AI readiness assessment
Sixteen statements. About two minutes. No email required. See where your business systems stand and the most practical place to start.
Ready when you are.
You'll see one statement at a time. Rate it from 1 (Strongly disagree) to 5 (Strongly agree), then move to the next. At the end you'll see your stage and a concrete next step.
Data
Our key business data lives in a single system, not spread across multiple tools and files.
Data
Anyone in our organization can get the data they need to make decisions without waiting for a specific person to pull it.
Data
We trust our core business data enough to make important decisions without manually verifying it first.
Data
Data quality is consistent across departments.
Process
Our core workflows are documented well enough that a new employee could follow them independently.
Process
Our core processes are followed as designed, not through informal workarounds.
Process
We regularly review and update our processes based on measurable outcomes and business objectives.
Process
Each of our key processes has a clear owner accountable for how it performs.
Culture
Our leadership actively sponsors technology adoption, not just delegating it to IT or individual teams.
Culture
When we introduce new tools or ways of working, adoption is real, not surface compliance with people reverting to old habits behind the scenes.
Culture
When technology implementations hit obstacles, we work through them rather than scaling back or abandoning the effort.
Culture
People across the organization can name specific, practical ways AI might improve their own work, not just talk about it in abstract terms.
Tech
Our core business applications share data with each other automatically, without manual re-entry or import/export.
Tech
Automation is a standard part of how most teams work day to day, not limited to a few departments.
Tech
Teams across our organization work from a shared view of the customer, not each maintaining their own version.
Tech
Our IT team can implement new integrations without it becoming a months-long project.
Stage —: —
You have software in place, but most of the work still happens through manual coordination. Data sits in different places, processes live in people's heads, and no single system is fully trusted as the source of truth.
What this looks like in practice
- Different teams keep their own version of the same data.
- Key information lives in spreadsheets, inboxes, and individual desktops.
- Reporting means compiling from multiple sources by hand.
- Status tracking happens through memory or informal chat.
- Work moves through email and meetings, not systems.
- Records get updated after the fact, not as part of the work.
- No system is trusted as the source of truth.
- When something breaks, tracing what happened is hard.
What AI can do at this stage
Handle self-contained tasks that don't depend on clean data or connected systems. Drafting communications, summarizing documents, answering general questions. Real value, but limited. AI here works as a personal productivity tool, not an operational layer.
What AI can't do at this stage
Anything that requires knowing your business. AI can't automate a process that isn't documented, surface insights from data that isn't captured, or make reliable recommendations from messy inputs. So AI stays tactical, helping individuals work a little faster, not connecting dots across the business.
Next steps
Pick one core process and map every place its data currently lives. Consolidate around a single system of record for that process before expanding to others. Set basic data entry standards so information gets captured the same way every time. Make visibility your first goal. Automation comes after.
Your teams have moved off manual processes into proper software, but each tool operates on its own. Getting a complete picture of anything still means jumping between platforms and stitching data together by hand.
What this looks like in practice
- Each team has a system, but those systems don't share data.
- Customer context has to be assembled across multiple tools.
- Teams switch platforms to understand a single situation.
- Handoffs between teams rely on manual updates or notifications.
- Duplicate records exist across systems with small inconsistencies.
- Dashboards exist but need interpretation, not action.
- Data is available, just not at the moment it's needed.
- If AI is in use, it works inside individual tools, not across workflows.
What AI can do at this stage
Deliver real value inside individual tools where those tools have clean data. If your CRM is well-kept, the AI features inside it probably work fine. Same for support, marketing, finance. Each one as a standalone.
What AI can't do at this stage
Work across the boundaries between tools. The valuable use cases (churn prediction, intelligent routing, personalized outreach, flagging at-risk accounts) need a connected view of the customer that siloed systems can't provide. AI at this stage is a series of disconnected features, not a coherent layer.
Next steps
Audit the handoffs between your most critical teams and find where context gets lost or recreated by hand. Prioritize integrations that eliminate duplicate data entry, not just integrations that share reports. Define what a complete record looks like for your key objects (customers, deals, tickets) and enforce it at the point of capture. The goal is tools that work together, not tools that work independently.
Your systems are mostly connected and largely capable, but adoption is uneven. Outcomes still depend more on who's doing the work than on the process itself. People bypass the system when it gets in the way, and the system gets in the way often enough that the bypasses have become standard practice.
What this looks like in practice
- Systems are integrated, but data quality varies a lot.
- Teams follow different versions of the same process.
- Critical fields get skipped or filled inconsistently.
- Exceptions are handled outside the system (email, chat, side processes).
- Automation exists, but gets bypassed or overridden often.
- Time spent reviewing AI outputs eats into the efficiency gains.
- Reports give a general picture, but the numbers need verification.
- Workarounds that started as temporary fixes have become permanent.
What AI can do at this stage
Generate useful output often enough that people notice when it works. Recommendations, summaries, and predictions are mostly directionally correct. Teams with strong individual contributors get real value by combining AI output with their own judgment.
What AI can't do at this stage
Operate reliably enough to be trusted without supervision. Inconsistent data produces inconsistent results, and once a team learns not to trust AI output, that skepticism is hard to walk back. The risk isn't that AI fails outright. It's that uneven results create uneven adoption, and uneven adoption makes it nearly impossible to tell whether AI is actually working.
Next steps
Identify the processes that get bypassed most often and figure out why. Tighten data quality at the source by making the right way to work the easiest way to work. Reduce reliance on individual expertise by documenting and systematizing how exceptions get handled. Build a short track record of reliable AI output in a low-stakes area so teams start to trust the recommendations.
Your systems reflect how the business actually operates. Data flows automatically, context is shared across teams, and AI runs on top of a foundation it can trust. At this point, the question is no longer whether AI can help. It's where to point it.
What this looks like in practice
- Teams work from a shared, consistent view of customers and operations.
- Workflows are followed because they're embedded in the system, not enforced manually.
- Clean, relevant data gets captured as part of the work, not after.
- Cross-team handoffs happen automatically and predictably.
- Exceptions are visible in the system and handled systematically.
- AI recommendations are part of regular decision-making.
- AI output is trusted because results are consistent over time.
- Teams spend less time gathering information and more time acting on it.
What AI can do at this stage
Function as an operational layer, not just a feature. With reliable data, connected context, and consistent processes, AI takes on work that previously required human coordination (routing, prioritization, anomaly detection, next-best-action recommendations) and does it at a scale and speed no team matches by hand.
What AI can't do at this stage
Replace judgment on decisions that need context your systems don't capture. Relationships, organizational dynamics, strategic nuance, edge cases with no precedent. The point at Stage 4 isn't to remove people from the loop. It's to make sure people spend their time on the decisions that actually require them.
Next steps
Shift from running your systems to building on top of them. Identify where autonomous action (not just insight) creates the most value. Invest in feedback loops so AI output improves over time based on real outcomes. Expand connected context to teams or processes still operating at lower stages. Move from using AI reactively to deploying it as part of how work gets initiated and prioritized.
Want a hand with your next step?
Wherever you landed, the next move depends on more than a score. A short consult covers what you're working with, where the foundation needs work first, and the most practical place to start.