TotalTek Blog | SAP, PMO, Maritime & AI Insights

The SAP AMS Iceberg: The Costs You Weren’t Meant to See

Written by Brad Nicolaisen | Aug 6, 2026, 4:00:57 PM

By Brad Nicolaisen, Senior Vice President, Strategic Growth & AI Innovation, TotalTek

Across industries and budgets, I’ve seen the same pattern repeat: leaders sign for a ‘predictable’ SAP AMS agreement with a capped-hours, hourly-billed model, and within a few months they’re explaining why actuals keep blowing past the base fee. Month after month, the variance memo writes itself. After hearing this enough times, I stopped accepting the surface-level story. I spent weeks inside statements of work and invoices, replayed ticket data, and interviewed delivery managers. What emerged wasn’t a random set of edge cases, it was an iceberg. The tiny, visible tip was the base fee. The mass below the waterline was a dense stack of direct and indirect costs the sales cycle rarely names and the monthly P&L can’t ignore.

The Pattern Behind the ‘Predictable’ Variance

Capped T&M looks tidy until real life arrives: a critical incident weekend, a cutover that runs long, a burst of small fixes, a few ‘must-ship’ change requests. Individually, every one of these is defensible. Collectively, they compound into a structural gap between the fee you were quoted and the money you actually spend. Finance calls it variance. Operations calls it whiplash. The root cause is simpler: the model rewards activity, not stability.

The Hidden Ledger: Direct Costs You See on Invoices

  • Overage premiums for hours beyond the cap, usually at a higher rate.
  • After-hours multipliers that apply to both capped work and overage.
  • Minimum billing increments that round micro-tasks to 15–30 minutes.
  • Expedite fees layered on top of labor for rush work.
  • Change-request processing time (scoping, approvals, deployment) billed as additional hours.
  • Specialist surcharges when architects or SMEs replace the blended rate.
  • Pass-throughs and extras: monitoring, testing, observability, travel, and FX swings.

The Second Ledger: Costs You Pay Inside Your Organization

  • Timesheet policing and invoice reconciliation to chase clarity and fix errors.
  • Backlog triage to ‘fit the cap,’ often at odds with business priority.
  • Vendor management: QBRs, exception handling, escalations, and corrective actions.
  • Audit and compliance evidence pulls across tickets, approvals, and deployments.
  • Finance forecasting and accruals for inherently volatile invoices.
  • Downtime impact measured in dollars per minute, not just SLA penalties.
  • Rework and context switching as retickets and handoffs consume capacity.
  • Breakage—unused hours that expire when carryover isn’t allowed.

Why Hourly AMS Resists Stability

Under a pure Time & Materials construct, revenue expands with motion: incidents, retickets, after-hours callouts, change cycles. There’s little structural incentive to eliminate root causes, simplify run, or reduce ticket volume. The result is predictable and corrosive: a slow, grinding drift toward firefighting. Preventative work slips. Technical debt accrues. Engineers normalize rework. You get more of what the model pays for—activity—rather than what your business needs: reliability.

From Anecdotes to Math

To stop debating feelings, I built a calculator to expose the full stack beneath the waterline. The goal was precision with usability: leaders should be able to answer, in under five minutes, what their effective monthly cost really is and which three drivers matter most.

The method is straightforward. First, gather inputs: cap hours and rates, overage rate, after-hours fraction and multiplier, expedites and fees, micro-tasks and minimum increment, change-request hours, SME hours and differential, pass-through tooling/travel/FX, internal loaded rates and the time your teams spend on admin and governance, downtime minutes and business cost, and whether carryover exists. Second, model both the Expected view (probability of breach × average overage hours; no breakage if carryover) and the Actual view (month-specific values, with breakage when carryover is disallowed). Finally, roll everything up: All-in Monthly Cost = Base Fee + Vendor-Billed Hidden Costs + Internal Costs + Downtime Impact + Breakage. Express the uplift as a percent of base and rank-order the contributors. The argument usually ends right there.

What Leaders Do When the Numbers Land

They stop normalizing variance. They ask different questions: Why are after-hours rates the norm for routine maintenance? Which tickets recur, and what would it take to erase their root causes? Why pay a minimum increment for micro-fixes in a continuous support context? Which pass-throughs duplicate tools we already own? Who benefits from breakage when carryover is off?

Then they change the incentives. Some negotiate away a few obvious line items; the durable fix is to govern to outcomes: stability, SLA attainment, change success, mean time to restore, first-time-right, and business uptime. Quarterly reviews shift from hour accounting to reliability and quality. Reticket rates and defect escapes become central metrics. Reliability work is funded explicitly. The North Star becomes fewer incidents next quarter—not more hours next month.

Why Some Providers Won’t Applaud

For a long time, the iceberg has been kind to hourly vendors. Dozens of ‘reasonable’ line items disperse attention and blunt accountability. When you consolidate them into an all-in number and measure uplift over base, the picture sharpens—and the economics change. You stop being the ‘difficult client with edge cases’ and become a rational buyer who has finally priced the model as it operates, not as it’s pitched. That clarity can be uncomfortable. It’s also overdue.

The Bottom Line

Budgets hate surprises. Operations hate firefights. The business hates watching spend rise while stability stagnates. Hourly AMS—especially capped hours with overage—can look orderly on paper and still undermine all three. You don’t have to accept that trade-off. Put the math on the table. Measure the hidden ledger with the same rigor you apply to the base fee. Then decide, with eyes open, whether the structure you’re funding creates the behavior your SAP landscape needs. If you reward activity, you’ll buy more activity. If you reward durable fixes and fewer incidents, you’ll buy a calmer, more reliable SAP.