AWS runs a stack of programs that fund a move onto its cloud: MAP credits scaled to your workload, free Assess and paid Mobilize phases, partner-side incentives, and OLA for Windows and SQL Server licensing. The money is real and large — but it is partner-filed through the AWS Partner Network, which is exactly why DIY migrators almost never see a dollar of it. This page maps every program, who qualifies, the realistic amounts by migration size, and the first step.
AWS will spend real money to move a workload onto its cloud because the lifetime value of a committed customer dwarfs the one-time cost of the migration. Understanding that incentive is the key to understanding why the money exists, how much there is, and why you have to be pointed at it.
A workload that lands on AWS and stays there generates years of consumption revenue — compute, storage, database, data transfer, managed services. Against that, the cost of funding the cutover (a few weeks of partner engineering plus a slug of usage credits) is a rounding error. So AWS built a set of programs that pay for the migration directly, on the theory that the credits run out right about the time you are too embedded to leave. That is not cynicism — it is the structural reason the funding is generous and durable.
The programs are real, public in name, and almost entirely opaque in mechanics. AWS publishes that the Migration Acceleration Program exists; it does not publish a "claim your migration funding" button, because the money is intentionally routed through partners. AWS would rather a vetted partner own the relationship, do the assessment, run the cutover, and carry the support load than hand cash to a customer migrating solo and hope it goes well.
The result is a structural gap. The teams most likely to migrate themselves — a strong engineering org moving off Heroku, a platform team lifting VMware into AWS — are exactly the teams with no path to file for the funding, because filing happens inside the AWS Partner Network. They do the migration competently, pay full price for the engineering hours and the AWS usage during cutover, and never learn that a partner-filed MAP record could have credited a large share of it back.
This page exists to close that gap. The next sections walk through each funding program, who qualifies, the realistic dollar amounts by migration size, and the partner-filed mechanic that turns "AWS theoretically funds migrations" into credits sitting in your billing console.
MAP is the program people mean when they say "AWS funds migrations." It is structured in three phases — Assess, Mobilize, Migrate & Modernize — and the funding scales up as you move through them. Knowing what each phase delivers (and what AWS pays for in each) is what lets you forecast the real number.
MAP is partner-led by design. A partner with the relevant AWS competency files a MAP record on your behalf and is credited by AWS for the engagement; you receive AWS usage credits scaled to the migration. The three phases are gates: you do not jump to production credits without first establishing the business case and proving a pilot.
What it is: a structured discovery of your current environment — application inventory, dependencies, a directional TCO comparison, and a migration readiness score. The partner uses AWS Application Discovery Service and Migration Hub to inventory servers and map dependencies, then builds the business case.
What AWS funds: the Assess phase is typically delivered at no cost to you — AWS funds the partner to produce the assessment, because a credible business case is what unlocks the later phases. This is the phase DIY migrators skip entirely and, with it, the document that justifies all the downstream credit.
Duration: 2–4 weeks for a mid-size estate; longer for sprawling on-prem environments with hundreds of servers.
What it is: build the foundation. Stand up the landing zone (multi-account structure, networking, guardrails, security baseline), run a pilot migration of a representative workload, and validate the runbook and rollback plan before touching production. See the AWS landing-zone guide for what a production-grade foundation actually contains.
What AWS funds: AWS credits a portion of the Mobilize work — directionally around 25% of qualifying migration costs at this stage — to de-risk the foundation phase. The exact percentage is set per engagement; treat it as a representative range, not a contract.
Duration: 4–8 weeks, depending on landing-zone complexity and how many workloads the pilot covers.
What it is: production cutover at scale, plus selective modernization (re-platforming a database to a managed service, containerizing onto ECS/EKS, replacing self-managed components with managed ones). This is where the bulk of the work — and the bulk of the funding — lives.
What AWS funds: the Migrate phase carries the largest credit, directionally up to ~50% of qualifying migration costs, scaled to your projected post-migration AWS consumption. Larger committed spend pulls a larger credit; this is the lever that turns a substantial migration into a low-cost or near-$0 one.
Duration: 8–24 weeks for production cutover, longer for estates with strict downtime windows or heavy data-sync requirements.
MAP funding scales with your post-migration AWS commitment. A migration projecting meaningful steady-state AWS spend (roughly $5K/month and up) is what unlocks the larger credits; very small workloads see Assess/Mobilize support but little Migrate-phase credit. If your migration is below that line, the value CloudRoute delivers is a vetted partner who de-risks the cutover — not a six-figure credit. We tell you which case you are in before you commit a minute.
MAP is the centre of gravity, but it is not the whole story. Three other layers stack on top, and the difference between a partner who knows them and a team migrating solo is often tens of thousands of dollars — or a six-figure licensing saving.
A good partner sequences these so they reinforce each other rather than collide. Each has its own qualification logic and its own filing path inside the Partner Network.
If you run Windows Server or SQL Server, OLA is where some of the largest savings hide. An OLA is an AWS-funded assessment that right-sizes your fleet and models the licensing options — bring-your-own-license vs license-included, and whether moving SQL Server workloads to a managed or open-source engine eliminates licensing entirely.
The headline move OLA surfaces is migrating SQL Server to Aurora PostgreSQL (the same SCT + DMS path used for Oracle) or to RDS, converting recurring licence cost into either zero (open-source engine) or pay-as-you-go. For estates with significant SQL Server licensing, the license saving alone can exceed the entire migration cost — and an OLA is what quantifies it credibly enough for AWS to fund the move.
AWS funds partners directly to do migration work through programs like Partner Funding (POA/MDF-style mechanisms) and migration-specific incentives tied to MAP. You never see these line items, but they are why a partner can run an Assess phase at no charge to you and price the Mobilize/Migrate work below what the same engineering would cost as a naked consulting engagement.
This is the mechanic that makes "$0-cost migration" honest rather than a sales line: the partner is paid by AWS for a qualifying engagement, the customer is funded by AWS credits, and the economics close without you in the payment loop. The same structure underwrites AWS POC funding for proof-of-concept work.
If you are a funded startup migrating to AWS, MAP and Activate Portfolio credits ($100K) are not mutually exclusive — they fund different things. Activate funds your general-purpose AWS consumption (the steady-state bill); MAP funds the migration project itself. A startup migrating off Heroku can have a partner file MAP for the cutover and Portfolio for the ongoing infrastructure, and the two stack cleanly because they describe different spend.
What does not work is double-counting: filing the same workload under both programs to inflate the total. AWS reviewers see both records. A partner who knows the rules scopes them so MAP covers the migration and Activate covers the run-rate — which maximizes the legitimate total without tripping the double-count filter.
The single most useful thing this page can give you is a grounded sense of the dollar amount for a migration like yours. The driver is not your headcount or your funding stage — it is the projected post-migration AWS spend and the size of the workload being moved.
The figures below are representative ranges from how migration funding typically lands in 2026, not quotes. Your number depends on the source environment, the modernization scope, your licensing position, and the AWS commitment you are willing to make. Use them to calibrate expectations before the assessment produces a real figure.
Every program above shares one mechanic: it is filed by an AWS partner inside the Partner Network, not by you on a public form. Understanding this mechanic is what separates teams that get funded from teams that pay full price for an identical migration.
The funding lives behind APN — the AWS Partner Network — and its engagement portal, ACE (APN Customer Engagements). A partner with the right competency and tier registers your migration as a structured opportunity and files the MAP record against it. The record is not a form you can reach; partner tier gates access, and Advanced or Premier tier partners are the ones who can file MAP-scale engagements.
When a partner files a migration funding record, it carries a specific set of fields. AWS reviewers pattern-match against them, which is why a well-constructed record gets funded and a thin one gets downgraded:
A DIY migrator has none of this and no way to submit it. That is the entire reason the funding goes unclaimed: not that the team is ineligible, but that there is no door for them to knock on. The fix is not "apply harder" — it is to have a vetted partner file the record. That is the single thing CloudRoute exists to arrange.
| Program | What it covers | Typical value | How to unlock |
|---|---|---|---|
| MAP — Assess | TCO + readiness assessment, application discovery, business case | Usually $0 to you (AWS-funded) | Partner files MAP record via APN |
| MAP — Mobilize | Landing zone, security baseline, pilot migration, runbook | ~25% of qualifying migration cost | Partner advances MAP record to Mobilize gate |
| MAP — Migrate & Modernize | Production cutover at scale + selective modernization | Up to ~50% of qualifying migration cost; $25K–$500K+ by size | Partner advances to Migrate gate; scales with committed spend |
| OLA | Windows/SQL right-sizing + licensing optimization | Six-figure licence savings on large SQL estates | Partner runs AWS-funded OLA assessment |
| Partner incentives | Funds the partner to deliver Assess/Mobilize/Migrate work | Invisible to you; enables low/$0 cost migration | Built into a partner-led MAP engagement |
| AWS Activate | General-purpose steady-state AWS consumption (not the project) | $5K–$100K+ depending on stage | Partner-filed via ACE; stacks with MAP |
| EDP | Committed-use discounts on the ongoing AWS bill | Negotiated; enterprise scale | AWS account team; for large committed spend |
Capable engineering teams migrate to AWS unfunded every week. It is rarely a skills gap — it is a structural one. Here is exactly where the money leaks, point by point.
Hour 0 — You submit an inquiry to CloudRoute. Four things: company, source environment (Heroku / on-prem / GCP / Azure / etc.), rough workload size, and target timeline. Total form time: about 3 minutes.
Hour 0–4 — CloudRoute scores the inquiry for funding eligibility — is this a MAP-scale migration, an OLA candidate, an Activate-stack case, or a smaller migration where the value is a de-risked partner-led cutover? You get an honest read on which bucket you are in.
Hour 4–24 — Routed to a vetted AWS partner matched to your source platform (Heroku, VMware, GCP, Azure, on-prem), your database estate (Oracle, SQL Server, Postgres), and the AWS competency that qualifies them to file MAP. You receive a scheduling link.
Hour 24–48 — Discovery call (30–45 minutes). The partner confirms eligibility, outlines the funding programs that apply to your migration, and sketches the Assess → Mobilize → Migrate plan with a directional funding range. You decide whether to proceed.
Hour 48–72 — If proceeding, the partner kicks off the Assess phase and files the MAP record through APN. Application Discovery Service starts inventorying the source environment; the business case and TCO take shape over the following 2–4 weeks.
Throughout — CloudRoute is paid by the partner as a routing commission. You pay $0 to CloudRoute. On a qualifying migration, the migration itself is funded by AWS through MAP and partner incentives — so your cost for the cutover is low to nothing.
There are three ways to get onto AWS. Founders often assume DIY is the cheapest because it has no invoice — but DIY forfeits the AWS funding entirely, which usually makes it the most expensive option in total.
| Variable | Funded partner-led (CloudRoute) | DIY in-house | Unfunded consultancy |
|---|---|---|---|
| Who files MAP / OLA | Vetted AWS partner via APN | Nobody — no access | Sometimes, if they are an AWS partner |
| AWS migration funding captured | Yes — MAP + OLA + incentives | None | Only if the firm files it |
| Assess-phase business case | AWS-funded, partner-built | Skipped | Billed to you |
| Migration engineering cost | Low → $0 on qualifying migrations | Your engineers full-time | Full consulting rate |
| Licensing optimization (OLA) | Modeled + funded | Usually unmodeled | Maybe |
| Cutover risk | AWS-vetted runbook + rollback | On you | Depends on the firm |
| Cost to start | $0 (CloudRoute paid by partner) | Opportunity cost of your team | Engagement fee upfront |
Situation: Data-center lease expiring in 9 months, forcing a cloud move. Strong in-house platform team fully intended to migrate themselves — VMware-based estate plus a large SQL Server licensing bill they assumed would simply carry over to AWS. No business case, no awareness that AWS would fund any of it, and a licence renewal quote that made the whole project look expensive.
What CloudRoute did: Routed within 20 hours to an EU partner with the Migration competency and SQL Server depth. Partner filed the MAP record and ran a free Assess phase (Application Discovery Service inventory + TCO) in week 1–3, then an OLA in parallel. The OLA modeled moving the bulk of SQL Server to Aurora PostgreSQL via SCT + DMS, eliminating most of the licence cost; the in-house team kept ownership of the application layer and ran the cutover alongside the partner.
Outcome: MAP credited a large share of the Mobilize + Migrate work; combined with the OLA-driven licence elimination, the migration ran at near-zero net cost against the avoided licence renewal. Production cutover completed inside the lease deadline. Total CloudRoute cost to the customer: $0 — the commission was paid by the partner from AWS engagement funding.
engagement window: ~5 months · funding: MAP Assess+Mobilize+Migrate + OLA licence elimination · cost to customer: $0 to CloudRoute, near-$0 net migration · servers moved: ~70
CloudRoute routes you to a vetted AWS partner who files the MAP record, runs the funded Assess phase, and de-risks the cutover. You pay $0 to CloudRoute — and on a qualifying migration, little-to-nothing for the migration itself.