DataGovernance.ca working methods · Original planning guidance. Examples are illustrative, not client results or a regulatory framework.
Choose a decision, not a transformation slogan.
A useful first brief identifies a decision that is slower, less reliable or harder to explain because of a data problem. For example: the monthly customer report requires repeated reconciliation before anyone will use it. Define the report, the audience, the period and the consequence. “Improve our data” is too broad to approve or finish. Start with one domain and record what the pilot will not attempt.
Establish a small, honest baseline.
Agree a short observation window. Ask the people doing the work to record reconciliation time, the reason for each correction and what evidence was available. Distinguish a one-off migration task from recurring rework. Do not extrapolate an unusually difficult week across a whole year without saying so. Record missing observations as missing; do not replace them with a convenient zero. The baseline is a working estimate, not an audited cost account.
Separate capacity from cash.
Consider an illustrative team of five people spending two hours per week on reconciliation, at an assumed loaded labour rate of C$50 per hour for 48 working weeks. The annual capacity value is C$24,000. A hypothetical 25% reduction represents C$6,000 of potentially released capacity, not C$6,000 automatically removed from payroll. The time might instead support faster reporting, more analysis or a better customer response. Use the rework estimator to change the assumptions rather than borrowing these example figures as your own business case.
Put the cost of change beside the benefit.
Include discovery and implementation work, internal staff time, platform or vendor costs if relevant, training, review and ongoing maintenance. State which costs are estimates and which have been quoted. Avoid double-counting the same hour as both saved capacity and additional revenue. Keep a conservative scenario where the expected time reduction does not materialise. A decision maker should see the assumptions they are accepting, not only the optimistic outcome.
Approve the smallest testable commitment.
A useful pilot might standardise one disputed metric, assign its business owner and run two reporting cycles with a documented quality check. Agree the acceptance evidence before the work starts: an approved definition, a traceable calculation, an issue owner and a repeatable process. The pilot is not a promise to eliminate every data problem. If an essential dependency is missing, describe a stop or rescoping decision rather than assuming more budget will solve it.
Review what actually changed.
Repeat the baseline observation after the pilot under comparable conditions. Note changes in team size, reporting volume, systems or business definitions. Distinguish recorded time, people’s estimates and management judgements. Report both improvements and work that remains. Expand the programme only when the sponsor understands the evidence, the operating cost and the next decision. A candid small result is a stronger foundation than an impressive return assembled from untested assumptions.
Start with one concrete example.
Use the related resources to prepare a bounded conversation. Keep confidential records in your approved internal environment.