1
Demand
Is there enough demand to justify a build, CRM layer, or automation?
Process
Diagnose, position, build, measure, improve.
The process is designed to avoid vague website projects. Every stage makes the next commercial decision clearer and keeps the first useful step visible.
Strategy compass
Build only what improves the next decision.
The point is not to add more pieces. It is to find the piece that reduces confusion, leakage, or repetitive work without adding avoidable risk.
Decision
2
Data
Are fields, sources, and ownership reliable enough to make decisions?
3
Risk
What should stay manual until the process is trustworthy?
4
Leverage
Which change makes the next commercial decision easier?
Step 1
Diagnose
Clarify the offer, audience, current assets, constraints, and what must improve first.
Output: Fit call notes, priority map, and scope recommendation.
Decision gate: Continue, reduce scope, or pause.
Step 2
Position
Turn the business case into page hierarchy, CTA logic, evidence requirements, and message direction.
Output: Offer structure, page map, and conversion path.
Decision gate: Message and page direction approved.
Step 3
Build
Create the site, landing paths, handoffs, and practical launch materials with a tight scope.
Output: Working pages, handoff notes, and launch checklist.
Decision gate: Launch checklist and handoff are clear.
Step 4
Measure
Define the events and reporting view that help you decide what to improve next.
Output: Decision-ready measurement plan, not vanity dashboards.
Decision gate: Events and reports answer a real decision.
Step 5
Improve
Use the first signals to prioritize conversion repair, content, automation, or advisory cadence.
Output: Backlog, next tests, and optional evolution hours.
Decision gate: Next sprint, advisory cadence, or stop.
When the issue is bigger than the page
The deeper process starts with control, not a giant build.
For AI Workflow / Cost-Leak Audit, customer intake, dashboards, product systems, or custom conversion paths, the first job is not to sell a giant build. It is to demonstrate the workflow, source material, owner, approval step, and follow-up rhythm.
Boundary
This is not sold as autonomous support, guaranteed savings, a CRM migration, a product rewrite, or a lead-volume guarantee. Those become custom scope only after fit is clear.
1. Fit gate
We check whether one workflow can be isolated, whether an owner exists, and whether the problem is real enough to justify system work.
2. Client inputs
We collect channels, example requests, source material, escalation rules, tool constraints, and privacy boundaries before building.
3. Build the operating layer
We map intake, triage, approved source material, human-reviewed drafts, dashboard fields, and follow-up dates into one visible workflow.
4. Launch with boundaries
No autonomous sending, no regulated advice, no guaranteed lead volume. The first version proves the workflow before deeper automation.
5. Improve from signals
The first weekly note shows repeated questions, missed owners, overdue follow-ups, and the next useful improvement.
Operating principle
Scope follows risk.
Start with Base Digital Refresh when the problem is trust, clarity, mobile, or contact. Move to Enhanced Digital Refresh when the page needs stronger hierarchy, evidence, polish, or one low-risk integration. Use AI Workflow / Cost-Leak Audit first when the problem is repeated work, tool handoff, or AI adoption risk.
Ecommerce, payment, localization, Google Ads, and custom conversion paths enter only when the brief shows that extra depth is justified.