside outsourced case work [A, company; B, Samedan]. EXTEDO's Safety Management Hub is a smaller line next to its eCTD business. Much of the spend is labour, and it flows to services firms. Cognizant says it processes over 4 million adverse-event case versions a year and has helped more than half of the top 10 pharma companies with their PV environments [C]; Accenture sells AI-driven pharmacovigilance as R&D outsourcing [C]; Genpact launched its PVAI case-processing automation in 2017 and spun PVAI out as an independent company in February 2023 [A, Genpact release; B for the spin-out]. Clarivate sells the data around the case — Drug Safety Triager for literature review and OFF-X safety intelligence [C]. The AI attack is real but already funded and already copied. Graph AI (Graph Safety) raised a $13.3M Series A led by Insight Partners with Bessemer in September 2026 for an AI-native intake, safety database and aggregate-reporting stack, and claims case turnaround cut from over three hours to under ten minutes [B for the round, C for the claim]. Every incumbent ships the same promise: Oracle Safety One Intake, ArisGlobal's NavaX generative case processing, Veeva AI in safety, IQVIA's AI-assisted processing. No Canadian vendor of note. Canadian marketing authorisation holders (Innovative Medicines Canada has 40+ member companies) buy the same global databases and file to Canada Vigilance from them. Entry cost and regulatory drag decide it: a new safety database must be validated, certified on the FDA, EMA and Health Canada gateways and survive inspection before its first live case; a sponsor moves its case history only in a multi-year validated migration; the top tier is now a $1.8B Dassault asset, a $1.75B-line Veeva product and Oracle's installed base; and the only opening — AI-native case processing — is already a venture-funded race the incumbents are matching feature for feature.
Veeva owns the top of this market, and the rest of it is being bought. Veeva RIM (Registrations, Submissions, Submissions Publishing, Submissions Archive) sits inside Veeva's R&D and Quality Solutions line, which earned $1,748.4M in FY2026 (subscription $1,426.6M plus services $321.8M, year to 31 January 2026) [A, 10-K]. Veeva does not break RIM out, and it names IQVIA, Dassault Systèmes, OpenText, Oracle and Honeywell as the vendors that compete with parts of its Development and Quality clouds [A, 10-K]. The second tier has changed hands three times in thirty months. Bertelsmann bought EXTEDO (Munich, founded 1996, eCTDmanager and the EXTEDOpulse RIM platform) in February 2024 [A, EXTEDO's own about page]. That same month Ennov (Paris), which also sells PV, quality and clinical software, bought Calyx's Enterprise Technology division, whose RIM and InSight publishing tools Ennov still sells [B, Manatt release; C, Ennov's site]. On 23 July 2026 Dassault Systèmes agreed to buy ArisGlobal (LifeSphere: RIM, submissions, labelling, safety) for about $1.8B in cash plus up to $200M in earn-outs. ArisGlobal is expected to make $175M of revenue in 2026 across safety, regulatory, quality and medical affairs, and claims 200+ customers, half of them among the top 50 biopharmas [A, Dassault release]. That deal prices a mixed safety-and-regulatory suite at roughly 10x revenue. On authoring, Certara sold its Regulatory and Medical Writing business (the old Synchrogenix) to Veristat. It was announced at up to $135M and closed on 8 May 2026. That business earned $50M of revenue and $17M of adjusted EBITDA in 2025 and had about 220 staff [A, 8-K]. The writing work there was mostly done by people, not by software. LORENZ (Frankfurt; docuBridge) claims 2,000+ paid installations in 48 countries, including 19 regulators, Health Canada and the FDA among them [C]. IQVIA sells SmartSolve RIM (launched as RIM Smart in 2019) for MedTech, IVD and pharma [C]. DXC still sells FirstDoc, FirstPoint, TRS, Writer and ToolBox [C]. The start-ups sit below all of these. Rimsys (MedTech RIM) raised a $16M Series A led by Bessemer in 2021 [A, company post]. Kivo targets clinical-stage biotech. Yseop (Paris) sells AI-written CSRs and CTD documents. Freyr mixes regulatory services with its freya fusion software. The regulators set the switching moments. FDA has accepted new applications in eCTD v4.0 since 16 September 2024 but has set no date for making it mandatory [A, fda.gov]. EMA made v4.0 optional for new centralised MAAs on 22 December 2025. It becomes strongly recommended from Q1 2027 and mandatory from Q1 2028; the forward-compatibility date for existing v3.2.2 dossiers is still pending [A, EMA eSubmission]. EMA's Product Management Service (PMS, the IDMP piece of SPOR) went live on 31 May 2024 [B, EMA guidance via search]. Health Canada's v4.0 Module 1 guide is still a draft from 2019 [A, canada.ca]. Each cutover is a reason to re-buy publishing software, and every incumbent already markets its v4.0 readiness. How it differs from its neighbours. The life-sciences commercial record is CRM and data: Veeva CRM against Salesforce. The clinical-trial record covers EDC, CTMS and eTMF. This record is the regulatory department's own system of record: what is registered where, what has been submitted, and the dossier itself. Pharmacovigilance (ArisGlobal's larger half, Ennov PV, Oracle Argus) and serialisation are separate records still to come. Incumbent vulnerability decides it. Veeva is growing and bundles RIM with its clinical and quality vaults, and three strategic buyers are absorbing the independents, so a new RIM vendor has an exit market but little open ground.
TraceLink is the network incumbent for US DSCSA data exchange, and the line-level systems are sold by packaging-equipment groups that keep consolidating. TraceLink (North Reading, Massachusetts; founded 2009) had raised $167M by its 2018 Series D, a $93M round led by Georgian Partners, and said it connected more than 270,000 drug companies, distributors, logistics providers, hospitals and pharmacies [B, MedCity News; network count C]. No later round or any revenue was found, and tracelink.com refuses automated retrieval. The other big seller is now part of a US-listed group. Antares Vision (Travagliato, Brescia) bought the assets of the liquidated French vendor Adents for about €1.5M in November 2020 [B, Pharmaceutical Commerce] and agreed to buy rfxcel for $120M in February 2021 [B]. Antares reported €90.4M of revenue in H1 2025 [B, Soldionline], across inspection and traceability, not serialisation alone. Crane NXT (NYSE: CXT) then bought Antares in stages, completing it on 31 March 2026 for about €362M in cash. It booked $98.9M of Antares backlog at that date [A, Crane NXT Q1 2026 release on SEC], and Antares was delisted from Euronext Milan. Systech now sits with Dover's Markem-Imaje (its site carries the Markem-Imaje mark [C]; the January 2020 deal pages refused retrieval). Movilitas, bought by Engineering Group in 2022, now redirects to Engineering's IndX division, which sells Movilitas.Cloud and SAP ATTP projects [C]. Tjoapack, a contract packager rather than a software vendor, is now part of Alcami [C]. Körber's PAS-X Track & Trace (March 2025 release, Lüneburg) runs on-premise, in the customer's cloud or as Körber-managed SaaS [C]. Optel (Canada, founded 1989) claims 6,000 systems installed and 25 billion products tracked a year [C]. SAP sells Advanced Track and Trace for Pharmaceuticals (ATTP) to its own ERP base. LSPedia (OneScan) claims 1,600+ companies and sells to pharmacies [C]. Axway, now part of 74Software, still sells an EPCIS repository product, Axway Track & Trace [C]. The mandate deadlines have mostly passed. DSCSA's enhanced drug distribution security requirements took effect on 27 November 2023, followed by a one-year stabilisation period. FDA's October 2024 exemptions then ran out for manufacturers and repackagers on 27 May 2025, for wholesale distributors on 27 August 2025, and for dispensers with 26 or more full-time employees on 27 November 2025 [B, Quarles summary of the FDA notice]. Small dispensers (25 or fewer pharmacists and technicians) are now exempt until 27 November 2027 [A, FDA exemptions page, content current 26 Aug 2026]. The EU's EMVS links the national systems of 29 countries to the European Hub, with 2,900+ marketing authorisation holders, 4,000+ wholesalers and 115,000+ commercial pharmacies connected [A, EMVO]. The compliance wave that funded this category is over in the EU and nearly over in the US, so the remaining buyers are mainly renewing systems they already run. The lock-in has three layers. Line-level serialisation (L1–L3) is GxP-validated equipment on packaging lines. The enterprise serial repository (L4) holds EPCIS connections to every trading partner. A network vendor like TraceLink adds partner-to-partner links on top. Replacing any layer means revalidating and re-onboarding counterparties. How it differs from its neighbours. Food safety and traceability (311) is lot-level FSMA 204 record-keeping, not unit-level serials with regulator verification. Life-sciences commercial software (Veeva CRM) and regulatory information management (RIM, submissions) handle sales and the dossier, not the pack. A warehouse management system runs the building and consumes the serial data. Incumbent vulnerability decides it: one network is entrenched, the line-level makers are now inside Crane NXT, Dover and Körber, and the last open segment (small US pharmacies, exempt to November 2027) already has LSPedia, plus wholesaler-provided tools.
Catalogued categories — named, not analysed
And what every business buys · 33 generic categories
Sold to every industry rather than this one, so they are filed against the software industry's own code. The same few vendors recur across most of them.
Companies in this industry · 622
Every company this research names that is filed here or beneath — the operators, and the vendors that sell to them — largest disclosed revenue first. The rank is within the company’s own six-digit industry.
| Company | Filed under | Revenue | Rank |
|---|---|---|---|
| Smurfit WestrockPrivate | Converted paper product manufacturing3222 | $31.2B | 1/4 |
| MagnaNYSE:MGA | Forging and stamping3321 | $42.0B | 1/4 |
| LyondellBasellNYSE:LYB | Resin, synthetic rubber, and artificial and synthetic fibres and filaments manufacturing3252 | $30.2B | 1/4 |
| International PaperNYSE:IP | Pulp, paper and paperboard mills3221 | $23.6B | 1/4 |
| Sherwin-WilliamsNYSE:SHW | Paint, coating and adhesive manufacturing3255 | $23.6B | 1/8 |
| AmcorNYSE:AMCR | Plastic product manufacturing3261 | $23.5B | 1/5 |
| PPGNYSE:PPG | Paint, coating and adhesive manufacturing3255 | $15.9B | 2/8 |
| WhirlpoolNYSE:WHR | Household appliance manufacturing3352 | $15.5B | 1/7 |
| AlcoaNYSE:AA | Alumina and aluminum production and processing3313 | $12.8B | 1/3 |
| Motorola SolutionsNYSE:MSI | Navigational and guidance instruments manufacturing334511 | $11.7B | 1/5 |
| CelaneseNYSE:CE | Resin, synthetic rubber, and artificial and synthetic fibres and filaments manufacturing3252 | $9.5B | 2/4 |
| SuzanoSUZBY | Pulp, paper and paperboard mills3221 | BRL 50.1B | 2/4 |
| Packaging Corporation of AmericaPrivate | Converted paper product manufacturing3222 | $9.0B | 2/4 |
| TapestryNYSE:TPR | Other leather and allied product manufacturing3169 | $8.0B | 1/4 |
| SynopsysNASDAQ:SNPS | Semiconductor and other electronic component manufacturing334410 | $8.0B | 1/5 |
And 607 more on the companies page.
Who works here
The occupations employed in Manufacturing, most concentrated in it first. The share is measured against the whole sector, not this industry — the published cross-tabulation stops there.
Tagged to this industry
Concentrated in this sectorA
These jobs are mostly done here. An operator in this industry is competing for them against others in the same industry, not against the whole economy.
And the jobs every business has
Found across at least fourteen of the twenty sectors. In a small establishment several of these are usually one person, or an outside provider.
Inside this industry
21 rows sit directly beneath 31-33, and 534 in all once every level is counted. Each has a base report of its own.