I’m at AMS HQ, stood next to a big hole in the ground. On one side: new saplings. On the other: fresh landscaping. It’s a simple picture that explains how to implement a business database the right way.

Think of your system as a plot of land. You don’t build the entire garden in a day; you plan it, start with one area, and grow into it.

Here are the takeaways:

Start small and plan phases
Don’t try to do everything in phase one. Pick a solid platform like Microsoft Dynamics 365, launch the essentials, and add more in phases two and three. Doing too much too soon increases risk and can sink a project.
Put change management in place from day one
Decide who’s involved, how you’ll communicate, and how often. Set up a regular rhythm (daily or weekly) to share progress, collect feedback, and agree changes. Keep the right people informed and engaged.
Prototype early, refine often
Before you “fill the pond with water”, build a prototype. See it at the right scale, test it with real users, and tweak it before you commit. Prototypes save time, money, and stress.
Keep phase one lean
Focus on the critical outcomes. Get the core team on board, prove value quickly, then iterate. When people can see it working, they’ll help shape what comes next.
Iterate with feedback loops
As the system rolls out, keep listening. Make small, frequent improvements. The goal is a better product over time, with users who feel heard and invested in the result.

That’s the heart of a successful implementation: start focused, communicate clearly, prototype early, and improve as you go.

More ideas

Phase planning made simple
- Define phase one as your “minimum lovable product” only what users need to succeed on day one.
- Map phase two/three in advance so you don’t lose momentum.
Change champions and comms
- Pick a few respected users as champions to test prototypes and spread good practice.
- Keep updates short and regular: what’s live, what’s next, what feedback changed.
Data readiness
- Cleanse duplicates, agree naming and picklist standards, and decide who owns data quality.
- Start small: migrate just what you need for go‑live, archive the rest.
Success measures to track
- Adoption: active users per week, records created, data completeness.
- Outcome: time to create a quote/case, lead response time, first‑contact resolution.
- Quality: error rates, rework, duplicate records.
Go‑live approach
- Pilot with one team, fix issues fast, then roll out broadly.
- Plan “hypercare” for the first two weeks with clear support routes.
Governance and roles
- Agree who approves changes, who manages the backlog, and how often you release.
- Keep a single, prioritised list of improvements to avoid scope creep.
Integrations and “system of record”
- Decide which system owns customers, products, and invoices.
- Build in error handling and simple monitoring so failures are visible and fixable.
Security and compliance basics
- Use role‑based access, enable auditing where needed, and set sensible data retention rules.
- Document who can see/export sensitive fields (e.g., pricing, personal data).
Performance and usability
- Keep forms light, hide fields users don’t need, and test search performance.
- If it feels slow, it won’t get used, optimise early.
Environments and releases (Power Platform/Dynamics)
- Use Dev → Test → Prod environments, manage changes in Solutions, and release on a schedule.
- Keep a short demo video or screenshot in each release note to show what changed.