Quadient's CMO Petra Wolf doubled marketing productivity with agentic AI across the GTM engine. Then she published the part most marketing leaders leave out: "You've made the old operating model faster. You haven't changed it."
That line describes roughly 90% of ABM programs running AI right now. Faster research. Faster personalization. Same disconnected motion underneath.
The five-stage curve, applied to ABM
Wolf maps AI adoption across five stages: Assisted, Augmented, Orchestrated, Autonomous, Agent-native. The first two make people faster. Output goes up. The account motion doesn't change.
Most ABM teams are stuck between stages two and three. Marketing runs air cover against a wish list of sales accounts. Sales runs outbound. AI accelerates both in parallel, but they stay parallel. The only alignment is a shared spreadsheet of account names.
The economics shift at stage three, where two siloed programs collapse into one motion: account research, prioritization, buying-committee mapping, always-on touches, and sales outreach running under shared KPIs. AI handles the multi-step work between human decisions. Stage four means agents support the program end to end under guardrails. Stage five means the org chart follows the system — marketing and sales become a cross-functional team with a revenue target.
Why Quadient's context matters
Quadient is a €1bn+ company (€1,036M revenue in FY2025) in a hard strategic pivot. CEO Geoffrey Godet stated on the FY2025 results: "As we are ending 2025 at a record €250 million ARR level, we are raising our financial ambitions in Digital, which is now set to become our largest and most profitable Solution by 2030." Digital EBITDA margin hit 18.0%. Transactional mail volumes are forecast to decline through 2030.
Wolf joined from AWS in January 2024 and owns revenue and growth marketing. Her doubled-productivity number isn't the destination. It's the funding mechanism for the operating model change.
Three structural decisions from Quadient's ABM approach stand out.
Use-case clusters instead of verticals. Quadient's digital and financial automation use cases cut across banking, insurance, utilities, construction, healthcare, and SaaS. Cluster targeting lets you define unified qualification criteria around shared buyer triggers. Credit-to-cash, for example: high DSO, manual reconciliation, lean finance teams. The buying committee (CFO, VP Finance, Credit and Collections Manager, AR leads) is consistent regardless of industry. France's B2B e-invoicing mandate, effective 1 September 2026 for large enterprises, is a cluster with a built-in deadline.
Small-scoped pilot in a stronghold market. Quadient is number one in worldwide CCM, and France and Benelux are home turf. The pilot team: an ABM lead at 80% capacity, an SDR with at least eight months tenure at 60%, a content marketer at 40%, plus operational support from a subject-matter expert and RevOps to wire Salesforce to intent data, event signals, and content engagement.
Internal center of excellence before regional expansion. The pilot team becomes the framework owner. Without that, regional teams operate in silos with different GTM challenges and the program dies on arrival.
What the AI layer actually costs: Backbase's numbers
Backbase CMO Tim Rutten built the operating-model version of this curve. Backbase targets roughly 3,000 accounts worldwide. When they started, 60 to 70% of US wealth management targets were cold. A small pilot produced six discovery calls with enterprise accounts in eight weeks, proving the methodology and creating internal demand to expand.
The AI layer went on top of the proven process. Rutten built a signal engine in-house pulling marketing, third-party, LinkedIn, and sales signals in real time across every account. Monthly infrastructure cost: about $3,000 to $4,000. An agent reviews and prioritizes all 3,000 accounts continuously and flags the next action each week. When an account heats up, AI runs deep research and pushes buying-committee contacts into Salesforce. Around 80% of ABM content — account briefs, cluster webinar decks, newsletter drafts — starts as AI-generated raw material that a human refines.
What stays human: deciding which accounts get 1:1 attention, relationship building, and the executive layer. Rutten hosts a podcast with target-account executives and runs a community for women in banking.
The metrics that actually move
Wolf's curve is a cost story: CAC falls, cost-to-serve decreases. ABM at stage three and beyond moves the other side of the equation. Account-to-pipeline ratio, average deal size, and win rate. More named accounts convert to qualified pipeline because you're multi-threaded across the buying committee instead of single-threaded on one champion. Win rates climb because sales and marketing engage the same account with the same message.
Multi-threading a full buying committee across dozens of accounts while keeping research fresh was never affordable at scale. Human bandwidth was the bottleneck. That's exactly the work AI agents handle well, under human strategy and guardrails.
Rutten's CEO made the call: "How fast can all of my teams, all of my regions, all of our territories work this way?" Not a productivity report to the board, but an operating model the whole company reorganizes around. The AI just runs the plumbing.
