Pricing AI Automation Services: A Value-Based Playbook
Nate Herk | AI Automationgo watch the original →
the gist
Stop billing hourly and start pricing based on the measurable business value—specifically 10-20% of the first-year annualized savings or revenue impact—to ensure profitability and client alignment.
The Case Against Hourly Billing
Nate Herkelman argues that hourly billing is inherently flawed for AI automation because it penalizes efficiency. When a developer gets faster, their income drops, creating a perverse incentive to work slowly. Hourly models also fail to capture the massive delta between the cost of development and the actual value delivered to a client's bottom line. He recommends moving to value-based pricing as soon as a consultant has a few successful projects to point to as proof.
The Value-Based Pricing Framework
Pricing should be anchored to the client's 'ceiling'—the total financial impact of the solution—rather than the developer's 'floor' (cost). Herkelman suggests a simple heuristic: identify the annual savings or revenue gain the system will generate, then price the project at 10% to 20% of that total.
- Discovery: Use the 'LRP' (Listen, Repeat, Poke) framework to uncover the business problem. Ask 'Why this, why now, and why me?' to surface hidden objections.
- Sizing: If a client refuses to share financial data, use proxy questions (e.g., 'How many employees touch this process?', 'What is the volume on your busiest day?') to estimate the scale of the problem.
- Proposal: Present three tiered options (Starter, Growth, Scale) to shift the client's decision from 'Should I hire this person?' to 'Which engagement level is right for us?'
Managing Scope and Payments
To avoid 'scope creep' and cash flow issues, structure projects into 30-day milestones. These milestones must be strictly objective (e.g., 'The agent pulls from the database and responds within 60 seconds') to prevent subjective arguments over whether a deliverable is 'done.' When clients request new features mid-project, acknowledge the value but move them to a 'Version 2' backlog to keep the current project on track. Maintenance retainers should be kept simple and standardized; they cover keeping the system running as scoped, not building new features.
Measuring Success
Crucially, consultants must capture a 'baseline' before implementation and measure the 'after state' to prove ROI. If the consultant does not surface these numbers, the client may forget the value provided, making it harder to win future business or justify higher fees.