Application Modernization And Migration Services

The category name contains two words describing two different businesses, and a great deal of enterprise money has been spent on the confusion between them.

Migration moves an application. Same code, same architecture, different infrastructure. Modernization changes it. New architecture, new patterns, frequently a rewrite of things that worked fine and had simply stopped being maintainable.

One is a logistics exercise with a defined end. The other is open-ended engineering with a business case that has to be argued for. Both were sold under the same banner for most of a decade, and the shift between them is the story this category has been telling since at least 2021.

What the work actually involves

An enterprise arrives here with a portfolio, not an application. Several hundred systems of varying age, some critical, some nobody is certain about, a few that only one person understands.

The first deliverable is usually assessment. Which applications exist, what they depend on, what they cost to run, what business capability they support, and what the realistic disposition for each one is. That inventory is frequently the most valuable single artefact of the engagement, and organisations are consistently surprised by what turns up in it.

After that the work sorts into a handful of dispositions. Retire what nobody uses. Replace what a commercial product now does better. Rehost where the application is fine and only the infrastructure is wrong. Replatform where a few changes buy meaningful operational improvement. Refactor or re-architect where the application matters enough to justify the engineering. Reimplement where the thing needs rebuilding on modern foundations.

The provider brings the assessment tooling, the migration factory, the accelerators, the vertical-specific assets, and the several hundred people required to do this across a portfolio in a defined window. That last item is the actual product. Most enterprises could modernise any single application themselves. What they cannot do is modernise two hundred of them while continuing to run the business.

Why the distinction is commercial rather than semantic

Move-then-improve remains a viable approach, and Forrester says so. The reason it persists is that migration has a clean business case. Data centre exit, licence expiry, hardware refresh avoided. You can put it on a slide with a date.

Modernization has a messier one. The benefit is reduced technical debt and a path forward, which is real and difficult to price. Forrester's position is that this is where the actual business value sits, and its 2024 commentary is unusually direct in describing the market focus as shifting toward modernization at the expense of migration.

At the expense of is the phrase to notice. Not alongside. The lift-and-shift business is being cannibalised by the harder work, which is unusual in enterprise services, where the easier revenue normally survives.

There is a reason for it that shows up in the 2021 research. Enterprises that moved workloads unchanged discovered their cloud bills were higher than the data centre they left, because an application architected for fixed capacity behaves badly on metered infrastructure. Migration without modernization frequently transfers the problem and adds a monthly invoice.

The hyperscaler entanglement

One structural feature of this market deserves more scrutiny than buyers usually give it.

By 2021 Forrester was noting that the hyperscale cloud providers had become central to migration and modernization journeys, not only as strategic partners to large enterprise customers but by participating in dedicated go-to-market units that the large systems integrators were forming.

Those units are jointly staffed, jointly funded, and jointly incentivised. The cloud provider frequently subsidises assessment and migration work, sometimes heavily, because the return is consumption revenue over the following decade.

This is not hidden and it is not scandalous. It is, however, a conflict that sits inside the advice you are receiving. A provider running a funded practice with one hyperscaler has a structural reason to arrive at a target architecture on that hyperscaler. The assessment will be competent and the recommendation will be defensible, and it will also be the recommendation the funding arrangement makes easiest.

Ask directly which partner programmes fund the engagement and what the provider's revenue mix looks like across the major clouds. A provider genuinely balanced across three will give a different answer from one with ninety percent of its practice on a single platform.

Inside The Forrester Wave: Application Modernization And Migration Services, Q1 2024

The current evaluation is The Forrester Wave: Application Modernization And Migration Services, Q1 2024, authored by principal analyst Bill Martorelli, scoring fifteen providers against twenty five criteria across current offering, strategy, and market presence. It is addressed to cloud and vendor management professionals.

Six Leaders: Tata Consultancy Services, Accenture, Infosys, IBM, Capgemini, and Wipro, with five further vendors placed as Strong Performers.

Forrester identified three things separating the Leaders.

They combine delivery at scale with a genuinely differentiated vision, where the vision has to fit the customer's journey rather than exist as a marketing position. Scale is measurable. Vision is not, which is why it produces most of the separation.

They have demonstrated the ability to invest in emerging technology, generative AI in particular. Forrester attached an open question to this one, asking whether that attribute would reorder tomorrow's Leaders and answering that time would tell.

And they are conversant with modern delivery structures, specifically platform-based rather than factory-based approaches, with pricing models evolving as team-based structures adapt to product-oriented delivery.

Infosys placed in the top two on strategy and top three on current offering, taking the highest possible score on innovation, global delivery strategy, talent strategy, and twelve further criteria, and was positioned as a fit for most modernization opportunities, particularly application reimplementation. Wipro's profile described a provider extending from application management incumbency into modernization capability, which is a fair description of how several of these firms reached this market.

What changed at the top between 2021 and 2024

The predecessor was The Forrester Wave: Application Modernization And Migration Services, Q3 2021, scoring fourteen providers against twenty six criteria: Accenture, Atos, Capgemini, Cognizant, Deloitte, EPAM, EY, HCL Technologies, IBM, Infosys, Larsen & Toubro Infotech, NTT DATA, TCS, and Wipro. That edition was addressed to infrastructure and operations professionals.

Its Leaders were Accenture, Infosys, IBM, TCS, Deloitte, and HCL Technologies.

Compare the two lists. Accenture, Infosys, IBM, and TCS held Leader positions across both. Deloitte and HCL Technologies did not appear among the 2024 Leaders. Capgemini and Wipro moved up into them.

That is meaningful churn in a market where the same fifteen firms compete for everything, and it happened in under three years. It also happened during the period when the criteria shifted toward vision, emerging technology investment, and delivery model, which is exactly the kind of change that reshuffles firms of similar scale.

Worth noting alongside this: Hexaware was invited to participate in the 2021 evaluation, declined, and Forrester could not make sufficient estimates to include it. That is the participation policy operating in the open, and a reminder that absence from a Wave is not always a judgement about capability.

Factory versus platform

The delivery model distinction Forrester flagged in 2024 is the most practically useful thing in the research, and it survives translation into a procurement conversation.

The factory model is how this industry industrialised. Standardised process, defined roles, offshore delivery centres, throughput measured in applications migrated per quarter. It works, it is predictable, and it prices cleanly per unit of work.

The platform model treats modernization as product engineering. Persistent teams own outcomes rather than tickets, tooling and reusable assets carry more of the load, and the engagement is structured around capability delivered rather than tasks completed.

The reason this matters commercially is that the two models fail differently. A factory optimises for the applications that fit the pattern, and every portfolio contains a tail of systems that do not. Those are usually the important ones. A platform approach handles the tail better and costs more per unit on the applications that would have fitted the factory fine.

Most large engagements need both, and the negotiation worth having is about where the boundary sits and who decides which applications go through which route.

The question Forrester asked in 2024 has aged

Forrester's 2024 commentary described the market as on the verge of dramatic change from generative AI, promising to transform the nature of these services as AI gets embedded in accelerators and supercharges productivity. It also noted, carefully, that ambitions were large while established use cases were only firming up and case histories remained few.

That caution was correct at the time and it is now out of date, which is the main reason to treat this Wave as historical.

Legacy code comprehension and translation turned out to be among the more credible applications of the technology. Explaining what an undocumented COBOL module does, generating tests for code with no test coverage, and producing a first-pass translation into a modern language are exactly the tasks these engagements are full of, and they were previously the tasks that consumed the most expensive human hours.

Forrester has since established separate categories for agentic development platforms and for AppGen, which tells you the tooling side matured enough to warrant its own coverage.

The commercial implication is the one to carry into a negotiation. If a meaningful share of the effort in a modernization engagement is now performed by tooling rather than by people, a per-person rate card is measuring the wrong thing. Forrester saw the pricing question coming in 2024, noting that pricing models were evolving alongside team-based structures. Two years on, that evolution is a live negotiating position rather than a trend to watch.

Where this leaves you

Treat the Q1 2024 Wave as a reliable guide to who the serious providers are and an unreliable guide to their relative positions. The vendor set has not changed much. The capability that differentiates them has changed considerably, and there is no current scored evaluation to reflect it.

Four things to settle before signing.

Establish who owns the assessment output. The portfolio inventory a provider builds is genuinely valuable and genuinely portable. If it lives in the provider's proprietary tooling and leaves with them, you have rented knowledge about your own estate.

Fix the migration and modernization boundary in writing, with a mechanism for moving applications between categories. Every large programme discovers systems that were scoped for rehosting and need rebuilding. The question is whether that discovery triggers a conversation or a change order.

Ask what proportion of the work is now tool-assisted and how that shows up in the price. If the answer is that productivity gains accrue entirely to the provider, you are funding their margin improvement.

And ask about the tail. Providers will discuss the eighty percent of applications that follow the standard path. Your risk sits in the twenty percent that do not, and the provider's approach to those is where the programme succeeds or quietly runs two years long.

Forrester does not endorse any provider named here, and tier placement should not be read as a recommendation to buy.