Stop Buying AI Efficiency Discounts: Why Marketers Should Pay for Deliverables Instead

It is budget planning season for 2027, and marketing procurement teams have a shiny new weapon handed to them directly by agency pitch decks: the promise of AI-driven efficiency. Every major agency holding company, from WPP Open and Omnicom Omni to Publicis Marcel and Dentsu.Connect, now centers its pitch on a proprietary, "end-to-end, agentic AI" platform.

Procurement heard those pitches, took them at their word, and is now using those exact claims to justify slashing agency retainers across the board. But the entire conversation is built on a misunderstanding of productivity data, according to Maarten Albarda, founder of Flock Associates, writing in MediaPost on October 2, 2026.

The Productivity Gap: Headlines vs Reality

When marketers and procurement leads push for fee reductions, they often cite McKinsey's widely quoted estimate that generative AI could automate work consuming 60% to 70% of employee time. But nobody spends 100% of their workday on tasks that can be handed off to AI.

Empirical studies tell a more nuanced story:

With these numbers in mind, procurement might go in with a 30% reduction based on agency promises. The agency accepts the cut, reduces headcount for that client, and soon both sides realize that AI cannot write strategic briefs, navigate internal client politics, or manage complex stakeholder alignment. Operational quality collapses, the relationship sours, and both sides end up frustrated.

The Real Problem: Pricing the Wrong Thing

The core issue is that the negotiation focuses on theoretical efficiency rather than operational scope. Financial savings should be the natural result of fixing processes and scope, not the starting point.

Albarda's recommendation is to shift from asking for blanket AI efficiency discounts to requiring an operational audit that breaks work into three buckets:

1. Strategic Work

This includes brand strategy, campaign development, creative briefs, market analysis, and stakeholder management. AI can assist with research and ideation but cannot lead this work. The budget here should be protected, not cut.

2. Repetitive Implementation Work

This includes media buying setup, report generation, A/B test configuration, creative resizing, and routine copy variations. AI genuinely saves time here, and you should see less cost in this bucket. This is where procurement's 30% discount request is actually justified.

3. Complex Tasks Requiring Heavy Human Touch

This includes crisis management, regulatory compliance reviews, complex negotiations, and high-stakes creative decisions. AI can inform these tasks but should not execute them autonomously. The budget here should be maintained or increased.

Why Time-Based Pricing Fails in an AI World

Attempting to measure AI efficiency through agency timesheets is not going to give you the clarity you need. When an agency uses AI to write a campaign brief in two hours instead of eight, the timesheet shows two hours. But the deliverable's value has not decreased. In fact, the agency's investment in AI tools, training, and infrastructure means their cost per deliverable may have increased even as their hours decreased.

The fundamental problem is that time-based pricing rewards inefficiency in a pre-AI world (more hours = more revenue) and punishes investment in AI in a post-AI world (fewer hours = less revenue, despite higher tool costs).

The Deliverable-Based Pricing Alternative

The solution is to migrate agency contracts toward deliverable-based pricing or fixed retainers tied to specific business outcomes. This means:

• Define deliverables, not hours. Specify what outputs you expect: number of campaigns, creative assets, media plans, reports, and strategy documents.
• Tie payment to completion, not time. Pay when the deliverable meets an agreed quality standard, not when the agency logs enough hours.
• Let the agency own AI adoption. Paying for deliverables forces the agency to adopt AI to protect its own margins. They capture the efficiency gain, not the client, which gives them incentive to invest in better tools.
• Pay for value, not activity. The client pays for the value of the output rather than the hours spent sitting at a desk or churning tokens.

The Agency Platform Landscape: What Procurement Heard

To understand why procurement is asking for discounts, it helps to see what agencies are pitching. Every major holding company now markets an AI platform:

Beyond the holding companies, platform-level tools are also driving the efficiency narrative:

• Meta Advantage+ for automated ad creative and targeting
• Google Performance Max (PMax) and AI Max for automated media buying
• Amazon Marketing Cloud for data-driven advertising
• Salesforce Agentforce for marketing automation
• Microsoft Copilot Studio for custom AI workflows

The common thread: every tool promises faster, cheaper, and more automated marketing. Procurement heard the pitch and is now using it as leverage.

A Practical Framework for Marketing Teams

If you are a marketing leader planning your 2027 budget, here is a practical framework for restructuring your agency relationships:

Step 1: Conduct an Operational Audit Break your current agency scope into the three buckets: strategic, implementation, and complex. Map current spend, hours, and deliverables for each bucket.

Step 2: Renegotiate Implementation Work For the implementation bucket, ask the agency to show their AI-assisted workflow. This is where efficiency gains are real. Negotiate lower costs here, but do not expect the full 30%. The BCG/Harvard data suggests 25% is a more realistic target for tasks AI can handle.

Step 3: Protect Strategic Budget For the strategic bucket, maintain or increase budget. AI does not replace strategic judgment. If anything, the increasing complexity of AI-mediated marketing requires more strategic oversight, not less.

Step 4: Move to Deliverable-Based Pricing For each bucket, define specific deliverables with quality standards. Move away from hourly billing toward fixed-fee or deliverable-based contracts. This gives the agency incentive to use AI efficiently while ensuring you pay for what you need, not for time spent.

Step 5: Build AI Adoption Into the Contract Include requirements for the agency to report on AI tool adoption and its impact on deliverable quality and speed. This creates transparency and gives you data to inform future negotiations.

What Not to Do

• Do not ask for a blanket AI efficiency discount. It sounds good in a negotiation but destroys the relationship when the agency cuts the wrong resources.
• Do not trust timesheets. They measure activity, not value. In an AI world, the most efficient agency logs the fewest hours.
• Do not assume all work is automatable. The 60-70% McKinsey figure refers to tasks, not jobs. No one's entire role is automatable.
• Do not forget that AI tools cost money. Agencies investing in AI platforms need to recover those costs. If you squeeze fees too hard, they stop investing, and quality drops.

The Broader Context: Marketing AI in 2026

This pricing debate sits within a larger transformation of marketing technology. The Global AI MarTech market is forecast to grow from $28 billion in 2025 to $74.3 billion by 2031, driven by agentic AI, platform consolidation, and privacy-first personalization. Kantar's Media Reactions 2026 study found marketers are betting on AI but only 24% trust it for creative decisions, a gap that supports the argument for protecting strategic budgets.

The companies that will win the 2027 budget cycle are not the ones that squeeze the hardest. They are the ones that restructure their contracts to align incentives: pay for deliverables, let agencies capture efficiency gains, and protect the strategic work that AI cannot do.

If your marketing team is planning 2027 budgets and wants help restructuring agency contracts for an AI-powered world, Vivid can help you design a deliverable-based pricing framework. We build marketing operations that capture AI's real productivity gains without destroying the strategic work that drives growth.

Read Next

• Argo-Bench: Enterprise AI Agents Clear Just 34.8% of Real Operational Tasks
• AWS Strands Decider 2B: The Decision Model That Never Writes Text
• Shopify Canvas and Meta Muse: The Agentic Commerce Alliance
• Kantar Media Reactions 2026: Marketers Bet on AI, But Only 24% Trust It