AI may be the mother of reinvention at Accenture
Saturday, 6 December 2025
Accenture has rebranded again. Strategy, consulting, creative, and tech now sit inside a single unit called "reinvention services," and consultants are being labeled "reinventors." The rollout drew smirks, but the message is clear: AI is no side project. It's the core offer.
This isn't the firm's first linguistic leap. The Accenture name itself-born from a staff contest in 2001 and mocked at launch-outlived the jokes. Compare that to PwC Consulting's short-lived attempt to rename itself "Monday." Sometimes audacity pays off.
The real question isn't branding. It's whether clients will still hire big consultancies to figure out AI-or just use AI to do the work themselves. That challenge goes straight at the profit engine of the industry, especially the entry-level work that AI already handles well: research, slides, synthesis, and first-draft analysis.
What this means for managers
- Expect fewer junior-heavy teams. The traditional "pyramid" (many analysts, few partners) gets flatter as AI automates baseline tasks. Push vendors to bring senior, outcome-focused talent and fewer bodies.
- DIY vs. consultants vs. a hybrid. Many use cases are now feasible in-house with strong data, a small platform team, and AI copilots. Use consultants for complex integration, governance, and change-where institutional experience still matters.
- Outcomes, not decks. Tie work to measurable business targets: cycle time reduced, cost per ticket lowered, lead-to-close improved, model accuracy increased within acceptable risk limits.
How to evaluate proposals now
- Scope: Insist on working prototypes, not just "recommendations." Include data pipelines, prompts/agents, and handover plans.
- Time-to-value: 4-8 week sprints with executive demos. No multi-month discovery phases without a build.
- Pricing: Blend fixed fees for pilots with gainshare or KPI-linked fees for scale. Cap day-rate work.
- Models and IP: Clarify model choices (open vs. closed), fine-tuning strategy, prompt assets, telemetry, and who owns what.
- Security and compliance: Data boundaries, red-teaming, audit trails, PII handling, and fallback procedures. Require documented model risk controls.
- Change and adoption: Training plans, role redesign, frontline enablement, and metrics for sustained usage.
The talent shift you should plan for
Entry-level consulting work is getting automated. That hits the training ground most firms rely on. Expect vendors to bring smaller teams and more automation. Internally, redesign roles so managers and ICs use AI directly-no more waiting on a slide factory.
Build an internal "AI enablement" spine: a lean platform team, data owners in each function, and a playbook for safe experimentation. Without this, you'll stay dependent on external help and pay more for basics.
Three likely buying models
- DIY with targeted support: You staff a small in-house team and bring in experts only for audits, security, and tricky integrations.
- Hybrid pods: Joint teams deliver sprints with shared tools and shared accountability for KPIs.
- Full-service programs: Useful for heavy legacy estates or regulated environments-but demand clear milestones and exit criteria.
Market signal: the stock slide
While AI platform players soared this year, Accenture's share price fell from $389 to $269, cutting its market cap by more than 30% to about $167bn. Investors are asking if the old delivery model still fits the new demand. If this continues, the industry will shrink headcount and automate harder-which will change what you actually buy from them.
Practical next steps for Q1
- Select 3-5 use cases with short payback: customer support deflection, sales assist, invoice matching, policy Q&A, knowledge retrieval.
- Run one 6-8 week pilot per function. Require live demos and real users, not demo-only environments.
- Stand up guardrails: access controls, prompt logging, content filters, human-in-the-loop for sensitive actions.
- Retool vendor panels: mandate outcome pricing options, model risk documentation, and handover of prompts/agents as deliverables.
- Upskill your managers so they can evaluate solutions, not just presentations. Consider focused learning paths by role.
Context and further reading
The "reinventor" move fits a long-running pattern of big-bet branding and service realignment. For background on critiques of the consulting model, see The Big Con. For official positioning, check the Accenture newsroom.
If you need structured upskilling
To get your team fluent enough to judge vendor claims and run small in-house pilots, explore role-based paths at Complete AI Training.
Bottom line: AI won't kill consulting, but it will rewrite who you hire, what you pay for, and how fast you expect results. Set your standards now, or you'll keep buying decks while competitors ship working systems.
Your membership also unlocks:
