From AI ambition to readiness, action and continuous value.
October 2026
Part of Preparing for the Autonomous EnterpriseCrossfuze · ServiceNow Elite Partner
02
THE PROBLEM
We’ve invested in AI. Where are the results?
!CIO
Pilots never reach production
!Platform Owner
Unreliable data limits AI decisions
!AI / Knowledge Lead
Knowledge and agents lack clear owners
!CISO
Unapproved agents expand risk
!Finance / Procurement
Consumption and value remain uncertain
Five leaders. One question. What can you safely hand over?
81%
of executives believe more AI is running than they can account for.
ServiceNow with ThoughtLab, October 2026 · 550 executives
69%
of organisations suspect or have evidence of prohibited public GenAI use.
Gartner, November 2025 · 302 cybersecurity leaders
03
Your ServiceNow AI Readiness Gap
What’s stopping us from letting AI do the work?
Can we trust the data, rely on the workflow, control the action and justify the consumption?
Exhibit 01· The execution gap
ServiceNow AI is ready. The question is whether you are.
Crossfuze assesses readiness across the foundations of data, AI and agents, workflows and governance to identify what must change before autonomy can safely increase.
Three steps. The first two cost nothing. Explore with the tools, find your first outcome in 90 minutes, then deliver it with AI RunState.
01Self-serve toolsAbout ten minutes
ServiceNow AI ReadinessAre we ready?
ServiceNow Assist EstimatorWhat might it consume?
02Assess · 90 minutesClear next steps
The AI Readiness Assessment
Where you standWhat needs attentionWhat it may consumeWhere to start
90DAY CYCLES
03Design and executeOutcome-based delivery
Crossfuze AI RunState 12 MONTHS
Plan An AI Action Plan sets the direction
Prioritise Prioritization & Design picks what is next
Deliver Outcome Delivery ships it, in 90-day cycles
Clear next steps in 90 minutes. Outcomes every 90 days.
05
PREPARING FOR THE AUTONOMOUS ENTERPRISE
Autonomy is earned.
“
AI makes execution easier. Trust becomes the harder problem.
You prove the data can be trusted. You prove the workflow behaves as expected. You prove the controls work. Then you expand the boundary. Permission to automate is earned.”
Oliver NowakAI Practice LeadCrossfuze
Crossfuze helps you decide what AI is ready to do.
06
THE FOUR FOUNDATIONS
What does AIneed to stand on?
AI that can safely actBuilt on four strong foundations
Foundation 01AI and agents15%of the score
Foundation 02Data20%of the score
Foundation 03Workflows30%of the score
Foundation 04Governance25%of the score
AI can only act on what sits beneath it.
Your weakest foundation sets your autonomy ceiling.
Named people remain accountable.
Weights are each foundation’s share of the seven-dimension readiness score on page 08; dimension 06, platform and entitlement (10%), is assessed per Zero.
07
Your maturity
How ready are we?
ServiceNow’s Enterprise AI Maturity Index gives the direction. Our evidence determines where our AI readiness actually sits on the five-stage path to greater autonomy. Autonomy becomes practical at stage 03, and each stage is earned by lifting your weakest foundation.
The readiness journeyAutonomy becomes practical at stage 03
AI in isolated pockets. Data untrusted, work manual, and no one owns what AI does.
Pilots and assistants in places. Foundations patchy, so a person still checks every AI action.
Foundations connected across teams. Work redesigned so AI can start to act within limits.
Agentic workflows run end to end in production, with owners and controls.
AI predicts, decides and resolves by default. People set the limits and own outcomes.
Crossfuze readiness scale. Stages 02–05 align to the Experimenter, Scaler, Advancer and Transformer stages of the ServiceNow Enterprise AI Maturity Index 2026. The Five Zero outcomes and the Autonomous IT ambition are ServiceNow’s. The readiness scale is Crossfuze’s.
Every AI decision starts with an AI Readiness score. Each dimension is scored against evidence, weighted, and shown with its status, so nobody has to take the number on trust.
Seven weighted dimensions
01
Data qualityData Reliable fields, records and classifications
20%
02
Process determinismWorkflows Decisions, exceptions and rollback understood
15%
03
Knowledge and contentAI and agents Approved, current content to ground answers
15%
04
Workflow and integrationWorkflows The technical estate can run the work end to end
15%
05
Governance and securityGovernance Access, approval, monitoring and stop controls
15%
06
Platform and entitlementEstimator Package and Assist allocation fit the plan
10%
07
Ownership and adoptionGovernance Named owners, real users, capacity to measure
ProvisionalAt least 80%, including every 20% dimension
UnavailableOtherwise
Missing evidence is never scored as zero.
The score decides what happens next.
Not readyFix the limiting foundation first, then reassess.
ReadyDesign the roadmap for the chosen use case. Confirm the required controls before autonomous action.
Crossfuze AI Readiness Assessment method. Incomplete findings remain provisional and may need technical validation.
09
THE BILL
What will AI cost us as it scales?
The agentic cost paradox. The more AI does, the more it consumes. Readiness decides whether that consumption turns into value or into an unplanned bill.
THREE PACKAGES
AI is included at every level
FoundationAssistedSummarisation, categorisation and task-based help.
AdvancedAgenticAgentic workflows that understand context and automate a whole process.
PrimeAutonomousCustom AI specialists and agents that make decisions with role-based expertise.
Sources: ServiceNow New Commercial Model Overview, April 2026; ServiceNow assist-consumption guidance, 2026. Entitlements vary by product and package. Assists and Data Fabric credits are separate meters. Total consumption depends on run volume and mix.
10
The delivery problem
Traditional project plans don't work for AI.
Readiness tells you what you could do safely. It does not solve the delivery problem. Traditional projects lock decisions in early, make changes expensive and delay value until the end, exactly when the market may already have moved on.
Decided too earlyScope is locked before the platform and the market settle.
Expensive to changeEvery pivot needs a change order, with more cost and more time.
Value arrives lastThe payoff lands at go-live, if it is still relevant.
Illustrative. Not every long programme fails, but AI moves faster than a fixed plan can.
11
The 18-month trap
Why do most AI visions stay in the slide deck?
The problem is not a lack of vision. An 18-month fixed-scope project can lock priorities before the market, the platform and the business have finished changing. Each pivot becomes harder, while measurable value arrives late.
The board deckVision and outlook
Slow to commitWrong direction or too slow?Or both.
Dev teamWaiting
ProductionValue never arrives
A traditional AI programme
M0M3M6M9M12M15M18M21
Planning workshops
18-month projectChange orderChange orderGo-live
Frozen priorities
Month 1 decisions drive the outcomes in month 18.
Change orders
Anything new costs extra time and money.
Late ROI
Most of the payoff lands at the end, if it is still relevant.
Outdated on arrival
By go-live, the platform and the market have moved on.
So the vision waits. And the market keeps moving.
Illustrative timeline of a traditional fixed-scope AI programme.
12
Introducing CARS
The delivery methodology built for AI
Crossfuze AI RunState (CARS)
Built for how AI moves. CARS is a delivery service built for AI. It manages your ServiceNow AI Action Plan, prioritises what gets built and delivers it to production every 90 days.
Built for today's challenges
Where should we use AI?AI Action PlanUpdated every 6 months
What do we build next?Prioritization & Design SprintsQuarterly
How do we get it live?Outcome Delivery CyclesContinuous
90MINUTES
It starts from your free 90-minute assessment. Your readiness score, expected Assist usage and prioritised 12-month roadmap become the brief.
One team. One service. One flat monthly rate.
Board-ready AI Action PlanA strategy leadership can stand behind, in 4 weeks.
Pivot without change ordersPriorities reset every 90 days, at no extra cost.
Innovation every 90 daysSenior advisors and hybrid delivery, working as one team.
The free 90-minute assessment and the paid AI RunState service are separate. You decide whether to continue.
13
What you get
What do we get at every step?
Every step ends in something you can sign off. You make the decisions. We carry the work.
01Step 1 · 4 weeks
Plan.
Leadership decides the role of AI before anyone builds anything.
What you get
12-month incident and request clustering, turned into an AI opportunity and value map
End-to-end agent journey maps with a friction log
Executive AI Action Plan with future-state process and ServiceNow capability map
02Step 2 · Every 90 days
Prioritise.
Priorities updated every quarter, without new paperwork.
What you get
Approved three-month outcome set, capacity plan and signed-off backlog, handed to delivery
Outcomes scored on value, confidence and effort, with a parking-lot register and return points
Build-ready stories with technical feasibility, estimates and a sequenced dependency map
03Step 3 · 90-day cycles
Deliver.
Innovation into production every 90 days.
What you get
80 hours a month of hybrid delivery capacity, led by a dedicated onshore Technical Lead
Sprints, showbacks, peer review and testing to ServiceNow best practice
Release management, production verification and monthly status reporting
04Step 4 · Mid-year
Recalibrate.
Your strategy never goes stale. It is re-tested against what has shipped and what is new, and feeds straight into the next sprint.
Seven chances to pivot in 18 months. No change orders.
A sprint every 90 days, a mid-year recalibration and a fresh AI Action Plan at renewal, while delivery keeps running.
Indicative structure. Deliverables, hours and timings are confirmed in your statement of work.
14
THE FIRST 12 MONTHS
What does our first year look like?
Months 1 to 3 establish the AI Action Plan and approve the first build-ready priorities. Delivery cycles then bring approved changes into production every 90 days, with a mid-year strategy recalibration.
M1M2M3M4M5M6M7M8M9M10M11M12
AI Action Plan4 weeks, recalibrated mid-year
Prioritization & DesignRe-prioritised every 90 days
Outcome Delivery90-day cycles from month 4
Q1Q2Q3Q4
Active month
Recalibrate
Example · one illustrative ITSM year
Months 1 to 3Build an executable plan
Kickoff and instance access
4-week AI Action Plan: use cases ranked, baselines set
Sprint 1 approves Cycle 1 stories
Cycle 1 · M4 to 6Foundations and first wins
Top requests turned into one-click self-service
Tickets routed right the first time
Password reset automated
Now Assist for agents*
Cycle 2 · M7 to 9Ticket volume starts dropping
Knowledge rebuilt for top contact drivers
Virtual Agent resolving common requests
AI Search*
Predictive routing to cut reassignments
Cycle 3 · M10 to 12Scale AI across the desk
Now Assist extended to more agents*
AI agent pilot on a proven use case
Gains measured against baselines
One engagement. The effort moves to where the value is.
Illustrative ITSM year, not a commitment or scope of work. Nothing is built until you approve it. *Requires the right Now Assist licensing.
15
WHO DOES THE WORK, AND WHEN
How do we get started, and keep going?
One 12-month engagement. Plan in month 1, then prioritise and deliver every 90 days.
Oliver Nowakwith Crossfuze Advisory
01
Plan · Month 14 weeks
AI Action PlanWorkshops, use cases, roadmap
02Prioritise · every 90 daysDesign sprint
Prioritization & Design
Your team sets the priorities. Crossfuze turns them into stories and a three-month capacity plan, ready to build.
3CYCLES
12MONTHS
03Deliver · 90-day cycles80 hours a month
Outcome Delivery TECHNICAL LEAD
Build Enhancements and new capability
Run Administration, defects and support
Prove Hours tracked, reviewed before acceptance
Your teamOwns the priorities and signs off every cycle
Crossfuze AdvisoryOliver Nowak leads the AI Action Plan and each quarterly reset
Hybrid delivery podA dedicated onshore Technical Lead leads 80 hours of delivery a month
You own the priorities. We ship the outcomes.
Indicative structure. Scope, workshop numbers and hours are confirmed in your statement of work.
Autonomy is earned.
Start by knowing where you stand.
Book a free 90-minute AI Readiness Assessment. Leave with your readiness score, expected Assist usage and the first outcome worth prioritising. You decide what comes next.
