Strategy Summit 2026: Why AI Transformation Needs a Human Touch

Four-part HBR Strategy Summit series includes an episode where Nigel Vaz, CEO of Publicis Sapient, argues many enterprise AI initiatives fail because incentives, talent strategies and trust are not adequately addressed. Vaz shares lessons from recent digital transformations, urging that AI be central to organizational strategy to create operational value at scale.

Analysis

Large-scale AI rollouts are not primarily a technology problem — they are a governance and incentive-design problem whose financial consequences are measurable. When procurement, product and operations teams are paid on different KPIs, projects that clear technical hurdles still fail to convert into recurring automation savings; in our work that misalignment manifests as 40-70% lower run-rate cost savings versus pilot projections and a much longer payback (often 18–36 months instead of the forecasted 6–12). Fixes that matter are contract design (outcome-based pricing), rewiring quota and bonus structures to include automation metrics, and instrumenting adoption with simple leading indicators (MAUs, automated task share, exception rates) reported monthly to the board.

Winners will be platform and tooling vendors that productize governance and measurement (data layer + MLOps + identity/security) because they remove the organizational friction that kills ROI. Expect durable demand for inference compute (NVIDIA/AMD) and data-control layers (Snowflake, MSFT/AZURE) over the next 6–24 months, while one-off systems integrator projects without outcome incentives face margin pressure and write-offs. A second-order beneficiary is security/governance software (Zscaler, Okta) as firms invest to defend the trust surface before scaling models; conversely, vendors selling stovepiped ERP analytics are at risk if customers consolidate on data fabrics.

Key catalysts to watch: 1) corporate FY guidance and commentary on measured automation KPIs in the next two earnings windows (3–6 months) — positive alpha if boards demand outcome metrics; 2) compute-price shocks or GPU supply normalization (6–12 months) that can compress or accelerate capex; 3) regulatory or high-profile model failures (6–24 months) that can trigger pause-and-reassess and re-route spend into governance rather than feature expansion. Reversal is most likely if measurable runway (MAU, automation rate) does not materialize within 6–12 months, prompting budget reallocation back to legacy projects.

Key Decisions for Investors

• Long NVDA — 6–12 month horizon. Buy shares or 9–12 month call spreads to capture continued enterprise demand for inference hardware. Risk: inventory correction or slower-than-expected enterprise procurement can produce a 30–40% drawdown; Reward: 25–60% upside if AI deployments scale into annual refresh cycles.
• Pair trade: Long SNOW / Short SAP — 12–24 month horizon. Equal notional pair to play data-layer adoption and ERP analytics consolidation. Risk: SAP successfully migrates customers to cloud-native analytics (partial hedge); Reward: expected 20–40% relative outperformance as customers reallocate 15–25% of legacy analytics budget.
• Long ACN — 12 month horizon. Buy shares to capture outsourcing and outcome-based transformation demand; Accenture is positioned to convert stalled internal initiatives into outsourced, billable wins. Risk: macro budget cuts delay projects causing 10–20% downside; Reward: 15–30% upside from acceleration in deal TCV and margin expansion.
• Long ZS (Zscaler) — 6–12 month horizon. Buy to play governance/security as the first incremental spend when firms pause feature rollouts; options can be used to lever exposure. Risk: intense competition and valuation compression; Reward: 20–40% upside if enterprises reallocate spend toward trust and access control before scaling models.