The benchmarking trap
At some point in every AMS negotiation, a slide appears: “market benchmark” rates, “industry standard” SLAs, a staffing pyramid presented as best practice. Here is the uncomfortable fact about that slide: the genuinely authoritative benchmark data for SAP application management — the analyst work by firms like Gartner, ISG, and Everest Group — sits behind paywalls, and what circulates free is overwhelmingly vendor marketing dressed as data. We verified this the direct way. Gartner’s “Seven Key Performance Metrics in SAP Services” is subscriber-only. Everest Group’s Application Management Services PEAK Matrix 2024 is a qualitative vendor-positioning assessment — publicly, it contains no pricing, SLA, or staffing-ratio numbers at all.
So when a bidder’s deck quotes a precise “market” day-rate or a standard senior-to-junior ratio with no attributable source, the right response is not to lower your guard — it is to raise it. This article will not add another invented number to the pile. It offers the framework instead: the levers to negotiate, the metrics that resist gaming, and the questions that reveal whether a vendor’s claims are real.
The four SLA levers, unbundled
A workable SLA is not one number; it is four independent levers, and conflating them is the most common buyer error. First, priority — which should be derived, not asserted: the established service-management model derives priority from impact (how much of the business, how many users) crossed with urgency (how fast it must be fixed), mapping to tiers commonly labeled P1–P5. Second and third, response versus resolution — acknowledgment time and fix time are different commitments and must be separately specified. Fourth, coverage — business-hours 8×5 versus 24×7, which typically differs by tier: around-the-clock for P1/P2, business hours for the rest.
On the actual times to attach to each tier: the sources that publish reference tables are explicit that these are reference SLAs to be calibrated to your business hours, service type, and contractual obligations — not standards. Treat any vendor’s tier table as their opening negotiating position. Your counter comes from your own ticket history and business calendar, not from someone else’s table.
Priority done right: making “critical” auditable
The impact-times-urgency matrix earns its place for a governance reason, not an aesthetic one: it makes “critical” a defined term. Without it, P1 is whatever the vendor’s duty manager says it is at 02:00 — or whatever your angriest stakeholder insists it is. With it, priority assignment becomes auditable: you can sample a month of tickets and check whether the definitions were applied. A payroll defect on payroll-run day is the canonical example the framework handles well — modest impact by user count, maximal urgency by clock. Write the definitions with examples from your own landscape into the contract schedule.
Quality SLAs, not just clock SLAs
Time-based SLAs alone are gameable: close-and-reopen games, ticket splitting, priority inflation and deflation. Balance the clock metrics with quality metrics. First-time fix is among the metrics most strongly correlated with satisfaction, with rates below roughly 70% signaling training, process, or empowerment problems — a figure from the contact-center discipline, flagged as such: directional for an SAP desk, not an SAP-specific standard. Cost-per-ticket — total monthly support cost divided by tickets resolved — normalizes bids onto a comparable unit that headline day-rates obscure. Backlog aging and reopen rate complete the set. And the metric almost nobody contracts for, though it defines a good AMS partner: did future ticket volume fall? A desk that only resolves tickets is running your queue; a desk that eliminates ticket causes is running your estate.
The seniority question — benchmark outcomes, not ratios
Buyers intuitively ask for a staffing-ratio benchmark: what senior-to-junior mix should I demand? The honest answer is that no defensible public source produces a citable ratio — and the intuition behind the question is still correct: the dominant cost variable in AMS is who does the work, not the posted rate. Senior-led delivery resolves in hours what junior-heavy teams take days on, and a strong team drives future volume down rather than billing on the current queue.
The way out is to benchmark the outcomes that seniority produces instead of a ratio nobody can source: time-to-resolve by priority tier, reopen rate, escalation rate, and the trajectory of ticket volume over the contract. A vendor confident in its seniority mix will contract on those outcomes. A vendor that resists outcome metrics while asserting a pyramid diagram has told you something useful.
Where the real numbers live
If your negotiation genuinely needs hard market figures, they are obtainable — honestly, they usually cost money. The analyst houses (Gartner, ISG, Everest Group) sell exactly this research; a sourcing advisor can run a structured RFP that generates real competitive data specific to your estate; and SAP user groups (DSAG, UKISUG, ASUG) periodically survey members on support economics. What none of these will give you is the free, universal, precise benchmark the vendor slide implied — because estates, scopes, and coverage models differ enough that a single “market rate” is a fiction even before the marketing gloss. One adjacent, sourced data point shows what is knowable: organizations that implemented leading-practice shared-services delivery models have reduced overall operating costs by 20–40% — a shared-services figure, not an SAP-AMS benchmark, cited here only as evidence that delivery-model design moves cost materially. Note what we just did: named the source, named the domain, refused to reposition it. Demand the same of every number you are shown.
The buyer’s question set
Put these to any AMS bidder, and watch which ones get straight answers:
- For each number on your benchmark slide — what is the attributable source, and may we see it?
- Show your proposed P1–P5 definitions as impact-times-urgency rules with examples. Who assigns priority, and how do we audit it?
- Response and resolution, separately, per tier — and which tiers carry 24×7 coverage?
- What first-time-fix and reopen rates do you commit to, and how are they measured?
- What is your cost-per-ticket on comparable engagements, and how will it be reported to us monthly?
- Which named individuals work our estate in months one through six, and what happens to our SLA if they are substituted?
- Will you contract on ticket-volume reduction over the term? If not, why not?
- Which of your claims should we expect to verify against analyst or user-group data, and which are your own delivery statistics?
A vendor with real delivery strength answers these gladly — the questions favor them. Evasion is itself a data point.
Transparency as the differentiator
We have an obvious interest to declare: IOTEK sells senior-led SAP application management, so we are a participant in this market, not a neutral observer. That is exactly why this article quotes no number it cannot source. The vendors whose benchmarks are marketing artifacts cannot adopt this position without abandoning their own slides; anyone can adopt the question set above without buying anything. Use it on us, too.
The most honest sentence in AMS benchmarking is: “that number has no public source.” The buyers who insist on sourced numbers, defined priorities, and outcome-based seniority metrics do not just negotiate better contracts — they select for the vendors who can survive the scrutiny.
Sources
| # | URL | Publisher |
|---|---|---|
| 1 | https://blog.invgate.com/itil-priority-matrix | InvGate |
| 2 | https://rootly.com/incident-response/support-levels | Rootly |
| 3 | https://www.scottmadden.com/insight/designing-leading-practices-service-delivery-model-shared-services/ | ScottMadden |
| 4 | https://www.everestgrp.com/the-application-management-services-peak-matrix-assessment-2024/ | Everest Group |
| 5 | https://www.gartner.com/en/documents/3488126 | Gartner (paywalled — cited as verification of paywall/absence of public numbers) |
| 6 | https://daxsws.com/blog/ten-key-managed-services-delivery-metrics-to-measure | DAX |
| 7 | https://www.armatis.com/en/2026/02/25/how-to-measure-the-performance-of-your-outsourced-contact-centre-kpis-and-slas/ | Armatis |