Most writing on digital transformation describes the destination: the platform, the dashboard, the customer app. It is the easy part to describe and usually the easy part to buy. What decides whether a programme lands is less visible — how work actually moves through the business today, what the data really contains, and when security gets decided.
Start with what better means.
Before any platform is chosen, agree what better means in your business, in terms someone outside IT would recognise: fewer hand-offs, a month-end close that finishes in days rather than weeks, one version of the customer instead of four. Without that, nobody can say whether the programme worked, and the business case quietly turns into a licence comparison.
Describe the starting point honestly.
The target state is usually documented in detail. The current state rarely is. Somewhere in most organisations, work leaves a system and becomes a spreadsheet, an email or a specific person who reconciles two sets of numbers by hand. Those joins are where transformation programmes stall, because they belong to nobody and they are never in the plan.
Map them before you design anything.
Decide security at design, not at the end.
“Secure” should be a claim about sequence. Identity, access, segmentation and data protection are cheap to decide while the target is being designed and expensive to retrofit after it is built. If a plan shows security as a phase near go-live, it has been scheduled, not built in.
Prove it on real data before anyone commits to a date.
A dry run against real data — not a sample chosen to pass — is the cheapest insurance a programme can buy. It is where the undocumented fields, duplicate records and forgotten integrations surface: before the cutover weekend rather than during it.
Plan the exit as carefully as the entry.
Transformation fails slowly as well as suddenly. If ownership, documentation and decision rights are not settled at handover, the new estate starts to accumulate exactly as the old one did. Name who owns each system, who approves change and which standards apply, and write it down.
Where we are direct.
We will not fixed-price a scope nobody can yet describe. Where the estate is not understood, the discovery comes first and is priced on its own. A lift-and-shift is not a transformation. Moving the same process to a new platform moves the same problems with it. Sometimes the answer is to retire a system rather than replace it.
That advice is worth paying for too. If the target of your programme is clearer than its starting point, that is the conversation to have first. Speak with Toga — twenty minutes on what you are actually trying to do.