How a European insurer renegotiated CA AIOps from a $14M anchor to $6.2M.

The Desk took on a CA AIOps conversion in late 2025 for a multi country insurance group headquartered in Western Europe. The seller's opening proposal anchored at $14M across four years, built around a legacy contract structure that the buyer's IT operations team had inherited from a 2021 signature and had never re examined in any detail. The buyer's procurement leader brought the engagement to us with a sense that the number was high but without a framework to argue against it. The signed contract closed in March 2026 at $6.2M across the same four year term, on a converted subscription architecture matched to the buyer's actual operational footprint. The headline reduction was 56 percent. The work was almost entirely in the entitlement reconciliation that preceded the commercial conversation.

What follows is the negotiation arc of the engagement, presented without the insurer's identifying details. The figures are as the contract reads.

The Quote

The seller's opening proposal landed in October 2025 as part of the AIOps conversion motion that the seller has been running across its mid 2026 book. The proposal carried a four year total contract value of $14M, structured as $3.2M in year one, $3.5M in year two, $3.6M in year three, and $3.7M in year four. The conversion was positioned as a modernisation from the buyer's legacy AIOps contract structure (originally signed under the previous ownership and renewed in 2023) onto the seller's current subscription architecture.

The proposal carried a monitored node count of 18,400 nodes across the insurer's European, Latin American, and Asia Pacific operations. It carried an event ingestion volume cap of 4.2 billion events per month. It carried a feature bundle that included the full AIOps platform, the predictive analytics module, the integration platform for the buyer's ITSM and CMDB tools, and the topology mapping engine. The seller's narrative was that the conversion preserved continuity, normalised the pricing structure, and aligned the contract with the seller's current commercial architecture. The buyer's procurement team treated the proposal as a starting point for a 10 to 15 percent negotiation. The buyer's IT operations director suspected the node count was wrong but did not have time to prove it.

The Find

The Desk spent the first four weeks of the engagement on entitlement reconciliation. The first task was to pull the actual deployed AIOps node footprint from the buyer's configuration management system and compare it to the contracted count. The exercise produced a corrected count of 11,200 nodes against the contracted 18,400. The delta was 7,200 nodes, or 39 percent of the contracted population. The delta was explained by a 2023 divestiture of a Central European business unit that had carried 3,800 nodes, a 2024 decommissioning of a legacy on premise monitoring footprint in two regional offices (1,900 nodes), and an ongoing rationalisation of redundant agents across multi tier applications that had reduced the per host agent density.

The second task was the event ingestion volume. The contracted cap was 4.2 billion events per month. The buyer's actual trailing six month ingestion volume averaged 1.8 billion events per month, with a peak of 2.4 billion in a single month. The contracted cap was over twice the buyer's actual operational requirement. The buyer was paying for a capacity reservation that the buyer's own platform had never needed.

The third task was the feature bundle. The predictive analytics module had been deployed in pilot in 2023 and never moved to production because the buyer's data science team had built an equivalent capability in house on a different platform. The integration platform for the CMDB tool was actively used. The integration platform for the ITSM tool had been replaced in 2024 by a different vendor's integration layer that did not require the AIOps connector. The topology mapping engine was in production and was the feature the buyer most depended on across the bundle.

The fourth task was the credit for the unused entitlement on the legacy contract at the conversion date. The buyer's legacy contract had been pre paid in a structure that left roughly $1.1M of unused entitlement at the conversion date. The seller's proposal did not credit any of this back to the buyer. The buyer's procurement team had not noticed.

"Three years of operational change had moved the actual AIOps footprint a long way from the contracted one. The seller's proposal was built on a 2021 picture of the buyer that no one inside the buyer had bothered to update. The negotiation was almost entirely the reconciliation."CA AIOps Engagement Lead, The Desk

The Restructure

The restructure proposal that went to the seller in November 2025 had four commercial components and three structural components. The commercial components were a corrected node count of 11,500 (a small buffer above the 11,200 actual to absorb routine growth), an event ingestion volume cap of 2.5 billion events per month (sized at roughly the trailing peak plus a modest reserve), a feature bundle restructured to exclude the predictive analytics module and the ITSM integration platform, and a credit of $1.0M for the legacy entitlement balance at the conversion date.

The three structural components were an escalator capped at 3.5 percent compounding rather than the proposal's implied 9 percent year over year, a defined coverage tier with a 90 day cure window on any audit finding, and a documented data extraction obligation on exit that required the seller to provide a complete export of the buyer's telemetry, configuration, and topology data within 30 days of any termination at no additional cost.

The seller's deal desk rejected the first version in week one of the formal proposal. The grounds were that the conversion price was built around a bundle and could not be unbundled. The Desk presented the rate card for each individual module, which showed that the unbundled price of the components the buyer required was lower than the embedded bundle price by a meaningful margin. The deal desk reopened the conversation in week three. The escalation to the regional vice president happened in week six, primarily on the question of the legacy entitlement credit. The deal desk's initial position was that the legacy entitlement was not transferable. The buyer's contract counsel produced the original 2021 agreement, which contained a transferability provision the seller had overlooked. The credit was conceded in week eight.

The Outcome

The signed contract closed in March 2026 at $6.2M across four years, structured as $1.42M in year one, $1.50M in year two, $1.58M in year three, and $1.70M in year four. The escalator landed at 4 percent compounding, slightly above the requested 3.5 percent in exchange for the year one number coming down further. The contracted node count was 11,500. The event ingestion cap was 2.5 billion per month. The feature bundle excluded the predictive analytics module and the ITSM integration. The data extraction obligation on exit was included. The legacy entitlement credit was applied to the year one number as a $980K reduction (against the requested $1.0M). The audit clause was negotiated to include the 90 day cure window.

What we have seen on live deals

The CA AIOps conversion pattern in this case is repeating across our 2026 book. Buyers who have held legacy AIOps contracts since the previous ownership are being moved onto the current subscription architecture on opening proposals that are anchored to legacy entitlement assertions rather than to current operational footprints. The reductions available on the conversions are large because the operational footprint has almost always shifted since the legacy contract was signed, and the seller has no incentive to surface the shift. The buyer who insists on the entitlement reconciliation before the commercial conversation captures the largest reduction. The buyer who treats the conversion as a paperwork exercise renews the wrong number for another contract cycle.

Two procedural notes that travel across this book. First, the seller's deal desk in 2026 will accept entitlement corrections when they arrive with documented support but will resist them when they arrive as procurement assertions. The configuration management extract, the trailing event ingestion volume, the platform usage reports, and the original transferability provisions are the four documents that move the conversation. None of them is exotic. All of them require the buyer's IT operations team to spend the time. Second, the conversion conversation is a one time opportunity. Once the contract is signed onto the current architecture, the legacy entitlement assertions are extinguished and the next renewal will be anchored to the new contract. The reconciliation work has to happen at the conversion point or it does not happen at all. The buyer who postpones the reconciliation to the next renewal is postponing a conversation that will no longer exist.

The takeaway

• The 56 percent reduction on this AIOps conversion came mostly from four reconciliations done before the commercial conversation. Node count corrected against the buyer's configuration management system. Event ingestion cap sized against the actual trailing volume. Feature bundle reduced against the buyer's actual production use. Legacy entitlement credit applied against the original transferability provision.
• The structural concessions (escalator cap, audit cure window, data extraction obligation) were worth less in year one dollars but materially more in present value terms across the four year horizon. The escalator alone was worth more than 14 percent of the contract value over the term.
• The seller's conversion proposals across the AIOps book are anchored to legacy assertions, not to current operational footprints. The reductions are available to the buyer who does the reconciliation work. They are not offered to the buyer who treats the conversion as administrative.