The TCO categories SAP account teams do not show you.

The SAP TCO slide is one of the most polished documents in enterprise software. It runs to a single page, carries three or four bars, and resolves to a clean comparison between the buyer's status quo and the RISE proposal. The numbers in the bars are not wrong. They are also not complete. The slide shows subscription, infrastructure, and a thin line for services. It leaves out nine other categories that, across 500 plus engagements, account for between 28 and 41 per cent of the real seven year cost of running RISE. The work of buyer side TCO modelling is to find those categories, size them, and bring them into the same model the board will use to approve the deal.

Category one: integration and middleware

The first omission is the cost of integrating RISE into the surrounding application estate. The SAP TCO slide assumes that the integration layer carries no incremental cost under RISE, because BTP is included in the bundle. The assumption is wrong on two counts. The BTP allocation inside the bundle is sized for a baseline integration scope that almost no enterprise actually has. The real integration estate, including the bus, the API gateway, the file transfer infrastructure, the master data integration, and the analytics pipeline, requires a BTP consumption that runs two to four times the bundled allocation in years two and beyond.

The buyer side TCO model treats BTP overage as its own line. The line is sized from the current integration estate, mapped against the BTP service catalogue, and projected against a realistic growth assumption. The number is not small. On a five thousand user RISE deal, BTP overage typically runs to three to seven million dollars across the seven year term. It does not appear on the SAP slide.

Category two: hyperscaler egress and storage growth

The second omission is the cost of hyperscaler infrastructure that sits outside the bundled allocation. RISE includes a baseline infrastructure allocation that is sized at the start of the term. The allocation is rarely sized for the real growth of the data estate, the real volume of cross region traffic, or the real cost of backup retention. The hyperscaler bills for the overage at list rates, with no negotiated discount.

A workable buyer TCO model carries three hyperscaler lines outside the bundle. The first line is storage growth, sized at the historical compound annual growth rate of the database. The second line is cross region egress, sized at the current network volume. The third line is backup retention, sized at the policy the security team actually enforces. The three lines together typically add two to four million dollars across seven years on a mid sized RISE deal.

Category three: indirect access and digital access true ups

The third omission is the cost of Digital Access true ups across the term. The RISE proposal carries a Digital Access volume that is sized at signature. The volume is sized at a static document count, with no allowance for the real growth of integrated systems, the real expansion of EDI traffic, or the real volume of API calls from new applications.

The buyer TCO model treats Digital Access as a growing line, with a year one volume at the proposal level and a year seven volume sized against the integration roadmap. The growth rate is rarely below 15 per cent compound, and is often higher in organisations that are actively adding partners or expanding into new markets. The Digital Access true up across seven years typically runs to one to three million dollars.

Category four: internal team backfill and centre of excellence

The fourth omission is the cost of the internal team that runs RISE on the buyer side. The SAP TCO slide assumes that the move to RISE reduces internal headcount, on the logic that the platform is now managed. The reality is different. RISE moves operational responsibility but leaves architectural responsibility on the buyer. The buyer still needs a centre of excellence that holds the integration architecture, the security posture, the data model, and the release calendar.

The buyer TCO model carries an internal team line, sized at the headcount required to run RISE properly, with the salary loaded by full overhead. For most mid sized enterprises, the internal team for RISE runs to between eight and fourteen full time equivalents, including architecture, security, integration, release, and data leads. The line is not zero, and it is often higher than the team the buyer was running before.

Category five: change management and adoption

The fifth omission is the cost of change management across the user base. The SAP TCO slide treats the human cost of the migration as a one time training line inside the implementation services number. The line is almost always undersized. The real cost of moving five thousand users to a new platform, with new processes, new screens, new approval flows, and new analytics, runs to two to five million dollars across the first two years.

The buyer TCO model carries change management as its own line, with sub categories for executive sponsorship, super user enablement, end user training, communication, and post go live reinforcement. The model also carries an opportunity cost line, sized at the productivity dip during the cutover quarter. The dip is real, is measurable from prior migrations, and is rarely sized in the SAP proposal.

Category six: third party application licences and exit costs

The sixth omission is the cost of the third party application estate that sits around the SAP core. The estate includes the data warehouse, the analytics platform, the planning tool, the procurement platform, the EAM tool, the field service application, and the integration platform that connects them. The move to RISE often changes the integration points, sometimes triggers new licence demands from the third parties, and occasionally forces an early exit from a platform that no longer integrates cleanly.

The buyer TCO model carries a third party impact line, sized at the real cost of re integration plus any early exit fees plus any incremental licence changes. The line is rarely zero, and on complex landscapes it can run to four to eight million dollars across the term. The number is not visible on the SAP slide because the third parties are not SAP.

Category seven: contract uplift and renewal exposure

The seventh omission is the cost of the contract uplift mechanism over the term. The RISE proposal carries an uplift clause, typically linked to an inflation index, with a floor that is rarely below two and a half per cent and a ceiling that is rarely below five per cent. The SAP TCO slide carries the year one price, with a flat extension across the term. The buyer TCO model carries the uplift at the realistic mid point of the floor and the ceiling, compounded across the term.

The compounding matters. A three per cent annual uplift across seven years adds 22 per cent to the year seven invoice. The cumulative cost of the uplift across the term, against a flat extension, runs to between 8 and 14 per cent of the contract total. The buyer TCO model surfaces the number so that the negotiation can hold the uplift to the floor rather than the mid point.

Category eight: exit and reversibility

The eighth omission is the cost of exiting RISE at the end of the term. The SAP TCO slide treats the contract as if it renews automatically at the same terms. The reality is that the renewal is itself a negotiation, and a renewal that holds the buyer to disadvantageous terms carries its own cost. A reversibility option, including the cost of an internal estate that can receive the data if the contract is not renewed on acceptable terms, is part of the real seven year cost of the platform.

Conclusion: the slide is the start, not the model

The SAP TCO slide is the opening framing of the deal. It is competent, clean, and substantially incomplete. The buyer that approves the deal on the slide approves a number that is between 28 and 41 per cent below the real seven year cost. The buyer that builds its own TCO model, populated against the nine categories above, approves a number that holds against the real operating reality. The model is not difficult to build, but it does require discipline and the willingness to ask the questions that the slide does not invite. The work of the model is the work of the negotiation, conducted in numbers rather than in language.