Ask a manufacturing CIO about 2027 and you will get a fluent answer about S/4HANA. Ask the same person what happens to their PI/PO landscape on that date and the answer changes shape. There is usually a pause, then some version of: that is the middleware team's problem, and we think we have until 2030.
Both halves of that sentence are worth taking apart, because the second one is the expensive misunderstanding.
What actually happens to SAP PI/PO in 2027?
SAP PI/PO 7.5 reaches the end of standard maintenance on December 31, 2027. Extended support runs to December 31, 2030, and after that there are no updates, no patches, and no support of any kind.
The gap between those two dates is where most planning goes wrong. Extended support is a holding pattern, not a runway. It costs a premium, it delivers no new capability, and it does not protect you from the thing that actually breaks integration landscapes: the systems on either end of the interface moving on without you.
That is the part worth sitting with. Your PI/PO estate does not fail because SAP stops patching it. It fails because your S/4HANA core, your cloud applications, and your trading partners keep changing, and your middleware stops being able to follow them. An unsupported integration layer between a modern ERP and a modern supply chain is not a stable system that happens to be old. It is a widening mismatch.
The collision most roadmaps miss
Here is the detail that should reorder your planning: SAP ECC and SAP PI/PO lose mainstream maintenance on the same day.
Two large programs, one date, and in most mid-market manufacturers, one budget cycle and one small group of senior people who can credibly run either. The ERP program is visible, board-sponsored, and staffed. The integration program is a workstream inside someone else's plan, or it is nowhere.
The competition is not really for money. It is for architects. There is a fixed and shrinking population of people who have taken a manufacturing landscape through a cutover, and the same people are qualified for both programs. They are booked earliest by the clients who plan earliest. A manufacturer who scopes both in the same conversation gets a coherent sequence. A manufacturer who scopes them a year apart gets whoever is left.
There is also a straightforward technical reason to plan them together. If your S/4HANA migration is going to change the systems at one end of two hundred interfaces, and your integration migration is going to change the layer in the middle, doing them blind to each other means rebuilding some interfaces twice. Interfaces you were going to retire in the ERP program should not be migrated in the integration program. Interfaces that need redesign for S/4HANA should be redesigned once, on the target platform, not lifted to Integration Suite in their ECC-era shape and then reworked.
That sequencing decision is worth more than any tooling choice on this page, and it can only be made if both programs are on the same table.
What replaces PI/PO, and what it is not
SAP Integration Suite on SAP BTP is the strategic successor. It is not an upgrade. It is a different platform with a different architecture, which means every interface is a migration decision rather than a version bump.
What you get in exchange for that work:
Cloud Integration handles process orchestration and messaging, the direct functional descendant of what PI/PO does today.
API Management covers the layer PI/PO never really had. Governance, versioning, rate limiting, and OAuth-based security on the services you expose, which matters as soon as anything outside your firewall consumes them.
Event Mesh enables event-driven and real-time patterns rather than the batch-and-poll rhythms most ECC-era landscapes were built around. This is where the material process improvements usually come from.
Open Connectors ships more than 160 pre-built SaaS integrations, which removes a large category of custom adapter work.
B2B integration covers EDI, AS2, and SFTP with trading partner management, so your 850s, 856s, and 810s move without a bolt-on product.
Edge Integration Cell runs integration on your own premises when latency, data residency, or a plant-floor system requires it. For manufacturers with MES or SCADA connectivity that cannot round-trip to a public cloud, this is the component that makes the whole architecture viable, and it is frequently missed in early scoping.
How long does PI/PO to Integration Suite migration take?
Timeline scales with interface count more reliably than with anything else:
Read those against the calendar. From today, a manufacturer with 200-plus interfaces who has not started is already outside the window for a comfortable landing before December 2027, and is planning against the 2030 extended-support date whether they intended to or not.
The interface count is knowable this week. Most organizations discover their real number is higher than the documented one, because interfaces accumulate quietly and inventories go stale. That single number determines your entire planning horizon, and getting it wrong by fifty interfaces moves your end date by months.
What does it cost?
Published ranges, which vary considerably with scope and how much redesign you take on:
Return typically lands in 12 to 24 months, through reduced maintenance, retired infrastructure, and eliminated manual work around interfaces that were failing silently.
One structural warning. Integration Suite is consumption-priced, unlike the fixed licence you are used to. Organizations that lift high-volume interfaces without examining message design sometimes find the running cost unpleasant. Message volume is a design decision on this platform in a way it was not on PI/PO, and it belongs in the architecture phase rather than the first invoice.
Integration Suite, Workato, or custom: how to choose
Most manufacturers do not pick one. They place workloads.
The practical rule: SAP Integration Suite for the SAP core, because clean-core compliance, S/4HANA extensions, and SAP-to-SAP scenarios are what it is built for and what SAP will keep investing in. Workato for the periphery, where the value is rapid SaaS connectivity, low-code delivery to business teams, and agentic AI orchestration using the MCP protocol. Custom only where a proprietary protocol or a piece of specialized manufacturing equipment leaves no alternative, and with clear eyes about who maintains it in 2032.
Anyone who tells you one platform covers all three cases without trade-offs is selling a platform.
What the assessment actually tells you
SAP's Migration Assessment Tool analyzes your existing interfaces and sorts them into three buckets, and the distribution across those buckets is the single most useful number in your business case:
Ready to migrate. Standard patterns that convert largely automatically. Cheap, fast, low risk.
Adjustment required. Minor rework to fit the new platform. Predictable effort.
Redesign recommended. Interfaces that should be rebuilt as APIs or events rather than carried across as-is. Expensive, and where the actual modernization value sits.
An estate that is 70 percent ready-to-migrate is a different program, with a different budget and a different timeline, than one that is 40 percent redesign. You cannot estimate this from the interface count alone, and a proposal that prices your migration without having run this analysis is pricing a guess.
When you should not migrate an interface
The modernization pitch oversells redesign, so here is the counter-case.
Some interfaces should be retired, not migrated. In most landscapes a meaningful share of interfaces serve processes that changed years ago or feed reports nobody opens. Migrating them converts dead weight into modern dead weight at full cost. The inventory is the moment to kill them, and it is the only moment anyone will fund the review.
Lift-and-shift is often correct. A stable, low-volume, standard-pattern interface that works does not become better for being rebuilt as an event-driven flow. Redesign is justified where the pattern is genuinely limiting the business, not as a default posture. Redesigning everything is how an 18-month program becomes a 36-month one.
Integration Suite is not better at everything. Certain PI/PO scenarios, particularly deeply custom adapter modules and some stateful orchestrations, have no clean equivalent and need genuine rework rather than translation. If a partner tells you every scenario maps cleanly, they have not inventoried your estate.
And you cannot cut over all at once. Phased migrations run PI/PO and Integration Suite in parallel, which means a period of paying for both and operating both. That overlap is a real cost and belongs in the business case from the start. It is also the single largest risk reducer available to you, which is why it is worth paying for.
What good execution looks like
A phased program with a parallel run, sequenced by business criticality rather than by technical convenience. Interfaces move in waves, the old platform stays live behind them, and each wave validates before the next starts.
One industrial equipment OEM ran exactly this shape across 150-plus PI/PO interfaces supporting global manufacturing operations, connecting more than a dozen systems with GDPR exposure across the estate. Every interface moved to SAP BTP with zero production downtime, and integration costs came down 60 percent.
Zero downtime on a landscape that size is not a function of tooling. It comes from wave sequencing, parallel operation, and rehearsing cutover before it happens.
Security is architected in rather than reviewed at the end: TLS 1.2 or higher in transit and at rest, OAuth 2.0 and SAML authentication, role-based access control, rate limiting, threat detection, and audit logging. For regulated manufacturing that extends to ITAR data sovereignty, FDA 21 CFR Part 11 validation, and ISO 27001 controls, with penetration testing as standard rather than optional.
Frequently asked questions
Is SAP PI/PO still supported?Yes, but not indefinitely. Standard maintenance for SAP PI/PO 7.5 ends December 31, 2027. Extended support continues to December 31, 2030, after which there are no updates, patches, or support. Most organizations are migrating now to avoid competing for delivery capacity in 2027.
Can I upgrade PI/PO to SAP Integration Suite?No. There is no in-place upgrade path. Integration Suite is a different, cloud-native platform on SAP BTP, so every interface requires a migration decision: migrate as-is, adjust, redesign, or retire.
How many interfaces can be migrated automatically?It depends on your estate. SAP's Migration Assessment Tool classifies interfaces as ready to migrate, adjustment required, or redesign recommended. Standard patterns in the first category convert largely automatically. The proportion varies widely and can only be established by running the analysis against your actual landscape.
Do I need to migrate to S/4HANA first?No, but the two should be planned together. Both lose mainstream maintenance on December 31, 2027, and they compete for the same architects. Sequencing them jointly avoids migrating interfaces you are about to retire, or building them twice.
What if my systems cannot connect to the public cloud?Edge Integration Cell runs Integration Suite capability on your own premises for latency-sensitive, data-residency-restricted, or plant-floor scenarios such as MES and SCADA connectivity, while keeping a single management plane.
How do I start without committing to a full program?Begin with an interface inventory and a migration assessment. Two to three weeks establishes your real interface count, the readiness distribution, and a costed roadmap, which is enough to make the budget decision without committing to a delivery partner.
If you take one action from this page, make it the interface inventory. Not the platform decision, not the partner selection, and not the business case. Almost every bad PI/PO migration estimate traces back to an organization that did not know its own interface count and priced the program on the documented number rather than the real one.
Get Your PI/PO Migration Roadmap in 14 Days
A fixed-fee, two-week integration assessment for manufacturers planning the move off SAP PI/PO. Delivered by a senior integration architect, not a proposal team.
We'll map your:
- Complete PI/PO interface inventory and topology, including the interfaces your documentation has lost track of.
- Readiness scoring against SAP's assessment categories, so you know your ready-migrate, adjust, and redesign split before you budget.
- Recommended platform placement across SAP Integration Suite, Workato, and any genuinely custom cases.
- Phased migration roadmap with parallel-run strategy, sequenced against your S/4HANA plan rather than around it.
No sales drip. An integration roadmap on day fourteen, and one reply from a senior architect within one business day.
