Resources
Free checklists and templates
Four free downloads: a health check for a running integration estate, a readiness checklist and three working templates for a data migration, and a sample KPI dictionary. The integration and migration documents are adapted from my own delivery playbook. All are direct downloads — no form, no email address.
Data Integration Health Check
Six failure modes, 28 checks.
Six failure modes that recur in enterprise integration, each with the checks that prevent it, and a source data contract to agree with each system owner. Adapted from the article Enterprise data integration: what usually breaks.
Record identity
When several systems hold the same customer, student or supplier under different keys, matching on name and email works for most records, and the rest becomes the project: false duplicates, merges that destroy history, two real entities turned into one.
Late corrections
Most source systems allow backdated changes. An extract that only picks up new records misses them, and the target quietly disagrees with the source for that period.
Safe re-runs
A pipeline fails halfway, someone re-runs it, and part of the data now exists twice. Append-only loads are the easiest to build and fail only under conditions nobody tested.
Reconciliation
Without it, the first wrong number is found by a senior reader weeks later, and from then on every number in the platform is provisional.
Schema change
A source upgrade renames a field or adds a code value. The pipeline either fails in a confusing way or succeeds while dropping or mis-casting the new data.
Ownership and handover
The pipelines run and reconcile, then the person who built them changes role, and the estate becomes something the organisation is afraid to touch.
Download the health check (325 KB)
PDF · 4 pages · A4 · 325 KB · no sign-up
Data Migration Readiness Checklist
Six stages, 31 checks.
Each stage has short items you can tick off, and the gate at the end of it decides go or no-go. Use it before you commit to a go-live date and again at each stage gate: any item you cannot answer with a document, a date or a number is where the risk is. These are the six stages and why each one matters; the PDF closes with five questions to ask of any migration plan.
Scope and profiling
Estimates built on a count of tables rather than on profiled data are where migration schedules usually break, at cleansing or user acceptance testing, after the date is already committed.
Mapping
Transformation rules written from documentation or assumption surface as misclassified records after go-live, when the fix means correcting production data and rebuilding users' trust.
Data quality
The source keeps changing after you profile it. Without a re-check, the data you tested is not the data you cut over.
Test and reconcile
A job that finished without errors is not evidence that the data is right. Tests on samples miss rare historical values and the effects of scale, and without automated reconciliation, differences are found by users weeks later with nothing left to trace them back to.
Cutover
Without numeric triggers agreed in advance, the choice between pushing on and rolling back is made under pressure in the middle of the night, and sunk cost usually wins.
After go-live
A migration ends when the old system is switched off, not at go-live. Until then you pay for two systems, and some users keep entering data in the old one.
Download the checklist (244 KB)
PDF · 4 pages · A4 · 244 KB · no sign-up
Data Migration Templates
Three documents that carry the work.
Each template has its columns defined, a short note on how to fill it in, and blank rows; copy them into whatever tool your project uses. The article Data migration, stage by stage explains where each one fits, with a worked example.
Source-to-target mapping
The contract the business signs: one row per target field, with the source, the rule, the default and who approved it.
Rollback plan
The way back, decided before the night: numeric triggers, who decides, the point of no return, and the rehearsal record.
Cutover runbook
Every step of the cutover with an owner, a duration taken from rehearsal, how it is verified and when to stop.
Download the templates (318 KB)
PDF · 5 pages · A4 · 318 KB · no sign-up
KPI Dictionary Sample
What a KPI dictionary looks like.
Ten metrics from the SenMinh reference build, a Power BI report for a fictional school group on synthetic data. Each has one definition and one calculation, including the edge cases that usually make two reports disagree. It is the format the KPI Definition Audit delivers on your own metrics.
PDF · 2 pages · A4 · 230 KB · no sign-up
Integration & Migration Review
If a migration is coming and you want an independent read before you commit to a date, the Integration & Migration Review is a fixed-scope review built for that point. EthanCorp designs and runs data integration, analytics and AI automation systems for organisations that need their fragmented data to be trustworthy, built as durable systems, not demos. The review profiles your source systems as they actually run today, ranks the failure points by business impact, and produces a first source-to-target mapping for the domain that matters most.
- Response time
- Within two business days
- Based in
- Ho Chi Minh City, Vietnam — working across Asia and remote