Concession bands, read by product.

Red Hat publishes its list price in a visible price book. The published figure carries weight in the early posture of a renewal conversation, since it appears to set a ceiling. In practice, across signed Red Hat contracts in the trailing twelve months, the contracted figure sits well off list, and the distance off list varies materially by product line. The negotiation does not happen at list. The negotiation happens inside the concession band that comparable enterprises have already signed, which is precisely the figure the buyer does not have. Red Hat concession bands are the missing input.1

The structural reason is straightforward. List price is a published artefact. The contracted figure is a private artefact. The seller knows where the band sits, since the seller assembles it transaction by transaction. The buyer, encountering one renewal every one to three years, does not. The asymmetry compounds when the buyer treats the first concession offered as the floor, or treats the published list figure as gravity. Neither is correct, and the corrective is not a sharper opening posture; the corrective is the benchmark itself, read at the product level rather than at the portfolio level. The companion brief on Red Hat list price against concession bands sets out the gap in detail.

A blanket Red Hat discount benchmark obscures the structure that produces the figure. A portfolio that averages a sensible looking concession can hide a RHEL line at a band character much narrower than the buyer assumed, alongside an OpenShift Plus line where the bundle distorts the apparent concession upward. The two move in different directions and at different rates. Averaging them produces a number that supports no specific negotiation. The point of a benchmark is to support one specific negotiation, on one specific renewal, against one specific product mix.

The RHEL standard subscription band is narrow. The product is mature, the entitlement model is well understood inside Red Hat sales, and the comparable customer set is large enough that the contracted figure moves inside a tightly defined window. Renewal uplift on a stable RHEL footprint, absent a posture event, sits inside a familiar range and rarely surprises. A buyer who arrives at the negotiation with no benchmark will often accept a figure that sits a small step away from the achievable floor, but the absolute distance, in basis points, is modest by Red Hat standards.

The narrow band is not a reason to skip the benchmark. It is a reason to compress the engagement around posture and around the precise placement of the renewal inside the band. Two RHEL contracts of comparable size and region will rarely close at the same figure, but the distance between the lower and upper edges of the achievable window is smaller than for any other product line in the portfolio. The benchmark resolves the question of where inside that window the buyer is currently sitting, which is the question most buyers cannot answer from public material.

The narrow RHEL band also responds differently to deal size than the wider bands. A growth on a RHEL renewal that crosses a deal desk threshold opens a small additional concession surface, but the surface is bounded; RHEL is not the band on which deal desk latitude moves most. The lever that moves RHEL most reliably is the buyer's posture around exit, which is treated separately in the practice brief on exit planning and is also visible in renewal posture across the practice's benchmarking record.

The OpenShift Plus band is wide. The bundle is the reason. OpenShift Plus combines the container platform with Advanced Cluster Management, Advanced Cluster Security, Quay, and Red Hat Data Foundation into a single SKU, and the published list arithmetic for the bundle is structured so that the headline concession against list reads larger than the contracted concession against the underlying components would suggest. The bundle, in other words, distorts the apparent concession upward, and a benchmark that does not normalise for bundle composition will report a figure that flatters the deal.2

The corrective is to read the OpenShift Plus band against the constituent entitlements the buyer actually intends to use. A buyer who consumes only the container platform and the security component, and who has no realistic deployment plan for the storage or the management plane, is buying a bundle whose contracted concession looks generous and whose effective concession against consumed entitlement is materially smaller. The practice tracks the bundle band and the underlying entitlement band as two separate figures, and supplies both during a renewal engagement so the buyer can see the gap.

The wide band also creates the largest deal desk gradient in the Red Hat portfolio. Above the threshold the bundle moves into a different latitude window, and the achievable concession surface opens materially. Below the threshold the standard rep playbook applies and the surface compresses. The asymmetry is not visible to the buyer from the front of the negotiation. The benchmark, read by product and by deal size, makes it visible. Sister reading on the bundle at renewal is held in the OpenShift Plus bundle at renewal brief, and at the practice level on the OpenShift hub.

The Ansible Automation Platform band is moderate, and the moderation is conditional. The product carries a managed node entitlement model that interacts with the buyer's actual node inventory in ways the buyer is rarely able to verify without a subscription assessment. The contracted concession on the headline figure sits inside a wider window than RHEL and a narrower window than OpenShift Plus. The wrinkle is that the apparent concession can sit inside the moderate band while the effective cost per managed node, taken against the buyer's true node count, sits well outside the band that comparable enterprises with a clean node reconciliation actually pay.

The conditional character of the Ansible band is the reason the practice does not deliver an Ansible benchmark without a paired entitlement reconciliation. A benchmark figure attached to a node count the buyer cannot defend at audit is not a useful figure. The audit posture on Ansible has tightened across the trailing twelve months, and the band that comparable enterprises with clean reconciliations have signed has moved further from the band that those without have signed. The two are not the same engagement and should not be read against the same benchmark.

The Ansible band also responds to bundle posture, since Red Hat increasingly positions the platform as part of a broader automation surface that touches OpenShift and Satellite. A buyer who is renewing all three simultaneously will see a different Ansible band than a buyer who is renewing Ansible standalone. The practice tracks both, and supplies the band most relevant to the buyer's actual renewal sequence rather than the band most flattering to the seller's proposal. The Ansible practice hub sets the surrounding context.

The JBoss EAP and middleware band is wide. The width does not reflect a flexible deal desk; it reflects the legacy posture of the installed base. JBoss footprints in 2026 are concentrated in enterprises that have run the platform for many years and that have not made a decisive technical move off it. Red Hat's commercial posture around JBoss reflects that reality. The band moves on the buyer's posture around modernisation and around the alternative middleware estates that genuinely compete for the workload, rather than on the bundle dynamics that drive OpenShift Plus.

The practical consequence is that a JBoss benchmark read in isolation, without a parallel exit posture, will return a figure that sits high in the band. A JBoss benchmark read alongside a credible alternative posture, even a posture the buyer does not intend to execute, returns a figure that sits materially lower. The lever is buyer posture, not seller flexibility. The figure on the seller's first proposal reflects the seller's reading of where the buyer's posture actually sits, which is rarely where the buyer believes it sits.3

The JBoss band also carries the longest tail of conditional concessions tied to multi year commitment, since the seller's incentive on a legacy product is to lock the term rather than to widen the discount inside a single year. A buyer who accepts a longer term in exchange for a deeper headline concession often signs a figure that is, on a per year basis, indistinguishable from the figure available at a shorter term. The benchmark resolves the trade. The JBoss practice hub sets out the surrounding modernisation and concession dynamics.

Bands narrow at the top and at the bottom, and the threshold at which the band shifts is product specific. A RHEL contract that crosses the deal desk threshold opens a small additional surface; an OpenShift Plus contract that crosses the same threshold opens a much larger one. A buyer who reads the bands at the portfolio level, without regard to deal size, will misjudge the latitude available on the product line that actually drives the renewal. The benchmark, read by deal size, makes the latitude visible per product and per region.

Regional behaviour reinforces the same point. The North American band on OpenShift Plus differs materially from the EMEA band, and from the APAC band, in ways that reflect regional sales leadership latitude and the comparable customer set in each region. A US benchmark applied to a UK estate, or a European benchmark applied to a US subsidiary, returns a figure neither side will recognise as fair. The practice supplies the band that matches the buyer's actual contracting region, not the global average that flatters the seller's proposal.

The interaction of product, region, and deal size produces the actual figure that supports a negotiation. A single product specific band, read against the buyer's region and deal size, is the input that supports a credible counter. A blanket Red Hat discount figure is not. The benchmark is the ground on which the negotiation stands, not the answer to it. The figure on the seller's proposal is rarely the figure the seller is prepared to sign.

Preparing a response? The practice keeps a one-page Red Hat audit response checklist — what to acknowledge, what to preserve, and what not to volunteer in the first fourteen days after the letter arrives.