Constellation Software is the benchmark every serial acquirer pitch invokes. Investors have spent a decade searching for the next CSU — the company that buys small vertical market software businesses, never sells, and compounds capital indefinitely. The search has produced a long list of imitators and a short list of results.

Part of our Global Serial Acquirer Scorecard.

We screened 187 serial acquirers across 14 markets. The data answers the constellation software competitors question directly: which companies actually outperform CSU’s current returns, and do any of them use the CSU model?

Key Finding: Constellation Software earns 10.8% ROIC — below our 12% cost of capital threshold. Six companies from our global screen beat this figure. None of them replicates the CSU model. They use different target types, different geographies, and different operating approaches. The CSU model is specific, scale-constrained, and harder to replicate than most analyses acknowledge.

Constellation Software Competitors: What CSU Actually Earns Today

The starting point for any comparison is CSU’s current numbers, not its historical ones.

CSU’s goodwill-to-total-assets ratio is 11% — the lowest of any major acquirer in our screen. The company is not overpaying per deal. The problem is arithmetic: deploying $3B+ annually into VMS acquisitions, even at disciplined multiples, generates only 10-11% returns when the total invested capital base reaches $18B. CSU’s ROIC has declined from the high teens as revenue scaled from $4B to $14B.

CSU’s (GW+Int)/TA of 47% tells the fuller story. IFRS purchase price allocation classifies customer relationships and acquired technology as identifiable intangibles, not goodwill. For software acquirers, the combined ratio is the honest metric. At 47%, CSU carries substantial acquisition-premium exposure even though the headline GW/TA of 11% looks modest.

Topicus.com — CSU spun off Topicus in 2021 — earns 8.5% ROIC at 16% GW/TA and 33% combined ratio. Six data points, one anomalous year. Not yet a track record. Lumine Group — CSU spun off Lumine Group in 2023 — has no public ROIC data available in our screen.

The Six Companies That Beat CSU’s Current ROIC

We identified six companies from our 187-company screen with ROIC above 10.8% that investors and analysts group with or compare to CSU. Each uses a different model.

None of these companies is the “next CSU.” They each represent a distinct model, and the differences in model explain the differences in returns.

Four Models, Four Different Answers

Model 1: The Regulatory Moat (Wolters Kluwer)

Wolters Kluwer earns 19.7% ROIC at 50% GW/TA with 24.6% operating margins. It earns 19.7% ROIC at 50% GW/TA — one of the highest-ROIC acquirers in Europe carrying significant goodwill. The model: acquire information services businesses serving professionals in regulated industries (legal, tax, healthcare, finance), integrate them into workflow tools professionals cannot easily abandon, and reprice annually against switching costs they cannot afford.

The comparison to CSU is superficial. CSU targets small VMS businesses with low acquisition competition. Wolters Kluwer targets information services assets at premium multiples — its 50% GW/TA reflects years of paying full price. What sustains the returns is not acquisition discipline; it is regulatory lock-in. A law firm cannot switch its legal research platform mid-matter without disrupting workflow. A tax accountant cannot abandon the tool integrated with client data. The moat is regulatory complexity, not software switching costs.

At 24.6% operating margins, Wolters Kluwer absorbs a 50% goodwill load and still clears cost of capital by 8 percentage points. This is the margin path to high ROIC described by Mauboussin & Callahan (2023): sustained top-quintile performers average NOPAT margins 2.7x the universe. Wolters Kluwer’s margins are roughly 2.5x the European average.

Model 2: The Industrial Niche Compounder (AddTech)

AddTech earns 14.9% ROIC at 30% GW/TA with 12.6% operating margins. It is a Swedish industrial technology distributor with Bergman & Beving heritage — the founding family refined this model over decades. Lagercrantz, its Bergman sibling, earns 14.3% ROIC at 32% GW/TA with 15.0% operating margins.

The AddTech model acquires niche industrial component distributors across Scandinavia and Northern Europe. Targets are small, often owner-managed, operating in product categories narrow enough to avoid large-company competition. Price discipline holds: the Bergman companies historically buy at 6-8x EBIT, well below the 12-15x that software acquirers pay.

The comparison to CSU fails on target type. CSU buys vertical market software companies — subscription revenue, high switching costs, scalable. AddTech buys industrial distributors — project revenue, moderate switching costs, limited scalability. Both generate adequate returns, but through different mechanisms. AddTech’s margins (12.6%) are thin by software standards; the returns come from buying disciplined and keeping the goodwill base moderate.

AddTech is also not decentralized in the CSU sense. CSU operates a portfolio of businesses each managed autonomously; AddTech integrates acquired companies into a shared commercial infrastructure.

Model 3: Architecture Software (Nemetschek)

Nemetschek earns 14.5% ROIC at 53% GW/TA with 24.1% operating margins. The German building information modeling software company has one of the largest combined goodwill loads of any Tier A company in our screen — its (GW+Int)/TA reaches 71%. The ROIC survives only because of elite software margins.

This is the closest structural parallel to CSU among companies that beat CSU’s current returns: Nemetschek buys vertical software with sticky user bases, integrates them loosely under a holding company, and compounds through recurring revenue. But the target market is specific — architecture, engineering, and construction software — and the goodwill load now draws scrutiny. Nemetschek’s goodwill doubled in a single year (from EUR 552M to EUR 1,135M) following accelerated acquisitions. At 53% GW/TA and rising, any margin compression would push ROIC below 12%.

The risk profile is higher than CSU’s. CSU’s 11% GW/TA means the goodwill buffer is enormous — even if ROIC declines, it takes catastrophic overpaying to threaten capital recovery. Nemetschek at 53% has far less margin for error.

Model 4: The CSU-Adjacent Model (Enghouse Systems)

Enghouse Systems is the most direct CSU comparator in our screen. It earns 12.1% ROIC at 40% GW/TA with 18.0% operating margins — Tier A, barely, declining from 21%.

Enghouse buys small vertical market software companies, often the same categories CSU targets: contact center software, transit management, healthcare communications. It operates a decentralized holding model. The acquisition multiples are CSU-adjacent. The target geography overlaps.

And yet Enghouse earns 1.3 percentage points above cost of capital while CSU earns 1.2 points below it. The difference is not model fidelity — Enghouse copies the CSU model more faithfully than any other company in our screen. The difference is scale. CSU deploys $3B+ annually; Enghouse deploys far less. At smaller deployment volumes, the model still produces Tier A returns. At CSU’s current scale, it no longer does.

This is the core finding for the “next CSU” thesis: the model works at small scale. The constraint is not finding the model. The constraint is finding enough targets to feed it.

The US Comparison: Margins Over Model

The US serial acquirers frequently compared to CSU use fundamentally different approaches.

Roper Technologies earns 28.4% operating margins with 6.2% ROIC. The most extreme margin-ROIC disconnect in our entire screen across 187 companies. Roper pays 15-20x EBIT for niche software platforms — transformative deals, not tuck-ins. Its (GW+Int)/TA reaches 91%. The goodwill absorbs the margins. For Roper to reach 12% ROIC, it would need to cut its goodwill base roughly in half or nearly double its operating profit.

Danaher invented the operating system approach to serial acquisition — the Danaher Business System applies lean manufacturing principles to every acquired business. Post-spinoffs (Veralto in 2023, Fortive in 2016), Danaher retains its life sciences assets and earns 5.6% ROIC at 52% GW/TA. The operating system works on water treatment and industrial equipment; it has not overcome the capital intensity of pharma R&D.

TransDigm earns 10.7% ROIC at 46% GW/TA with 47.4% operating margins — the highest margins of any company in our global screen. The model: acquire aerospace component suppliers in sole-source positions, then raise prices aggressively because the FAA makes switching suppliers years-long. It is an acquisition model only in form. The returns come from pricing power, not acquisition discipline. TransDigm is not a constellation software competitor — it is an aerospace monopoly manager that happens to grow through acquisitions.

The “Next CSU” Trap

The most common error in serial acquirer analysis is comparing a company to CSU’s historical returns — the 25-35% ROIC from 2008-2020 — rather than its current trajectory.

From our Canadian market analysis: Canada has the lowest Tier A rate (15%) of any market we screened. The country that built the model produces the worst results. Seven of 13 Canadian serial acquirers earn below 5% ROIC. The one that epitomizes the model earns 10.8%.

The “next CSU” pitches cluster around two failure modes.

Failure mode 1: Wrong target type. CSU buys small VMS businesses at 0.7-1.5x revenue — businesses boring enough that no private equity firm competes for them. Companies pitched as “CSU for industrials” or “CSU for healthcare services” are buying different businesses at different multiples in competitive markets. Vitec Software in Sweden — often called “the Swedish Constellation Software” — pays 3-4x revenue for VMS targets. That 2-3x premium gap is the entire difference between 25% ROIC and 6.3% ROIC. Vitec earns 20.9% operating margins and 6.3% ROIC. The business performs. The acquisition pricing does not.

Failure mode 2: Wrong scale assumption. Every company pitched as the next CSU is smaller than CSU was when its returns were exceptional. CSU earned 27% ROIC at $1B revenue and 17% ROIC at $4B revenue. The trajectory is the product of scale, not discipline. A $500M serial acquirer buying small VMS companies at 4-6x EBIT can earn 15-20% ROIC. The same company deploying $3B annually into the same target pool — if such a pool existed — would earn less. Enghouse at $800M-$1B annual revenue earns 12.1%. CSU at $14B earns 10.8%. This is not a coincidence.

Mauboussin & Callahan (2023) find that 48% of top-quintile ROIC companies remain there after three years, and 15% fall to the bottom quintile. CSU has not fallen to the bottom — its current 10.8% ROIC still exceeds the Mauboussin median of 9.5% for the Russell 3000 — but the trajectory from 27% to 10.8% over fifteen years is precisely the regression to the mean their research predicts for companies sustaining top-quintile returns at scale.

A Direct Comparison: Nine Metrics Across Six Models

Three patterns emerge.

High margins compensate for high goodwill. Wolters Kluwer (24.6% margins, 19.7% ROIC) and Nemetschek (24.1% margins, 14.5% ROIC) carry the highest combined goodwill loads of any Tier A companies in our European screens — and clear cost of capital on the strength of their operating economics. Roper Technologies (28.4% margins, 6.2% ROIC) proves the limit: even 28% margins cannot overcome 62% GW/TA when deals are priced at 15-20x EBIT.

Low goodwill protects capital but does not maximize returns. CSU’s 11% GW/TA is the lowest of any acquisitive company in our screen. The discipline that protected CSU’s capital base during its scaling phase now constrains its ROIC — at $18B in invested capital, low goodwill as a percentage still means enormous absolute goodwill. The model’s arithmetic ceiling is lower than its history suggests.

The decentralized model is not unique to CSU. CSU, Enghouse, Nemetschek, and Diploma all operate decentralized holding structures. Decentralization is a common feature of successful serial acquirers, not a CSU differentiator. The differentiator was price: CSU entered a market with no competition for small VMS businesses in the 2000s. That market now has competition — from CSU’s own spinoffs, from private equity, and from imitators who have read the same playbook.

What the Data Tells Investors

CSU at 10.8% ROIC trades at a significant premium to book value. The premium assumes its returns will recover as it continues to scale. Our data provides reference points for that view.

The companies that beat CSU’s current ROIC do so through specific structural advantages: regulatory lock-in (Wolters Kluwer), long-term government contracts (CGI), or niche operating margins (Nemetschek, Diploma). CSU’s structural advantage — the VMS target pool — is intact but increasingly competitive.

The Swedish market shows what happens when acquisition competition enters a disciplined niche. When Swedish serial acquirer multiples inflated from 6-8x to 10-14x EBIT, ROIC across the market collapsed. Only companies with organic growth moats (AddTech, Lagercrantz) or exceptional margins (Lifco at 18.7%) survived with returns above 12%. CSU is not in a market with 24 direct competitors, but VMS multiples have risen as CSU’s own scale created a market signal.

The US market shows that sustained returns require either stopping acquisitions (Illinois Tool Works, 30.8% ROIC after stopping in 2012) or extraordinary operating margins. At 16.3% margins, CSU is not in the extraordinary margin tier. It is in the range where ROIC is sensitive to deal multiples and capital deployment volume.

The relevant question for investors is not “Is there a next Constellation Software?” It is “Can CSU’s current returns recover?” Our data cannot answer this. It can confirm that no other company in our screen sustained the returns CSU generated in 2012-2017 as it scaled past $10B in revenue, because no other company has attempted that scale in the same model. The arithmetic constraint is real. Whether CSU can overcome it through the spinoff model (Topicus, Lumine) is a different question from whether its imitators can.

Methodology

Data covers the latest fiscal year available as of February 2026. ROIC uses net operating profit after tax divided by invested capital with goodwill in the denominator. GW/TA is goodwill divided by total assets. (GW+Int)/TA adds identifiable intangible assets from IFRS purchase price allocation. Source: QuickFS (all markets), Yahoo Finance (drawdowns). Tier classification: A (>12%), B (10-12%), C (5-10%), D (<5%).

We screened 187 serial acquirers across 14 markets. This article draws on six market benchmarks: Canada, Sweden, the United States, the United Kingdom, Germany, and the Netherlands. Full global rankings are in the Global Serial Acquirer Scorecard.

Related Analysis

• The Goodwill Cliff — why 76% of companies above 40% GW/TA destroy value
• Serial Acquirer ROIC Analysis — the full ROIC decomposition across 187 companies
• Best Serial Acquirers 2026 — the 59 Tier A names ranked by quality