A ServiceNow technology Now Assist cost control case study.

How the ServiceNow technology Now Assist cost control case study unfolded

A technology firm reached renewal with a ServiceNow estate that had moved quickly into AI, running Now Assist across IT service management, operations and parts of engineering. The proposed renewal bundled an enlarged assist commitment, an 11 percent annual uplift and an open overage rate that would apply once metered consumption passed the committed volume. The account team framed the assist commitment as a floor the firm would inevitably exceed. The firm engaged us buyer side to size the commitment against real consumption before agreeing to it.

The situation

Technology teams overspend on Now Assist faster than most because they adopt AI features early and run agentic actions at scale. Under the 2026 commercial model AI is bundled in every tier but assists are metered, and a large agentic action consumes materially more assists than a simple summary or a single classification. The firm had let a few high volume engineering and operations workflows drive its consumption forecast, and the vendor had projected that curve straight upward into a commitment far above steady usage. Nobody had separated the heavy agentic workflows from the light ones, so the committed volume and the overage exposure were both sized on the wrong base.

Our first task was to model consumption properly. We broke the firm's assist usage into action types, separated the large agentic actions from routine assists, and built a forecast that reflected the real mix rather than the heaviest month projected forward. The mechanics behind that analysis are set out in our Now Assist consumption guidance.

What we found

The consumption model told a different story than the renewal quote assumed. The proposed assist commitment ran roughly a third above the firm's trailing twelve month draw, even allowing for planned expansion. The forecast had been anchored to two heavy agentic workflows that were already being optimised internally, so their per action draw was set to fall rather than rise. And the open overage rate, left unstated in the quote, would have applied to a commitment the firm was unlikely to reach, turning a safety margin into a standing liability. The real exposure sat in the overage mechanics, not the headline assist number.

The negotiation

We built the strategy around four moves, sequenced so consumption and rate were settled before total price. First, a right sized assist commitment anchored to the modelled mix rather than the heaviest month. Second, a fixed overage rate written into the agreement at signature, removing the open exposure. Third, a carry forward provision so unused assists were not simply lost at period end. Fourth, a capped annual uplift stated as a number, replacing the open 11 percent.

The firm's team led every conversation. We stayed behind the table, reviewing each proposal revision, drafting counters and briefing executives before each session, in the pattern set out in our ServiceNow renewal negotiation service and the wider ServiceNow renewal guidance.

The outcome

The agreement signed two weeks before deadline. The assist commitment was cut by roughly 35 percent to match the modelled mix, the overage rate was fixed at signature, unused assists carried forward within the term, and the annual uplift was capped at 5 percent, down from the proposed 11 percent. In total the renewal closed around 19 percent below the initial quote, and the firm kept room to expand AI usage without paying for headroom it did not need.

Lessons

Three lessons carry beyond this engagement. A Now Assist commitment should be sized on the real action mix, because large agentic actions consume materially more assists and a forecast built from the heaviest workflows overstates the steady draw. The overage rate matters more than the headline volume, since an open rate on an inflated commitment is where the cost actually lands. And the uplift is worth capping as a number so it cannot compound on a base that should have been smaller. The same consumption discipline runs across industries, as our insurance Now Assist cost control case study and media Now Assist cost control case study show.