Computer systems design and related services

The most fragmented generic category — the top five vendors hold only 20–25% — and still a walk, because the fragmentation is a symptom of low switching costs rather than an unserved segment.

Every suite vendor already ships this, and the independents are being squeezed from both ends. One analyst count puts the top seven — IBM, SAP, Oracle, DocuSign, Coupa, Icertis and Zycus — at over 30% of the market between them [C], which is a category where the buyer's existing vendor is always a credible option. Icertis claims more than a third of the Fortune 100 [C, vendor]. Meanwhile the AI drafting layer is arriving as a feature of the word processor rather than as a system of record. A new entrant would be selling a repository to buyers who already have three. The quoting and proposal end of CLM is a separate SMB and mid-market tier, and PandaDoc leads it. It sells proposals, quotes, contracts and e-signature from a public price list of $19–$65 per user a month, with quote builder and product catalogue on the Business plan and CPQ as a paid Enterprise add-on [C, vendor, checked 2026-10-08]. Its buyer is a sales team, not a legal department, and its lock-in is templates and a CRM connector rather than the executed-contract repository. That makes it cheaper to adopt and easier to leave than enterprise CLM. The same tier is crowded too: Proposify, Qwilr, Juro and DocuSign's own CLM Essentials all compete there, so it does not change the verdict.

A free incumbent runs most of the web, and every tier around it already has funded vendors. W3Techs' October 2026 survey puts WordPress on 40.1% of all websites and 58.6% of the sites whose CMS it can detect [B, measured survey]. The next largest are Shopify (7.8%), Wix (6.1%) and Squarespace (3.6%). WordPress is GPL software, so nobody can undercut it on price: it costs nothing. What holds buyers is the theme, plugin and agency ecosystem, not a contract. Above it, Adobe and private-equity firms own the enterprise digital-experience suites. Adobe's Digital Experience segment, which carries Experience Manager, reported $5.86B in fiscal 2025, up 9% [A, 10-K]. Sitecore (EQT), Optimizely (Insight Partners) and Acquia (Vista) each changed hands at about $1B or more. Beside it, the headless vendors are already funded to scale. Contentful has raised about $335M at a $3B-plus valuation, Contentstack $169M, Storyblok $138M and Hygraph $43M, and Sanity raised an $85M Series C in 2025. Strapi, Payload (now owned by Figma) and Directus offer the same architecture as free or source-available software. Below it, the hosted website builders are large. Wix reported $1.99B of revenue for 2025 [A], Squarespace went private to Permira at $7.2B, and Webflow last raised at $4B. That leaves no unserved tier for a new entrant: free at the bottom, bundled at the top and venture-funded in between. Incumbent vulnerability decides it. How this differs from the neighbouring records: DAM (541514-dam-digital-asset-management) covers the asset library of images, video and brand files that a CMS draws on. Knowledge Management (541514-knowledge-management-and-enterprise-search) covers internal wikis and search across what employees know. This record covers the system that writes and publishes an organisation's public websites and digital content.

The category just set the price of entry, and it is $32 billion. Google completed its purchase of Wiz on 11 March 2026 — the largest cybersecurity acquisition on record [B]. That settles who owns the independent leader, and it points at the structural problem underneath: the three hyperscalers each ship a native posture-management service with the cloud, so the buyer's default is already bundled. Palo Alto, CrowdStrike and Microsoft fold CNAPP into platform agreements a new entrant cannot price against. Unusually for this research, a strong open-source tier exists — Prowler, Cloud Custodian, ScoutSuite, Steampipe — because the artefact is a configuration scan rather than a network or a ledger, and that sets the floor at zero from below while the platforms squeeze from above.

Squeezed between a free tier that runs universities and an HR suite that gives it away. Moodle and Open edX are institution-grade and cost nothing, which puts the floor at zero; above that, SAP SuccessFactors Learning, Workday Learning and Oracle ship learning inside the HR system that already holds the employee record. The listed independent, Docebo, is reported at around $140M of ARR [C, vendor] — a real business, and also the ceiling for what independence buys in this category. Market figures here are unusually unreliable: estimates of the 2025 LMS market run from $30.9B to a projected $100.7B by 2032 [C], a spread wide enough to be useless for planning.

CPQ is a feature of the system that already owns the deal. Salesforce, Oracle and SAP each ship it inside the CRM or ERP the quote is being written in, so the standalone market is what is left after the suites take their share — and that residual is small enough that the one listed pure-play, PROS, guided to about 9% total revenue growth for 2025 [B] while selling CPQ alongside pricing and revenue management rather than on its own. The independents are analyst-recognised and commercially modest. A new entrant would be building a configuration engine to sit beside a CRM whose vendor gives one away with the renewal.

Salesforce earns more CRM revenue than Microsoft, Oracle, Adobe and SAP combined, and the credible free alternative has 45,000 GitHub stars. A new entrant is squeezed between an incumbent it cannot outspend and a substitute it cannot underprice.

A feature of the marketing suite that a handful of independents sell separately. Adobe ships Experience Manager Assets inside the stack that already holds the creative files, the CMS and the campaign, and the best-of-breed case for Bynder or Canto is integration quality rather than capability the suite lacks. The standalone category is what remains for buyers who do not want Adobe — a real but bounded market, with no vendor disclosing a DAM line anywhere. Open-source options are credible here (ResourceSpace, Pimcore), because an asset repository is a file store with metadata, which a community can build. Between a bundled suite above and a workable free tier below, the independent middle is thin.

Most buyers already own a DMS, because SharePoint comes with Microsoft 365. Microsoft reported 200 million monthly active SharePoint users in the cloud in December 2020 [C, vendor]. The storage, versioning and permissions that define a document management system are now part of the productivity licence. Above that default sit three large repositories. OpenText (Waterloo, Ontario) sells Documentum and Content Server inside a $5.25B business that grew 1.5% in FY2026 [A]. Hyland, owned by Thoma Bravo, bought both of the big open-source platforms, Alfresco (2020) and Nuxeo (2021) [B]. Box made $1.18B in FY2026, up 8% [A]. Below them, Paperless-ngx (GPL-3.0, about 46,000 GitHub stars) and Mayan EDMS make a self-hosted repository free. The one niche with its own leaders is legal, and it is already split between two firms. iManage claims 81% of the AmLaw 200 [C, vendor] and NetDocuments claims more than 7,000 customers [C, vendor]. Both are backed by private equity (Bain Capital Tech Opportunities; Warburg Pincus and Cove Hill). A new entrant would be selling a repository to buyers who already pay for one, and switching means migrating the records and their retention history. Incumbent vulnerability decides it. Scope. This is enterprise document management and ECM: storing, versioning, permissions, retention and search of documents. In automotive retail, "DMS" means Dealer Management System (CDK, Reynolds), which is a different product with a different buyer, covered by the 441110 Dealership Management Systems record. This record also differs from its neighbours. DAM manages rich-media brand assets. Knowledge management and enterprise search is the wiki and AI-answer layer over many repositories. CLM manages the contract record and its clauses. DMS is the general system of record for documents those tools may sit on.

Every independent EAM vendor that reached scale has already been bought, and the buyers are the industrial and suite companies the customers already use. IBM bought Maximo's maker, MRO Software, for about $740M in 2006 [B]. GE bought Meridium, the asset-performance (APM) leader, at a $495M enterprise value in 2016 [A, acquirer release]. Trimble bought Cityworks in 2019 [A, acquirer release]. Hexagon paid $2.82B for Infor's EAM business in 2021, a business expected to earn $184M that year, more than 70% of it recurring [B, release as reproduced by CIMdata]. IFS bought Ultimo in 2022 and Copperleaf in 2024. Siemens paid $1.575B plus an earn-out for Brightly, which expected about $180M of 2022 revenue [B, acquirer release]. Schneider Electric took full ownership of AVEVA in January 2023 [B]. The exit prices are high, roughly 9–15 times revenue, but they were paid for decades of reference customers in regulated, asset-heavy buyers, not for software a new entrant could rebuild. The buyer is a utility, a transit agency, a refinery or a public-works department. It buys by RFP on a five-to-fifteen-year cycle, wants a reference list in its own sector, and treats the system as the legal record of its asset condition and maintenance. How this differs from its neighbours. CMMS, the record written alongside this one, is the work-order tool for a single plant or facility team: priced per technician, bought by a maintenance manager, swappable in a quarter. EAM is the enterprise asset register: the hierarchy of every asset, its lifecycle cost, its condition and risk, capital planning (Copperleaf) and reliability analytics (APM). It is integrated with ERP finance and GIS and bought by the CIO or the asset-management director. The two overlap in the mid-market (IFS Ultimo, Brightly, eMaint all sell both). ERP (541514-erp-enterprise-resource-planning) holds the money; SAP and Oracle sell EAM as an ERP module, which is half the incumbent problem. Utility metering, billing and CIS (2211-utility-metering-billing-and-customer-systems) is the customer-and-meter side of the same utility buyer; EAM is the network-and-plant side. Incumbent vulnerability decides it: there is no weak incumbent to attack, only acquirers waiting for the next independent.

One vendor is pulling away and buying the rest. Deel passed $1 billion in annual recurring revenue in 2025, booked its first $100 million revenue month in September 2025 and its third straight profitable year, then raised $300M at a $17.3B valuation from Ribbit Capital, Andreessen Horowitz and Coatue [B]. It has spent the money on consolidation: it bought Safeguard Global's payroll division (March 2025) and Omnipresent outright (October 2025), after PayGroup, PaySpace and Zavvy [B]. G-P (Globalization Partners) claimed about $1 billion of ARR as long ago as January 2022, when Vista Credit Partners put in $200M at a $4.2B valuation [B; ARR is the company's own claim, C]. Rippling has bundled EOR into its HRIS and payroll suite, and the rivalry between Rippling and Deel is now in court: Rippling sued Deel for corporate espionage in March 2025 and Deel counter-sued, alleging that Rippling copied its EOR product [B]. A new entrant would be selling a commodity legal wrapper against a profitable leader that can undercut it and outspend it on acquisitions. What the category is. An Employer of Record (EOR) is the legal employer in the worker's country. It holds a local entity, signs the employment contract, runs payroll, withholds tax and pays statutory benefits, while the client directs the work. The service is a PEO-like arrangement and belongs to NAICS 561330 (professional employer organizations); this record covers the software platform sold to the hiring company, so it sits at 541514 with the other horizontal records. An Agent (or Agency) of Record (AOR) is the contractor version. The platform contracts with an independent contractor abroad, vets the classification, collects invoices and pays them, and takes on part of the misclassification risk. Atlas HXM, for one, now markets AOR as a separate product [C, vendor site]. Most platforms also sell global payroll for companies that already have entities, and visa sponsorship and relocation: Deel, Oyster and Atlas list visas on their own sites, and Envoy Global sells corporate immigration on its own [C, vendor sites]. Why it is horizontal. The buyer is any company that hires one person in a country where it has no entity. That might be a five-person startup with a developer in Portugal or an enterprise testing a new market, so demand is set by cross-border hiring, not by any one industry. How it differs from the neighbours. 541514-hrm-human-resource-management covers the domestic HRIS and payroll system of record (ADP, Workday, Rippling's core). This record covers the cross-border employment layer, where the platform itself becomes the employer. 5613-employment-services and 561320-temporary-staffing-agency are agencies that recruit and place workers, often on their own payroll. An EOR does not find the worker; the client brings a person it has already chosen. 541514-ats-applicant-tracking-systems is the step before the hire. The capital tells the story. Remote ($495M in total, SoftBank Vision Fund 2 leading its 2022 Series C) [B], Velocity Global, now Pebl ($500M, including a $400M Series B in 2022) [B], Papaya Global ($440M; $3.7B valuation in 2021) [B], Oyster ($227M) [B], Atlas (up to $200M from Sixth Street Growth) [B] and Multiplier ($77.2M) [B] all raised at 2021-22 prices. Omnipresent ($120M Series B in 2022) has since been absorbed by Deel. The only Canadian entrant found, Toronto's Borderless AI ($27M seed, March 2024) [B], had a 'coming soon' placeholder in place of its website on 2026-10-10. Incumbent vulnerability decides it: the leader is profitable, still growing and buying its rivals.

The open-source alternative in this category is not a hobby project. General Atlantic valued Odoo at €7B in January 2026. Competing here means facing SAP at 22% share and an $8B open-source company at the same time.

Five vendors hold roughly three-quarters of RMM and PSA and the top two are within a point and a half of each other, so the category is contested but closed to newcomers — NinjaOne needed $500M of ARR to reach third. Note a sizing contradiction: a published figure puts RMM/PSA vendor revenue above $20B, while the same source's shares imply roughly $2.8B once you divide NinjaOne's and Kaseya's disclosed revenue by their stated shares. The larger number is measuring MSP-delivered services, not vendor licences, and is recorded here only as a caution.