eazyware
Playbook·November 13, 2023·10 min read

Reporting AI progress to the board

Metrics, narratives, risk framing for board meetings on AI strategy. What directors want to hear and what signals credibility.

KR
Kushal R.
Engineering lead

Reporting AI progress to the board is part craft, part science. Directors want business outcomes, competitive positioning, and risk awareness — not technical architecture deep-dives. This post is the structure, metrics, and narrative framing that work in AI board meetings, and the common mistakes that erode credibility.

What resonates
Board reporting on AI — what resonates Common mistakes Too much model/tech detail Feature list without outcomes Hype-heavy framing Ignoring risks/gaps What works Business metrics (revenue, NPS, cost) Competitive positioning Risk register and mitigations Capability trajectory, not specs Core board slide structure 1) Business results this quarter 2) Key bets and where they stand 3) Competitive landscape 4) Risks to watch 5) Asks of the board
Common mistakes: too much tech, feature lists, hype-heavy framing, ignoring risks. What works: business metrics, competitive positioning, risk register, capability trajectory.

What directors actually want

Business results. Revenue growth, margin, customer acquisition, retention. AI strategy in service of these, not abstract technical achievements.

Competitive positioning. How do we compare to rivals? Where are we ahead? Where are we behind? Credible assessment, not marketing spin.

Risk visibility. Directors have fiduciary duties. Unknown risks create liability. Known risks can be managed.

Capital deployment efficiency. Are we investing in the right things? Getting returns? ROI on AI R&D.

Strategic alignment. Is AI work aligned with company strategy, or distraction?

Core slide structure

Business results this quarter. Core KPIs (revenue, growth, retention) with YoY and plan comparison.

Key bets and where they stand. 3-5 major AI initiatives; progress, blockers, decisions needed.

Competitive landscape. Who's winning where? How are we positioned?

Risks to watch. Regulatory, competitive, technical, financial. What might go wrong and what we're doing about it.

Asks of the board. Where do we need their help? Introductions, budget approval, strategic direction.

Metrics that matter to boards

Customer metrics. Number of customers, net revenue retention, expansion rate, logo growth. Tangible proof of value.

Revenue composition. What % of revenue is AI-driven? Growing or not? New product mix indicates momentum.

Unit economics. Gross margin, CAC, LTV. AI features often have high variable costs; unit economics scrutiny matters.

Retention and usage. Stickiness matters enormously. Daily/weekly active users, usage depth over time.

Competitive wins/losses. Deals won vs competitors; deals lost. Why we won or lost.

What not to show

Model training details. Directors don't need MLflow dashboards. Quality metrics summarized, not raw eval outputs.

Feature lists without outcomes. Shipped X, Y, Z features doesn't tell the board anything. Did they drive revenue? Retention? Market share?

Every experiment and pilot. Meaningful bets only. If every small pilot is reported at board level, signal-to-noise drops.

Hype language. 'Revolutionary,' 'transformative,' 'paradigm shift' — all words that drain credibility. Be measured.

Risk framing

Regulatory. AI Act in EU, state legislation in US, sector-specific rules. What's our exposure? Compliance readiness?

Competitive. New entrants, capability parity with leaders, category shifts. Candid assessment.

Technical. Model dependencies, provider concentration, data dependencies. Resilience posture.

Safety and reputation. Could an AI failure cause meaningful harm or reputational damage? Incident response capabilities?

Financial. Cost trajectory; margin compression from competition; capital requirements for next growth phase.

Cadence and rhythm

Quarterly board meetings. Full AI report typically quarterly, aligned with overall board cadence.

Monthly or intra-quarter updates. For fast-moving AI initiatives, short written updates between meetings help directors stay current.

Pre-read materials. Send materials 48-72 hours before meeting. Allows directors to prepare questions; meeting time spent on discussion, not presentation.

Executive sessions. Occasionally in closed session for sensitive topics (acquisitions, personnel, regulatory matters).

Tailoring to director types

VC directors. Want growth rates, competitive positioning, capital efficiency, exit scenarios.

Independent directors. Often bring operating or industry expertise. Relate to functional areas (go-to-market, operations, finance).

Strategic directors. Company-sponsored seats. Interested in relationship implications, joint work.

Founder/exec directors. Already know much; need strategic framing, not status updates.

Building credibility over time

Consistency. Same metrics every quarter; trend visible. Changing metrics every meeting erodes trust.

Accuracy. When you miss projections, acknowledge and explain. Cover-ups destroy credibility fast.

Preparation. Knowing your numbers cold. Inability to answer clarifying questions signals sloppy management.

Strategic thinking. Show you're thinking beyond current quarter. Board members want leaders, not just operators.

Read next
Getting executive alignment on AI strategy
Read next
Planning AI roadmaps: model progress, capacity, dependencies
Read next
AI product management: the craft in 2026
Tags
boardreportingexecutive
/ Next step

Want to talk about this?

We love debating this stuff. 30-minute call, no pitch, just engineering conversation.

~4h
avg response
Q2 '26
next slot
100%
NDA on request