Is Your Company's Data Ready for AI? A Calgary Consultant's Checklist

Last updated: March 2026

Most mid-size Calgary companies are not as far from AI-ready data as they think — but they do have 3–5 specific gaps that will undermine any AI initiative if left unaddressed. Gartner’s research shows that organizations lacking AI-ready data foundations are abandoning the majority of their AI projects — the firm estimated 60% would be scrapped, and current evidence suggests that prediction is playing out. Yet for companies already running on Microsoft 365, a CRM, and standard business applications, the path to data readiness is shorter than the enterprise-focused reports suggest. This checklist is built specifically for 50–500 person companies in Calgary and Western Canada, based on what Solway sees in practice during our Discovery and Baseline Scan engagements.

What "Data Readiness" Actually Means for a Mid-Size Company

Data readiness for enterprise AI at Google or Amazon involves petabytes of structured data, machine learning pipelines, and dedicated data engineering teams. Data readiness for a 75-person Calgary company deploying Microsoft Copilot and exploring AI-assisted workflows means something fundamentally different.

For mid-market companies, data readiness answers three questions:

Can your AI tools access the information your team works with? If your documents live in a shared drive that Copilot can't index, or your CRM data is siloed from your project management system, AI tools can't help because they can't see the data.

Is your information organized enough for AI to understand it? AI tools work best when files are named consistently, organized in logical folder structures, and tagged with meaningful metadata. A SharePoint site with 10,000 files named "Final_v3_REALLY_FINAL.docx" produces poor AI results.

Is your data clean enough that AI outputs will be trustworthy? If your CRM has duplicate contacts, outdated records, and inconsistent naming conventions, any AI-generated analysis will inherit those problems. The principle is straightforward: AI amplifies the quality of your data, good or bad.

The Cisco AI Readiness Index 2025 found that just 13% of organizations worldwide qualify as "Pacesetters" — companies leading on AI value. Of those, 99% have a well-defined AI strategy. They're four times more likely to move AI pilots to production. The differentiator isn't having perfect data — it's having a plan to make data usable.

Solway's Data Readiness Checklist for Calgary Mid-Market Companies

This checklist is organized into five areas that Solway assesses during the Discovery and Baseline Scan phase of the AI Clarity Sprint. Each item is scored as Ready, Partially Ready, or Not Ready. You don't need a perfect score to start with AI — but you need to know where the gaps are before you deploy.

Area 1: Document and File Organization

This is where most mid-size Calgary companies have the biggest quick wins. AI tools like Microsoft Copilot pull from your Microsoft 365 environment — SharePoint, OneDrive, Outlook, Teams. The quality of what Copilot can do depends directly on how well this information is organized.

Are your documents stored in a centralized, cloud-based system? Files on local hard drives, USB drives, or legacy network shares are invisible to modern AI tools. If your team's institutional knowledge lives on individual machines, that's your first data readiness priority.

Do you have a consistent file naming convention? AI tools use file names to understand content. "Proposal_ClientName_Date" is usable. "Doc1_final_v2" is not. A company-wide naming convention takes one afternoon to establish and immediately improves AI tool performance.

Are your SharePoint sites structured logically? Copilot searches across your SharePoint environment. If critical documents are buried in nested folders with unclear names, or scattered across dozens of SharePoint sites with no taxonomy, AI tools struggle to find relevant information. A clean SharePoint architecture doesn't require a migration project — it requires consistent organization going forward.

Do you have permissions set appropriately? AI tools respect your existing Microsoft 365 permissions. If an employee shouldn't see HR records, Copilot won't surface those records to them. But this means your permissions need to be accurate and current. Many mid-size companies discover during AI readiness assessments that their permissions are overly broad or outdated.

Area 2: CRM and Customer Data

Your CRM (whether Salesforce, HubSpot, Dynamics, or another platform) is one of the richest data sources for AI applications — but only if the data is maintained.

Are your contact records deduplicated and current? Duplicate contacts, outdated job titles, and inactive records reduce the quality of any AI-generated analysis. A quarterly CRM cleanup is basic data hygiene that dramatically improves AI tool effectiveness.

Are your pipeline stages and deal properties used consistently? If different salespeople define "Qualified" differently, or if half the deals are missing key properties, AI-generated pipeline analysis will be unreliable. Standardization matters more than perfection.

Are client communications captured in the CRM? Emails, meeting notes, and call summaries that live only in individual inboxes are invisible to AI analysis. CRM integration with email (automatic logging) and meeting tools provides the data foundation for AI-assisted account management.

Area 3: Financial and Operational Data

For AI to support operational decision-making, your financial and operational data needs to be accessible and consistent.

Are your financial systems producing clean, structured reports? AI tools can analyze financial data, identify trends, and flag anomalies — but only if the source data is structured. If your financial reporting involves manual Excel manipulation with inconsistent formatting, start by standardizing your reporting templates.

Is your operational data captured digitally? Manufacturing companies tracking production on paper clipboards, or service companies logging time in disconnected spreadsheets, have a digitization gap that must be addressed before AI adds value. The AI opportunity often starts with basic process digitization.

Can you export data from your core systems? Some legacy business systems lock data behind proprietary formats or limited export capabilities. Before investing in AI, verify that your ERP, project management, and operational systems can export data in formats that AI tools can process.

Area 4: Email and Communication Data

Email is the largest unstructured dataset at most mid-size companies — and one of the areas where AI delivers immediate value through summarization, drafting, and search.

Is your team on a cloud-based email platform? Microsoft 365 or Google Workspace are prerequisites for AI email tools. On-premises Exchange servers limit AI capabilities significantly.

Are email retention policies defined and enforced? Without retention policies, mailboxes grow indefinitely, making AI search less effective and increasing compliance risk. Define how long emails are retained by category and enforce it consistently.

Are meeting recordings and transcripts being captured? AI meeting summarization (through Copilot, Otter, or similar tools) requires recordings. If your team isn't recording meetings in Teams or Zoom, you're missing one of the highest-value AI data sources. A simple policy change — "we record all internal meetings" — unlocks significant AI capability.

Area 5: Security and Governance

Data readiness includes knowing your data is protected and governed appropriately.

Do you have a data classification scheme? Not all data carries the same risk. Client-confidential information, employee records, and financial data need different handling than internal process documents. A basic classification scheme (Public, Internal, Confidential, Restricted) helps employees make informed decisions about what data enters AI systems.

Are your backup and recovery processes current? Before adding AI to your technology stack, verify that your data backup and disaster recovery processes are tested and reliable. AI doesn't increase data loss risk, but any technology expansion warrants a backup review.

Is there an existing data governance owner? Someone in the organization should be accountable for data quality, access policies, and compliance. In many mid-size Calgary companies, this responsibility is informal or undefined. Naming a data governance owner — even as part of an existing role — is a foundational step.

The 80/20 Rule of Data Readiness

Here's the perspective that matters most for mid-size Calgary companies: you don't need perfect data to start with AI. You need good-enough data in the areas where AI will deliver the most value.

Solway's Opportunity & Risk Matrix explicitly maps this. Quick Wins — the AI use cases you start with — are deliberately chosen to work with your current data quality. Email drafting, meeting summarization, and document search don't require pristine data — they require a cloud-based Microsoft 365 environment with reasonable organization. Most Calgary mid-market companies already have this.

Strategic Upgrades — advanced AI applications like predictive analytics, automated reporting, and AI-assisted decision-making — require better data foundations. But you don't need those foundations on day one. Build them incrementally while capturing value from Quick Wins.

Shaheer Tariq, Co-Founder of Solway, notes in his AI briefings: "The biggest data readiness mistake I see is companies trying to boil the ocean — launching a massive data cleanup project before touching AI. The companies that win start with AI on the data they have today, and use the momentum from early wins to fund the data improvements they'll need tomorrow."

How the AI Clarity Sprint Addresses Data Readiness

Solway's AI Clarity Sprint includes a Discovery and Baseline Scan (Step 2) that functions as a practical data readiness assessment. Over 1–2 weeks, we evaluate your Microsoft 365 environment, CRM data quality, document organization, and communication patterns to determine which AI use cases your data currently supports and which require data preparation.

The Sprint's deliverables include the Opportunity & Risk Matrix, which categorizes AI use cases into Quick Wins (ready now), Quality Lifts (minor data prep needed), Strategic Upgrades (significant data work required), and Not Yet (data foundations must be built first). This gives leadership a clear roadmap that matches AI ambition to data reality.

Frequently Asked Questions

How do we know if our data is "good enough" for AI?

If your team works primarily in Microsoft 365 (Outlook, Teams, SharePoint, OneDrive) and your files are stored in the cloud rather than on local machines, you likely have sufficient data foundations for basic AI tools like Copilot. The checklist above identifies specific gaps to address. You don't need perfect data — you need to know which gaps affect your highest-priority AI use cases.

How long does a data readiness assessment take?

For a 50–200 person company, a practical data readiness assessment takes 1–2 weeks. Solway includes this as part of the AI Clarity Sprint's Discovery and Baseline Scan. The output is a clear picture of what's ready, what needs work, and in what order.

What's the most common data readiness gap in Calgary mid-market companies?

Document organization in SharePoint. Most 50–200 person companies have years of accumulated files with inconsistent naming, unclear folder structures, and outdated permissions. This is also one of the easiest gaps to address — a structured cleanup and naming convention can be implemented in 2–4 weeks without disrupting daily work.

Does CAPG cover data readiness work?

CAPG covers training under the Digital and Technological skills category. Training that includes data management, SharePoint organization, and data governance practices qualifies for 50% reimbursement. The training components of an AI Clarity Sprint, which include data readiness education for your team, are CAPG-eligible.

Do we need a data scientist to get our data AI-ready?

No. For mid-market companies deploying tools like Microsoft Copilot, data readiness is primarily an organizational challenge, not a technical one. File naming conventions, SharePoint structure, CRM cleanup, and permission management can all be handled by your existing operations and IT staff with proper guidance. Solway's workshops and Sprint engagements provide this guidance without requiring a data science hire.

What if our data is in legacy systems that AI tools can't access?

This is common in Calgary's energy, manufacturing, and professional services sectors. The approach is to start AI deployment in areas where data is already accessible (Microsoft 365 environment) while planning data migration or integration for legacy systems as a parallel initiative. Don't let legacy system data delay your entire AI program.

How does data readiness relate to AI policy?

They're two sides of the same coin. Your AI policy defines what data can enter AI systems and how AI outputs should be reviewed. Your data readiness determines what data AI tools can actually access and how useful that data will be. The Solway System's AI Policy Framework includes data handling components that align directly with data readiness requirements.

Can we do a self-assessment of our data readiness?

Yes — use the checklist above as a starting point. Have your IT lead and operations manager walk through each item and score it as Ready, Partially Ready, or Not Ready. This gives you a baseline picture. For a more detailed assessment with specific recommendations, Solway's AI Clarity Sprint provides an expert-led evaluation with actionable next steps.