The 90-Day AI Readiness Sprint: A Practical Playbook for Mid-Market Leaders

A phase-by-phase guide to building readiness while your competitors are still in committee meetings

---

MIT’s research found that mid-market companies implement AI in 90 days while enterprises take 9 months or longer.

This is not a small difference. This is your structural advantage.

While enterprises navigate approval layers, align stakeholders across regions, and manage pilot purgatory, you can be deployed, learning, and iterating.

But 90 days of unfocused activity produces nothing. Speed without direction is just faster failure.

What follows is a practical playbook for the 90-day sprint. Phase by phase. Week by week. What to do, why it matters, and how to avoid the mistakes that derail most initiatives.

This is not theory. This is the methodology I have developed across 25 years of transformation work, refined through successes and failures, designed specifically for mid-market organizations that need to move decisively with limited resources.

---

Before You Start: The Prerequisites

The 90-day sprint assumes certain prerequisites are in place. Without them, you are not ready to sprint. You are ready to prepare for a sprint.

Prerequisite 1: A committed sponsor.

Someone at the executive level must own this initiative. Not approve it. Own it.

The sponsor must have authority to allocate resources, remove obstacles, and make decisions without endless escalation. They must be willing to spend political capital when resistance emerges.

If you do not have a committed sponsor, do not start the sprint. Find the sponsor first.

Prerequisite 2: Defined scope.

The 90-day sprint is not “implement AI across the organization.” That is a multi-year program, not a sprint.

The sprint focuses on a specific domain, process, or use case. Customer service for one product line. Underwriting for one category. Operations for one facility.

If your scope is “AI transformation,” narrow it before you start. Broad scope guarantees slow progress.

Prerequisite 3: Available resources.

The sprint requires dedicated time from key people. Not “when they have time.” Dedicated time.

If your best people are 100% allocated to other work, the sprint will fail. They will be pulled away. Momentum will die. Progress will stall.

Ensure that sprint participants have capacity before you start.

Prerequisite 4: Tolerance for imperfection.

The sprint will produce imperfect results. Early deployments will have problems. Initial approaches will need adjustment.

If your organization punishes imperfection, the sprint will be paralyzed by fear of failure. People will slow down to avoid mistakes. The 90-day timeline will extend indefinitely.

Ensure that leadership has explicitly endorsed experimentation and learning from failure.

---

Phase 1: The Audit (Weeks 1-2)

The sprint begins with honest assessment. Not what you think is true. What is actually true.

Week 1: Leadership and Data Assessment

These are the two highest-weighted dimensions. Start here.

Leadership assessment questions:

Can the executive sponsor articulate in two sentences what AI will change about competitive position? Not generalities. Specific claims about specific changes.

Has the leadership team resolved strategic tensions about AI direction? Cost reduction versus capability building? Defense versus offense? If tensions remain, they will surface later and derail progress.

Are there executives who are skeptical or opposed? Name them. Understand their concerns. They will become obstacles if not addressed.

Does the sponsor personally use AI tools? If not, their credibility in driving adoption is compromised.

Data assessment questions:

For your defined scope, what data is needed? Where does it reside? Who controls it?

Can you access that data in days or does access require weeks of negotiation? Test this. Do not assume.

What is the quality of the data? Is it clean, consistent, and complete? Or is it messy, contradictory, and full of gaps?

Is there documentation of what the data means? Or does understanding require tribal knowledge?

Week 1 outputs:

Leadership readiness rating: Ready, needs work, or not ready.

Data accessibility map: What data exists, where it is, who controls it, how long access takes.

Data quality assessment: Honest evaluation of quality for the specific use case.

Identified blockers: Leadership gaps and data obstacles that must be addressed.

Week 2: Capability, Process, Governance, Culture Assessment

Capability assessment questions:

For the people who will work with AI in your defined scope, what is their current capability? Can they use AI tools? More importantly, can they evaluate AI outputs?

Do they have the domain expertise to catch AI errors in this specific domain? Or are they generalists who might miss domain-specific problems?

What training have they received? Tool usage training? Judgment development? Nothing?

Process assessment questions:

For the processes within your defined scope, are they documented? Actually documented, not “we know how it works.”

Were they designed or did they evolve accidentally? Is there rationale for why things are done the way they are done?

Where would AI intervene in these processes? Who would review AI outputs? What happens when AI is wrong?

Governance assessment questions:

Are there existing policies that apply to AI in this domain? Data usage policies? Decision authority policies?

Who decides what AI can do? Is this clear or ambiguous?

If AI makes an error that affects customers or operations, who is accountable? Is this defined?

Culture assessment questions:

In the team that will work with AI, is experimentation safe? Can people try things without fear of punishment for failure?

Will people question AI outputs or accept them uncritically? Is there a history of healthy skepticism or passive acceptance?

Is there fatigue from previous change initiatives? Will this sprint face cynicism from past failures?

Week 2 outputs:

Capability gap analysis: What capability exists and what must be developed.

Process readiness assessment: Which processes are ready for AI, which need redesign.

Governance requirements: What policies and accountability structures are needed.

Culture risks: What cultural factors might derail the sprint.

End of Phase 1 deliverable:

A consolidated readiness assessment across all six dimensions, with specific gaps identified and prioritized.

This assessment is not a thick report nobody reads. It is a focused document that answers: What must be true for AI deployment to succeed in our defined scope, and what gaps must we close?

---

Phase 2: The Alignment (Weeks 3-4)

Assessment without action is worthless. Phase 2 translates assessment into alignment and decisions.

Week 3: Leadership Alignment

This week focuses entirely on getting leadership aligned.

The alignment session:

Gather the leadership team. Not an update meeting. A decision meeting.

Present the assessment findings. Be direct about gaps. Do not soften the message to avoid discomfort.

Force decisions on strategic tensions:

Is this AI initiative primarily about cost reduction, revenue growth, capability building, or competitive defense? Pick one primary objective. Secondary objectives can exist, but primary must be clear.

What resources are we committing? Specific budget. Specific people. Specific time allocation.

What does success look like in 90 days? Define measurable outcomes. Not vague aspirations.

What are we willing to sacrifice? Every initiative requires trade-offs. Name them.

Addressing skeptics:

If skeptical executives were identified in the assessment, address their concerns directly.

Do not avoid conflict. Surface it. Unresolved skepticism becomes underground resistance.

Either convert skeptics through genuine engagement with their concerns, or get explicit commitment to support despite disagreement.

The commitment ceremony:

At the end of the alignment session, ask for explicit commitment.

Not “does everyone agree?” That gets passive head nods.

Instead: “I am asking each of you to commit to supporting this initiative for 90 days, including providing resources when requested, removing obstacles when they arise, and not pulling the plug based on early results that may not look good due to the J-curve. Can you make that commitment?”

Go around the room. Get individual commitments. Document them.

This feels awkward. That is intentional. Explicit commitment is more durable than implicit assumption.

Week 3 outputs:

Documented strategic direction with clear primary objective.

Resource commitments with named people and specific budgets.

Success metrics for day 90.

Individual leadership commitments documented.

Week 4: Gap Prioritization and Sprint Planning

With leadership aligned, now prioritize which gaps to address and plan how.

Gap prioritization:

Not all gaps can be closed in 90 days. Prioritize ruthlessly.

Which gaps are blocking? These must be addressed or deployment cannot happen.

Which gaps are degrading? These will hurt results but not prevent deployment.

Which gaps are acceptable for now? These can be addressed after the sprint.

Focus sprint effort on blocking gaps. Monitor degrading gaps. Accept acceptable gaps.

Sprint team formation:

Identify the specific people who will execute the sprint.

This is not a committee. It is a working team with dedicated time.

Assign clear roles. Who leads the sprint? Who handles data preparation? Who develops capability? Who manages the technology? Who handles governance?

Week-by-week plan:

Create a specific plan for weeks 5-12.

What happens each week? What are the deliverables? What are the dependencies?

This plan will change as you learn. But starting without a plan guarantees drift.

Risk identification:

What could derail this sprint? Leadership distraction? Data access delays? Technology problems? Resistance from affected workers?

Name the risks. Identify mitigation approaches. Assign ownership for monitoring each risk.

Week 4 outputs:

Prioritized gap list with clear categorization.

Sprint team with named members and defined roles.

Week-by-week plan for weeks 5-12.

Risk register with mitigation approaches.

End of Phase 2 deliverable:

A sprint charter documenting direction, resources, team, plan, and risks. This becomes the reference document for the remaining 8 weeks.

---

Phase 3: The Activation (Weeks 5-8)

Assessment is complete. Alignment is achieved. Now the real work begins.

Week 5: Foundation Building

Address the blocking gaps identified in Phase 1.

Data access:

If data access was a blocking gap, resolve it now. Meet with data owners. Get agreements. Establish access.

Do not accept “we’ll get you access soon.” Get specific commitments with specific dates. Escalate to sponsors if needed.

Basic governance:

If governance was a blocking gap, establish minimum viable governance now.

Who can approve AI deployment? Who reviews outputs? Who handles errors?

This does not need to be comprehensive. It needs to be clear enough to proceed.

Initial capability development:

If capability was a blocking gap, begin development now.

This might be training sessions on AI tools. It might be workshops on evaluating AI outputs. It might be practice exercises with feedback.

Focus on the people who will work with the initial deployment. Broader capability development comes later.

Week 5 outputs:

Data access established or firm commitments with dates.

Minimum viable governance documented and communicated.

Initial capability development underway for deployment team.

Weeks 6-7: Deployment Preparation

Prepare for initial AI deployment within your defined scope.

Technology configuration:

Select and configure the AI technology for your specific use case.

This might be adapting a platform you already have. It might be implementing a new tool. It might be building something custom.

Favor speed over perfection. You can improve the technology after learning from initial deployment.

Process integration design:

Design how AI will integrate into existing workflows.

Where does AI output go? Who reviews it before action? What happens when AI is uncertain? What happens when AI is wrong?

Document the integration clearly. Train the people involved.

Testing:

Test the AI in conditions that approximate real use.

Does it work with your actual data? Does it handle the variations that occur in real operations? Where does it fail?

Document what you learn. Adjust before deployment.

Communication preparation:

Prepare communication for people affected by the deployment.

What is changing? Why? What is expected of them? What support is available?

Do not surprise people. Surprises create resistance.

Weeks 6-7 outputs:

AI technology configured for the use case.

Process integration designed and documented.

Testing completed with issues identified and addressed.

Communication prepared for affected stakeholders.

Week 8: Initial Deployment

Deploy AI to initial users within your defined scope.

Deployment approach:

Start narrow. Not everyone in scope at once. A subset of users, processes, or transactions.

This limits blast radius if something goes wrong. It creates manageable scale for learning.

Support structure:

Provide intensive support during initial deployment.

Someone available to answer questions. Quick response to problems. Visible presence that signals this matters.

Do not deploy and disappear. The first days shape adoption patterns.

Learning mechanisms:

Establish how you will learn from initial deployment.

Daily check-ins with users. Tracking of issues and questions. Observation of how AI is actually being used.

The goal is rapid learning, not just deployment.

Week 8 outputs:

AI deployed to initial users.

Support structure active.

Learning mechanisms operational.

Initial observations documented.

---

Phase 4: The Breakaway (Weeks 9-12)

Initial deployment is complete. Now learn rapidly and build foundation for scaling.

Week 9: Intensive Learning

Focus entirely on learning from initial deployment.

User feedback:

Talk to users. Not surveys. Conversations.

What is working? What is frustrating? What did you expect that is not happening? What surprised you?

Listen without defending. Users are telling you what needs to change.

Performance analysis:

Analyze how the AI is actually performing.

Where is it accurate? Where does it fail? What patterns appear in errors?

Compare performance to expectations. Understand the gaps.

Process observation:

Observe how AI is being integrated into actual work.

Are people using it as designed? Have they developed workarounds? Are they skipping verification steps? Are they over-relying or under-relying?

Actual behavior reveals what training and design missed.

Week 9 outputs:

User feedback synthesized.

Performance analysis completed.

Process observation documented.

Prioritized list of changes needed.

Weeks 10-11: Iteration

Based on learning, iterate on the deployment.

Quick fixes:

Address issues that can be fixed quickly.

Configuration changes. Process adjustments. Additional training. Communication clarification.

Do not wait for perfect solutions. Improve continuously.

Deeper changes:

Identify issues that require more substantial work.

Technology limitations that need different approaches. Process redesigns that require more time. Capability gaps that need structured development.

Plan these for post-sprint work. Document clearly.

Expanded deployment:

Based on learning, expand deployment within scope.

More users. More processes. More transactions.

Expand deliberately, not all at once. Each expansion is an opportunity to learn.

Weeks 10-11 outputs:

Quick fixes implemented.

Deeper changes documented and planned.

Deployment expanded within scope.

Continued learning and iteration.

Week 12: Foundation for Scale

The sprint ends not with completion but with foundation for what comes next.

Results documentation:

Document what was achieved in 90 days.

What was deployed? To how many users? With what results?

Be honest about both successes and failures. Inflated claims undermine credibility.

Lessons learned:

Document what you learned.

What worked? What did not? What would you do differently?

This documentation enables future sprints to start ahead of where this one started.

Scaling plan:

Based on 90 days of learning, what should come next?

Expand this deployment? Start a new sprint in a different domain? Address the deeper changes identified? Build more capability?

Create a concrete plan with timelines and resources.

Leadership update:

Return to the leadership team with results.

This is the commitment they made in Week 3. Show them what their commitment enabled.

Ask for continued support for the next phase.

Week 12 outputs:

Results documented with honest assessment.

Lessons learned captured.

Scaling plan prepared.

Leadership updated and next phase approved.

---

Common Failure Points

The 90-day sprint can fail. Here is what typically causes failure.

Scope creep.

The sprint starts focused. Then someone adds another use case. Then another. Suddenly you are trying to transform everything and accomplishing nothing.

Protect scope ruthlessly. Additional use cases are future sprints, not this sprint.

Leadership distraction.

The sponsor gets pulled into other priorities. The weekly check-in gets skipped. Decisions wait for attention that is not available.

The commitment ceremony in Week 3 helps. But sponsors must actively protect their attention.

Data delays.

Data access takes longer than expected. The sprint stalls waiting for data that was promised but not delivered.

Escalate early. Do not wait politely while the timeline burns.

Premature evaluation.

The J-curve means results may look poor in early weeks. If leadership evaluates based on early results, they may kill an initiative that was on track.

Set expectations in Week 3. Remind leadership during the sprint. Measure leading indicators, not just lagging outcomes.

Invisible resistance.

Someone is quietly undermining the sprint. Not openly opposing, but not providing data, not making people available, not supporting when asked.

Identify resistance early. Address it directly. Escalate if needed.

---

Day 91

The sprint ends. What should be true on Day 91?

You have AI deployed and operating in your defined scope.

You have learned what works and what does not in your specific context.

You have developed capability in the people who will sustain and expand AI usage.

You have leadership commitment to continue and expand.

You have a concrete plan for what comes next.

Enterprises are still in committee meetings. You are learning from real deployment.

That is the 90-day advantage. That is what mid-market speed enables.

---

Are you ready for a 90-day sprint? What obstacles do you anticipate?

The AI Readiness Scorecard provides the assessment foundation for Phase 1 of the sprint. It takes ten minutes and shows exactly where your gaps are.

Comment “SCORECARD” below and I will send you access.

The 90 days start when you decide to start. The question is whether that is today.