
The SME Data Readiness Checklist: Is Your Data AI-Ready?
A practical guide to assessing whether your business data is ready for AI. Includes a downloadable checklist covering data inventory, quality, access, and governance.
Before you can leverage AI effectively, you need to answer a fundamental question: Is your data ready?
Organizations often find that AI projects struggle due to data quality issues. This guide provides a practical framework for assessing your data readiness across five critical dimensions.
Why Data Readiness Matters
AI systems learn from data. If your data is incomplete, inaccurate, or inaccessible, your AI will produce incomplete, inaccurate, or inaccessible results. This is a straightforward principle.
Industry research suggests data scientists spend a significant portion of their time on data preparation and only a smaller portion on actual analysis and modeling. Getting your data right is the majority of the work.
The Five Dimensions of Data Readiness
1. Data Inventory: Know What You Have
The first step is understanding what data assets you actually possess.
Key Questions:
- What data do we collect about customers, operations, and finances?
- Where does each data type live (CRM, spreadsheets, accounting software)?
- Who owns each data source?
- How old is our historical data?
Action Items:
- Create a data inventory spreadsheet listing all data sources
- Document the owner/steward for each data source
- Note the format and location of each dataset
- Identify any data that exists only in physical form
2. Data Quality: Assess Accuracy and Completeness
Poor quality data leads to poor quality AI outputs. Assess your data across these five quality dimensions:
| Dimension | Question | Red Flag |
|---|---|---|
| Accuracy | Is the data correct? | High error rates in reports |
| Completeness | Are there missing values? | Many blank fields |
| Consistency | Is the same thing recorded the same way? | Multiple formats for dates/names |
| Timeliness | Is the data current? | Last update was months ago |
| Uniqueness | Are there duplicates? | Same customer appears multiple times |
Quick Quality Audit:
- Export a sample of 100 records from your main database
- Manually check 10 records against source documents
- Calculate error rate and extrapolate
3. Data Access: Can You Get to It?
Data locked in silos is data that cannot power AI.
Access Assessment:
- Can data be exported in standard formats (CSV, JSON, API)?
- Are there technical barriers to accessing data?
- Are there legal or contractual restrictions?
- Do you have the right permissions and credentials?
Common Access Blockers:
- Legacy systems with no export functionality
- Data locked in SaaS platforms with limited API access
- Departmental silos with political barriers
- Vendor lock-in with proprietary formats
4. Data Governance: Who Controls What?
Governance is essential for data management.
Governance Checklist:
- Data ownership is clearly defined
- Access permissions are documented
- There is a process for data changes
- Sensitive data is identified and protected
- Retention policies exist
- Compliance requirements are understood (GDPR, etc.)
5. Data Integration: Can Systems Communicate?
AI often needs to combine data from multiple sources. Assess your integration readiness:
Integration Questions:
- Do your systems share common identifiers (customer ID, product SKU)?
- Can data flow between systems automatically?
- Do you have a single source of truth for key entities?
- Are there data transformation rules documented?
The Data Readiness Scorecard
Rate your organisation on each dimension (1-5 scale):
| Dimension | Score (1-5) | Notes |
|---|---|---|
| Inventory | ||
| Quality | ||
| Access | ||
| Governance | ||
| Integration | ||
| Total | /25 |
Scoring Guide:
- 20-25: AI-ready. Proceed with confidence.
- 15-19: Mostly ready. Address gaps before major AI projects.
- 10-14: Significant work needed. Start with data foundation projects.
- Below 10: Not ready. Focus on basic data management first.
Quick Wins for Data Improvement
If your score reveals gaps, here are immediate actions:
This Week
- Create your data inventory spreadsheet
- Identify your top 3 most critical data sources
- Check for obvious quality issues in those sources
This Month
- Establish data ownership for critical sources
- Clean up duplicate records in your CRM
- Document your data access credentials
This Quarter
- Implement data validation rules at entry points
- Create a simple data governance policy
- Evaluate integration options between key systems
Common Data Readiness Mistakes
Mistake 1: Assuming Clean Data Never assume your data is accurate. Always validate with spot checks.
Mistake 2: Ignoring Unstructured Data Documents, emails, and images contain valuable information. Include them in your inventory.
Mistake 3: Underestimating Integration Effort Connecting systems takes longer than expected. Budget extra time.
Mistake 4: Neglecting Data Security AI projects often require broader data access. Do not compromise security.
Next Steps
- Complete the scorecard for your organisation
- Identify your biggest gap and create an action plan
- Take our AI Readiness Assessment to see how data fits into your overall readiness
Data readiness is one of six pillars in our AI Readiness Assessment. Your data score directly impacts your ability to implement AI use cases effectively.
Ready to assess your complete AI readiness? Take the free assessment to get your personalised 90-day action plan.

