Practical guidance from implementation work.
The decisions behind successful CRM and AI projects: where to start, what to improve and how to prepare your team. If one of these saves you a meeting with us, it did its job.
Updated
Getting it live, and keeping it used
Migration, data and adoption — the delivery work that decides whether a system is used.
The CRM migration checklist we use
Moving a CRM is a decision about what to leave behind before it is a technical exercise. The list we run on every migration, in the order decisions get made.
Why CRM data goes bad, and the four fixes that hold
Bad CRM data is not a discipline problem. It is what a system produces when nobody owns a field, a list is a text box and there are four doors in. Four fixes.
CRM adoption is a design problem before it is a training problem
When a new system goes unused, the usual answer is more training. Usually it was designed for the report, not the person’s Tuesday. What we change instead.
Deciding what to automate
Which work an agent should own, and which should stay a person’s job.
How do digital employees differ from traditional software automation?
Automation executes steps you defined in advance. A digital employee is given an outcome and works out the steps — which changes what you can hand it.
What should an AI agent own?
Work that is continuous, cross-system and cheap to check. Four tests that separate the workflows worth an agent from the ones that cost you.
When should a workflow stay human?
Five signals that a workflow should not be given to an agent — and why an implementer who never says so is telling you something about themselves.
Assistive AI vs autonomous AI — which do you need?
A copilot makes your team faster; an autonomous agent owns an outcome. They are peers, not rungs on a ladder, and most firms need both.
What it costs, and what you get
The money questions — ROI, licensing and where the time actually goes.
How do AI coding agents speed up development cycles?
Coding agents that know a platform’s data model and UI framework produce work that lands natively, not snippets to stitch in by hand. Creatio reports up to 10x.
How do you calculate AI ROI?
Capacity returned, revenue unclaimed and cost taken out, against licences, AI actions and the build. The baseline is the whole exercise.
What does “unlimited CRM” mean?
Unlimited users, apps, workflows and API calls, priced per organization. AI actions are still metered — so forecast agent volume, not seats.
Where it runs, and whether you can trust it
Architecture and governance: permissions, placement and proof.
Do AI agents need their own permissions?
They need a role inside the permission model you already maintain — not a credential of their own. Anything else is a shadow access system.
When should AI live inside the CRM versus outside it?
If the work is CRM work it belongs in the platform, under its own approvals and audit trail. Outside is for what the platform cannot reach.
A straight answer, even when it doesn’t sell work
If an article raised a question it did not settle, bring it to a call. A principal consultant answers it.