What Happens After Software Launch? Maintenance, Costs and Responsibilities
A practical guide to post-launch software maintenance: who owns what, which running costs to budget, and what a support agreement should define.
To move from Excel to a CRM, inventory your spreadsheets, clean and identify records, map fields to the target system, test a representative sample, then reconcile the final import before switching teams over. The goal is to preserve customer relationships and useful history, not simply upload every row.
This checklist applies to planning a migration into an existing CRM or a custom system. Import formats, supported objects and rollback options vary by platform, so confirm the target CRM’s current requirements before changing production records.
1. Decide what the CRM needs to support
List the jobs your team must perform after migration: find a customer, assign an owner, update an opportunity, schedule a follow-up and report on a pipeline. Define which records and relationships each task needs.
A contact, company and deal are different entities. One contact may have several deals, and a company may have many contacts. Repeating a customer’s name on each spreadsheet row does not explain those relationships. Decide the target structure before preparing an import.
2. Inventory the source files and preserve a baseline
Identify the authoritative file for each dataset and appoint a business owner to resolve conflicts. Keep a dated, access-controlled copy of the original files. Record row counts, headers, formulas, hidden sheets and attachments so important information does not disappear during conversion.
Decide what to migrate, archive or investigate. A calculation may need to become a CRM formula; a note may belong in activity history. Transfer the information needed for the business process, and confirm your organization’s rules for sensitive records and retention.
3. Clean and standardize the data
- Use consistent column headers and one defined value per field.
- Standardize dates and verify ambiguous values such as 04/05/2026 with the data owner.
- Preserve identifiers and phone numbers as text where needed so leading zeros are retained.
- Align sales stages and owner names with valid target values.
- Separate missing values from intentional blank updates.
- Review stale records and conflicting details instead of guessing which value is correct.
If you export CSV, check character encoding, delimiters and the exported values. Formatting, comments and spreadsheet formulas should not be assumed to transfer as CRM behavior.
4. Define unique identifiers and duplicate rules
Choose a stable source identifier for each record and keep a mapping to the new CRM identifier. A person’s name is rarely sufficient. Email can be useful for contacts, but shared addresses, missing values and changes require an explicit handling rule.
For example, HubSpot’s import documentation explains identifiers used to update, deduplicate and associate records. Those rules are product-specific; do not assume that another CRM matches records in the same way.
Decide whether each batch creates records, updates records or does both. Define what happens when a possible match is uncertain. Keep a review queue rather than automatically merging people solely because they share a name.
5. Create a field-mapping sheet
Document the source column, target entity, target field, transformation and validation rule. Have a person who uses the customer data review the mapping before importing.
| Source column | Target | Rule to agree |
|---|---|---|
| Customer ID | Contact: external identifier | Preserve as text; require a unique value |
| Company | Company record and contact relationship | Match an approved company identifier |
| Sales Stage | Deal: pipeline stage | Map source labels to allowed stages |
| Account Manager | Record owner | Map to an active CRM user |
| Next Follow-up | Task or follow-up date | Confirm date, owner and time-zone behavior |
Required fields and import capabilities differ. Microsoft’s model-driven app import guidance describes source files and column mapping for its platform. Use your selected system’s documentation to validate your final import template.
6. Test a representative sample
Use a sandbox or isolated test environment where available. Include difficult records: missing emails, multiple deals, unusual characters, old dates and duplicate candidates. Use masked or synthetic data where possible, with controlled access for any real customer data.
Check whether imports trigger emails, tasks, integrations or automation. Put a controlled test plan in place before loading data so a rehearsal does not accidentally contact customers or update another live system.
7. Reconcile records, relationships and workflows
Record counts are necessary but insufficient. A successful import may still link deals to the wrong companies or assign records to an inactive owner.
- Account for source rows as created, updated, excluded, rejected or merged according to the agreed rules.
- Check required fields, identifiers and representative values.
- Compare deal totals by currency and stage where relevant.
- Verify contact-company-deal relationships and activity history.
- Test who can see and edit the imported records.
- Ask users to complete the workflows defined in step one.
8. Plan the final cutover and recovery
Agree when spreadsheet editing stops and how changes made during migration are captured. Rehearse the final import and define a go/no-go owner. Document how to reverse a batch or restore a known state using the platform’s supported methods; do not assume deleting imported records reverses updates to existing records.
After approval, make the CRM the working source and keep spreadsheets read-only for an agreed period. Train the team, assign an owner for rejected records and check early reports before retiring the old process.
Frequently asked questions about CRM migration
Can I import Excel directly?
Some CRMs accept spreadsheet files; others require CSV or a template. File support does not guarantee that formulas, attachments and relationships transfer. Test the exact format and objects you need.
How long does a migration take?
Time depends on data quality, entities, relationships, history, integrations and review requirements. A clean contact list is a different task from migrating several connected spreadsheets with conflicting records.
Do we need a custom CRM to leave Excel?
No. A configured existing CRM may fit. Consider a custom system when essential workflows or integrations cannot be handled adequately. Start by mapping the process and comparing configuration, extension and custom development options.
Plan the workflow as well as the import
Zeoark’s CRM development services address customer records, workflows, integrations and migration planning. Discuss your CRM requirements with a description of your current spreadsheets and the tasks your team needs to complete.
