Why embedded workflows, switching costs and operational dependence can create durable economics — and why investors still need to distinguish genuine indispensability from artificial lock-in
The economics of being hard to replace
Investors often describe attractive businesses as having “sticky” customers. The phrase is useful, but incomplete. A customer may stay because a product is good, because alternatives are weak, because a contract is long, or because switching would disrupt operations. Those are very different forms of retention, and they do not deserve the same valuation.
The strongest version of the investment case appears when a supplier has become woven into the customer’s operating system. Its software may sit inside a core workflow; its data may support regulatory reporting; its components may be qualified for a production line; or its service teams may be responsible for keeping mission-critical infrastructure running. Removing the supplier is technically possible, but operationally expensive, slow or risky.
That distinction matters because customer dependence can convert ordinary revenue into something more durable. Renewal rates can improve, revenue visibility can lengthen, and the economics of serving an established customer can become more attractive than the economics of winning a new one. For investors, the question is not simply whether customers renew. It is why they renew — and what would have to happen for them to leave.
Switching costs can be economic, technical and behavioural
Competition authorities increasingly treat switching costs as a central feature of market structure. The European Commission’s 2026 draft merger guidelines note that switching can be delayed by customisation, qualification procedures, technological barriers or legacy compatibility, while customer lock-in can arise when suppliers hold important data or operate interconnected ecosystems.
The OECD’s 2026 work on competition and consumer policy in digital markets makes a similar point: switching costs may come from data lock-in, loss of connections, relearning interfaces or reduced functionality. Even when monetary switching costs are low, time, effort and operational disruption can still discourage customer movement.
For investors, this creates a useful hierarchy. Contractual stickiness is the weakest form because terms eventually expire. Product habit is stronger but can be disrupted by a superior alternative. Integration-based stickiness is stronger still because replacement requires data migration, systems work and process redesign. At the top sits operational indispensability: the supplier is not merely used by the customer but is embedded in activities that the customer cannot easily pause.
Recurring revenue becomes more valuable when retention is structural
Recurring revenue is often awarded a premium because it makes future cash flows easier to forecast. But recurring billing alone does not create a moat. A subscription that can be cancelled in minutes has very different economics from a service whose removal would require a multi-month migration, retraining programme and redesign of connected systems.
Public-company disclosures illustrate the difference. SS&C Technologies’ 2025 annual report describes software-enabled services and maintenance contracts with average revenue retention above 95% in each of the prior five years. The company’s offerings are integrated into asset management, healthcare and other operational processes where complexity, scale and data continuity can make replacement demanding. The disclosure does not prove that switching costs alone cause retention, but it shows why investors examine retention together with embeddedness rather than in isolation.
A similar logic appears in cloud software. PagerDuty’s fiscal 2026 filing reported ARR churn below 10% of beginning ARR and described its role in customers’ digital operations. Again, the relevant investment question is not the headline retention metric by itself. It is whether the product is sufficiently integrated into operational processes that customers continue to expand usage over time rather than merely renew out of inertia.
The best businesses often become part of the customer’s process
This is why some of the most durable business models sell less like discretionary products and more like infrastructure. Enterprise software can become the system of record for a workflow. Industrial suppliers can become qualified vendors whose components have been tested against performance standards. Payments processors, custody platforms and compliance providers can become embedded in regulated operations. Maintenance firms can hold site-specific knowledge that competitors would need years to rebuild.
In each case, the supplier benefits from accumulated context. It knows the customer’s configuration, transaction history, equipment base, risk parameters or operating procedures. That knowledge improves service and raises the cost of replacement. Importantly, the advantage can compound without requiring the supplier to block competition. A competitor may be free to bid, but it still needs to reproduce years of integration, data and trust.
For investors, this can translate into several attractive characteristics: lower churn, greater pricing resilience, opportunities to cross-sell adjacent products, and a customer-acquisition model in which the lifetime value of a relationship rises as integration deepens. The result can be a business whose growth is partly protected by the difficulty of displacing what is already installed.
Data gravity can be as important as product quality
The rise of cloud and software ecosystems has made data portability a strategic issue. Once a provider hosts large volumes of customer data, applications and integrations, migration becomes more complex. The European Union’s Data Act was designed in part to reduce these barriers. The European Commission’s Data Act guidance says customers of cloud and other data-processing services have faced high egress charges, lengthy switching procedures and interoperability problems; the Act therefore introduces contractual and technical requirements intended to make switching easier.
Under the Data Act itself switching charges are scheduled to disappear from 12 January 2027, while providers must support portability and interoperability. The policy direction is significant for investors because it shows that some forms of customer dependence are not treated as benign. If retention relies primarily on technical obstruction or punitive exit costs, regulation can erode the advantage.
The stronger investment thesis therefore rests on value-based embeddedness rather than trapped customers. A business should ideally remain difficult to replace even after portability improves, because customers would still face meaningful operational change, retraining, process redesign or risk in moving to another provider.
Customer dependence can support pricing — but not without limits
A supplier that is deeply embedded may have more room to raise prices because the customer compares the increase not only with a competitor’s price, but with the total cost of switching. That cost can include migration, downtime, internal management attention, validation, retraining and the risk of disrupting a critical process.
Yet pricing power should not be confused with the ability to exploit customers indefinitely. Excessive price increases can motivate customers to multi-source, develop internal alternatives or support new entrants. The European Commission’s June 2026 preliminary DMA view on AWS and Microsoft Azure explicitly cited apparent lock-in effects and high switching costs as factors in its assessment of cloud market power. That is a reminder that very high customer dependence can become a regulatory liability.
Investors therefore need to look for a balance: enough embeddedness to support retention and economic durability, but not so much coercive dependence that customers, regulators or procurement teams actively seek to dismantle it.
What investors should examine beneath the retention number
A high retention rate is useful evidence, but it is not the end of the analysis. The most informative questions concern the mechanics of the relationship. How long does implementation take? How many systems connect to the product? How much customer data accumulates inside it? Does the product require certification, validation or retraining before replacement? Are customers expanding usage across departments, geographies or workflows? And can the supplier raise prices without materially increasing churn?
Investors should also separate customer value from customer captivity. Low churn can result from excellent service and genuine workflow integration, but it can also reflect poor portability, opaque contracts or exit penalties. The former may support a durable franchise. The latter may attract regulation, reputational damage or competitive disruption.
Customer concentration is another constraint. A business can be hard to replace for each individual customer and still be risky if a small number of customers dominate revenue. The ideal profile combines high retention with a diversified customer base, recurring economics, measurable expansion within existing accounts and limited dependence on any single buyer.
The counterargument: technology can destroy switching costs faster than investors expect
The investment case is not static. New technology can reduce migration costs, open standards can improve interoperability, and regulation can force portability. Artificial intelligence may also make it easier to translate data models, rewrite integrations or retrain users, reducing barriers that once looked permanent.
This is especially important in software. A product can appear irreplaceable because migration is painful, only for a new abstraction layer, open protocol or automated conversion tool to make switching much easier. Conversely, businesses that remain deeply embedded because they continuously improve the customer’s workflow are more resilient than those whose moat depends on technical inconvenience.
The distinction is similar in industrial markets. Qualification and certification can protect an incumbent supplier, but a major technology transition may reset the qualification process and allow new competitors into the system. A replacement cycle can strengthen incumbency, but a platform change can erase it.
Why “difficult to replace” can justify a valuation premium
A business that customers cannot easily replace can deserve a higher valuation because the durability of its cash flows may be greater than headline growth alone suggests. Lower churn increases customer lifetime value. Embedded relationships can support expansion revenue. Recurring contracts can improve visibility. And operational importance can reduce the sensitivity of demand to short-term economic cycles.
But the premium should be earned through evidence. Investors need to see retention, renewal behaviour, customer expansion, implementation depth, pricing resilience and low concentration risk. They also need to understand whether regulation or interoperability standards are likely to reduce switching barriers.
The most attractive businesses are not those that make leaving impossible. They are those that make staying economically rational. Their customers remain because replacing the supplier would require giving up accumulated integration, knowledge, reliability and operating confidence — not because an exit door has been deliberately blocked.
Conclusion
The investment case for difficult-to-replace businesses is ultimately a case for durable relevance. The deeper a supplier becomes embedded in a customer’s workflows, data and infrastructure, the more resilient the relationship can become. That resilience may support retention, recurring revenue and stronger long-term economics.
Yet customer dependence is not automatically a moat. Regulation is increasingly targeting artificial lock-in, cloud portability is improving, and technology can compress switching costs. Investors therefore need to distinguish structural indispensability from friction that can be legislated, automated or competed away. The strongest franchises are hard to replace because they keep creating value inside the customer’s operating system — not simply because switching is difficult.
References
1. European Commission — Draft revised merger guidelines (2026)
2. OECD — Competition and consumer policy in digital markets (2026)
3. European Commission — Data Act explained: switching between data processing services
4. EUR-Lex — Regulation (EU) 2023/2854, Data Act
5. European Commission — Preliminary DMA position on AWS and Microsoft Azure, 25 June 2026
6. European Commission — Cloud computing stakeholder roundtable under the DMA, 1 July 2026