Enterprise Ai

AI and Salesforce transformation — Where to start

By David Stott, MBA · July 27, 2026 · 6 min read

Leaders routinely rush to platform choices—Salesforce editions, third-party apps, or generative AI—before they understand the operating friction that prevents value. Start by making friction visible: name the business outcome, inventory recurring delays and rework, and assign single accountable decisions. Use a short evidence-driven readiness path (friction inventory → root diagnosis → sequence decisions → prepare or pause) so platform selection follows the operating fixes it must enable. This reduces waste, shortens decision cycles, and increases the probability that technology amplifies human judgment rather than accelerating broken processes.

Why this matters

Enterprise leaders face two recurring failures when pursuing Salesforce modernization or AI: (1) choosing technology to solve symptoms rather than root causes, and (2) scaling AI before readiness (ownership, data, controls, or redesigned work) exists. Both produce higher cost, escalating technical debt, frustrated users, and weak measurable outcomes.

Start with Business Before Technology™

Begin with a clear business outcome and the explicit question: Does this change reduce business friction and create measurable business value? Use the Business Before Technology principle and the Business Friction lens to keep decisions grounded on operating evidence—not vendor demos.

Recognize common friction signals (what to look for)

• Repeated escalations or slow executive decisions (decision friction).
• Work that loops, duplicates, or depends on fragile handoffs (process friction).
• Important information that is late, fragmented, or reconstructed at the point of decision (data friction).
• Platforms or integrations that create rigidity, frequent outages, or opaque ownership (technology friction).
• Teams preserving shadow processes because official workflows don’t produce results (trust and adoption friction).

A practical five-step diagnostic playbook (90 minutes to first evidence)

1. Anchor on one measurable outcome (15 minutes)

• State the business result you need in the next 3–12 months, the stakeholders it affects, and the evidence that currently shows value is lost.

1. Capture the friction inventory (30–60 minutes)

• Collect recurring pain signals across the dimensions in the Executive Friction Report (decision, process, data, technology, organizational, customer-value, trust/adoption).
• Record specific examples (a delayed approval, a repeated rework loop, an inaccessible dataset) without jumping to solutions.

1. Prioritize by impact and urgency (15 minutes)

• Score items by their immediate threat to the outcome and how frequently they recur. Prioritize items that block decisions or create customer harm.

1. Diagnose root causes (30–90 minutes per priority)

• Test hypotheses: is the fault unclear ownership, missing data, unnecessarily complex process, or brittle integration? Ask: who decides, what evidence informs the decision, and what changes would demonstrate reduced friction?

1. Assign accountable decisions and 30/60/90-day actions (15 minutes)

• Name one decision owner per priority, the decision required (not the solution), and the smallest responsible action that will show observable improvement.

When to consider Salesforce changes or AI

• Prepare the operating model first. Use Strategy and Systems Alignment and Enterprise Readiness Assessment to test whether the platform or AI will plug into a capable operating model.
• Consider platform changes when: clear capability gaps exist that cannot be addressed by process redesign and the product stewardship and integration obligations are owned.
• Consider AI when: a defined friction condition exists where AI can reduce repetitive work, improve decision evidence, or surface opportunities; accountable owners, data readiness, governance, and work redesign are demonstrably available (see Enterprise AI Maturity Model and AI Governance Framework).

Sequence: Redesign before automation

Follow the Redesign Before Automation approach: simplify the work, remove waste, clarify ownership, and only then automate or embed AI. Automation accelerates what you build—if the underlying process is broken, automation amplifies the problem.

Minimum launch controls for AI-informed workflows

• Named human accountability for each material decision informed by AI.
• Defined evidence and grounding for the model’s outputs at the point of decision.
• Clear action boundaries (what the system may recommend vs. execute).
• Monitoring and escalation paths; stop-or-adjust triggers tied to outcome measures.

A 30/60/90-day action plan template (example sequence)

30 days (stabilize)

• Complete friction inventory for the anchored outcome.
• Assign accountable decision owners for top 3 friction items.
• Run a light Enterprise Readiness Assessment on the highest-priority use case.

60 days (diagnose & redesign)

• Validate root causes with evidence (logs, handoffs, timelines, sample data).
• Redesign one critical process to remove waste and clarify handoffs.
• Define data fixes required at the point of decision (ownership, source, refresh cadence).

90 days (prepare & pilot)

• If readiness is green on critical dimensions, pilot a bounded platform change or AI assistant with controls and measurement.
• Collect outcome evidence (decision cycle time, rework incidents, customer/employee effort) and review against stop-or-adjust criteria.

Measures that matter (not activity)

• Decision cycle time for material approvals or customer decisions.
• Frequency of repeated handoffs or rework loops.
• Time-to-trust for a dataset at the point of decision (availability + acceptance by users).
• Customer or employee effort scores tied to the affected journey.
• Movement in the anchored business outcome defined at the start.

Avoid these common failure patterns

• Selecting technology to signal progress rather than to remove the material constraint.
• Averaging maturity by dimension and hiding a critical red condition.
• Creating committees or governance that collect status but do not resolve accountable decisions.
• Measuring activity (deployments, models trained) instead of capability and outcome.

Decision guide (short)

• If a friction item is a clear process or ownership gap → fix process and name owners before buying or building.
• If data is missing or disputed at the point of decision → assign data ownership and remediation; delay automation until data is trustworthy.
• If architecture or integration gaps prevent reliable context flow → use Architecture Decision Framework to select an option with explicit operating consequences and reversibility.
• If AI is proposed to reduce repetitive judgment and data is fit for purpose → apply AI Governance Framework controls and pilot under direct business ownership.

Closing recommendation

Do not treat Salesforce or AI as the answer to unclear decisions, poor data, or fragmented ownership. Start by diagnosing and reducing the friction that prevents execution. Sequence the work: clarify the outcome, inventory and prioritize friction, assign accountable decisions, redesign work, assess readiness, then select technology. That sequence increases likelihood of measurable value and reduces the risk of accelerating complexity or distrust.

David’s Perspective

Transformation succeeds when leaders focus first on the decisions that must change, not the technology they hope will make those decisions easier. Naming accountable decision rights, making data trustworthy at the point of decision, and simplifying work are investments that increase the return from any Salesforce modernization or AI initiative. When teams skip that order—design, own, measure—the technology simply accelerates the existing friction.