AI Readiness Assessment

Before the next AI budget, see what’s in the way.

Free. 24 questions about what your organization does today give you a six-area profile, not a peer rank, with research behind your two lowest areas.
Take the assessmentBook a 30-min call
A brass magnifying glass over handwritten notes on a worn desk, lit by one warm lamp

Most companies can’t yet show what AI returns.

Three large surveys, three different measures. In each one, most companies can’t yet show a return. Read them side by side, not as one trend.
56%
of CEOs say AI brought neither higher revenue nor lower costs in the past year.
Source: PwC, PwC's 29th Global CEO Survey: Leading through uncertainty in the age of AI (2026). Studied: 4,454 CEOs in 95 countries.
37%
of respondents attribute any EBIT impact to AI, about the same share as the year before.
Source: McKinsey & Company, The state of AI in 2026: On the road to ROI (2026). Studied: 1,719 participants in 97 nations.
60%
of companies reap hardly any material value from AI, despite substantial investment.
Source: Boston Consulting Group, The Widening AI Value Gap (Build for the Future 2025) (2025). Studied: 1,250 executives.

Where the hard part sits

Share of the challenges companies face implementing AI, as BCG reports it.
  • 70%People and process
  • 20%Technology
  • 10%AI algorithms
Most of the hard part is people and process, not the model.
Source: Boston Consulting Group, Where's the Value in AI? (2024). Studied: 1,000 CxOs and senior executives in 59 countries.

What the companies getting value do differently

Of 25 attributes McKinsey tested, redesigning workflows had the biggest effect on whether a company saw EBIT impact from gen AI. In its report published in early 2025, only 21% of respondents whose organizations use gen AI said they had fundamentally redesigned at least some workflows. In its 2026 survey, nearly three-quarters of AI high performers, a much smaller group, said the same.
Source: McKinsey & Company, The state of AI: How organizations are rewiring to capture value (2025); McKinsey & Company, The state of AI in 2026: On the road to ROI (2026).

Leaders and everyone else, on four practices

Share of each group that reports the practice. Each survey defines its own leaders.
LeadersComparison group
McKinsey
Leaders are AI high performers, about 6% of respondents, who report 5% or more of EBIT from AI and significant value from it. The rest are all other respondents.
Infrastructure supports core AI initiatives
More than 20% of the digital budget goes to AI
Source: McKinsey & Company, The state of AI in 2025: Agents, innovation, and transformation (2025). Studied: 1,993 participants in 105 nations, 2025.
Gartner
Leaders are high AI maturity organizations. The rest are low-maturity organizations.
Business units trust new AI and are ready to use it
AI initiatives stay in production three years or more
Source: Gartner, Gartner Survey Finds 45% of Organizations With High AI Maturity Keep AI Projects Operational for at Least Three Years (2025). Studied: 432 respondents in six countries, Q4 2024.

Their leaders commit, and measure

High performers are twice as likely to say their senior leaders show commitment to AI, and to have defined ways to measure its impact.
Source: McKinsey & Company, The state of AI in 2026: On the road to ROI (2026).

The CEO is in it personally

BCG reports that where the CEO personally owns the AI program, results are better on all three things it measured: value captured, employee joy and trust.
Source: Boston Consulting Group, AI at Work: Strategy Matters More Than Tools (2026).

They track outcomes

Of 12 adoption practices, tracking well-defined KPIs for gen AI had the most impact on the bottom line.
Source: McKinsey & Company, The state of AI: How organizations are rewiring to capture value (2025).

People get real training

Regular AI use is sharply higher among people with at least five hours of training plus in-person coaching. Only a third say they were properly trained.
Source: Boston Consulting Group, AI at Work: Momentum Builds, but Gaps Remain (2025).

No study settles how much each factor matters

Every link above is a correlation from surveys, mostly self-reported. McKinsey’s model of 25 attributes explains about a fifth of the variation in reported EBIT impact (an R-squared of 0.20). In its 2025 ranking, technology and data infrastructure matter but sit behind practices like workflow redesign. So our assessment gives you a profile across six areas, not a verdict on which one matters most.
Source: McKinsey & Company, The state of AI: How organizations are rewiring to capture value (2025); McKinsey & Company, The state of AI in 2025: Agents, innovation, and transformation (2025).

What the famous AI cases actually show

Read at their primary sources, the best-known enterprise cases report adoption and task time, nearly all self-reported. None publishes an audited return on AI. Here is what each one does show, and the caveat that goes with it.

Uber

What they report
16 agentic pods (two-week teams pairing engineers with a business expert) in two months. Capital allocation across 150 cities went from 15 hours to 30 minutes, and financial pacing reports from 2 days to 10 minutes.
Kind of evidence
Self-reported by Uber’s CTO, with no stated baseline method.
The caveat
Uber’s COO says it’s hard to link AI usage stats to more useful features. In the same interview he confirmed the CTO had said Uber blew through its 2026 AI budget early in the year.
What to take from it
The CTO’s own lesson: the wins come from redesigning a whole workflow, not automating one task.
Source: Uber, Uber CTO on Agentic Pods: taking agentic AI beyond engineering (2026); Rapid Response (WaitWhat), hosted by Bob Safian, Rapid Response: Uber's swerve on gas prices, hotels, and a driverless future, with Andrew Macdonald (2026).

JPMorganChase

What they report
Over 200,000 employees have access to its LLM Suite platform.
Kind of evidence
Statements at investor events. Usage, not a return.
The caveat
Jamie Dimon says about 150,000 weekly users believe they save about four hours a week, and the bank does not count that in its NPVs. The bank’s COO writes that the gains will free capacity for growth, and that the goal is not fewer people.
What to take from it
Count outcomes, not self-reported hours.
Source: JPMorganChase, JPMorganChase 2025 Investor Day Transcript (2025); JPMorganChase, JPMorganChase 2026 Company Update Transcript (2026); JPMorganChase, Line of Business CEO Letters to Shareholders (2025 Annual Report) (2026).

Moderna

What they report
A goal of 100% adoption and proficiency within six months, run by a dedicated team. Its internal chatbot reached more than 80% of employees.
Kind of evidence
A vendor customer story (OpenAI), plus Moderna’s leadership page, which in late 2024 showed one executive leading both people and digital technology.
The caveat
This is adoption. No business outcome is published.
What to take from it
The tool came with people work: a dedicated team ran the change alongside the rollout.
Source: OpenAI, Moderna and OpenAI (ChatGPT Enterprise customer story) (2024); Moderna, Tracey Franklin, Chief People and Digital Technology Officer (leadership bio) (2024).

Shopify

What they report
The CEO made AI use a baseline expectation, added it to performance reviews, and asks teams to show why AI can’t do the work before requesting more people.
Kind of evidence
The CEO’s published memo and the company’s annual filing.
The caveat
The filing shows about 7,600 employees at the end of 2025, down from about 8,100 a year earlier, while revenue grew from $8,880 million to $11,556 million. It does not attribute either change to AI.
What to take from it
A policy to copy or reject on its merits, not a measured result.
Source: Shopify, Reflexive AI usage is now a baseline expectation at Shopify (Tobi Lütke memo) (2025); Shopify Inc., Shopify Inc. Form 10-K for fiscal year 2025 (2026).

Klarna

The counter-case
What they report
Its IPO prospectus says the AI assistant handled 69% of customer service chats in the year to June 2025, did work equal to over 700 full-time agents, and saved about $39 million in 2024.
Kind of evidence
An SEC filing. The agent equivalence and the savings are Klarna’s own estimates. Audited customer service and operations costs fell $37 million, or 15%, in 2024, which the filing attributes to efforts to optimize customer service costs.
The caveat
In May 2025 the CEO admitted he had overstated how fast AI would take on the work, and said Klarna was hiring people for higher-end service conversations while expanding AI.
What to take from it
The honest model is two tracks: AI handles volume, and people handle the conversations that need judgment.
Source: Klarna Group plc, Klarna Group plc final IPO prospectus (Form 424B4) (2025); Big Technology, Klarna CEO: We're Giving AI More Customer Service Work, Not Less (2025).

What it looked like in our own work

Three engagements, with only the numbers each client approved for publication. Most are adoption and time figures. Two look forward and are not audited returns: Meridian’s hours are the client’s estimate, and Orion’s payout hours are projected.

Harbourside

A 70-person UK insurtech
Two weeks on-site, with the strategy run in parallel with building two prototypes. Their senior expert moved from spot-checks and answering questions to training the floor.
5% to 100%
of calls reviewed for quality and compliance, up from a monthly spot-check
1,441
answers in a knowledge base that started as a PDF of FAQs
Read the case

Meridian

A 50-person European real-estate firm
Twenty years of one person’s expertise made searchable and multilingual, with personal data stripped before anything else happens.
261
real answers in the knowledge base
7
people on their team trained to build with AI
Up to 528 hrs
saved per trained person per year, by the client’s estimate: 7 to 11 hrs a week x 48 weeks
Read the case

Orion Tech

A tech company of 100+ people, over six weeks
It opened with a readiness assessment and a shadow AI audit. Then strategy, people, governance and engineering ran at the same time.
About 80%
of the organization completed a readiness assessment in week one, across 20+ job functions
3,200+
personalized daily lessons completed
4.4 / 5
pulse score on how the change connects to day-to-day work
About 720 hrs
a year projected from one payout automation
Read the case
Our case studies use pen names. To speak with a client reference, email hello@vibrana.ai.

What the assessment measures, and what it doesn’t

Six areas, 24 multiple-choice questions about what your organization does today (22 for teams of 50 or fewer, who skip two that assume several teams). Next to each area: what we ask about, and how strong the published research behind it is, in our reading of it.

Leadership & Strategy

5 questions
Who is accountable for AI results, how many AI uses are part of daily work, whether results are measured against how the work went before, how many workflows were redesigned around AI, and whether work is timed before it changes.
Research behind it: Strong, replicated across surveys

Data Readiness

4 questions
How much of the information your work needs is digital, how people get data from other teams, how data is checked for errors, and whether each source has a named owner.
Research behind it: Moderate on how common gaps are, thin on effect

Technology Infrastructure

4 questions
An approved AI tool that works with your internal data, whether your systems connect to other tools, where they run, and how fast a working pilot reaches daily use.
Research behind it: Thin to moderate

Talent & Capability

3 questions
The share of staff who can use an approved AI tool, the share with at least five hours of training or practice, and who could build or look after an AI workflow.
Research behind it: Moderate

Governance & Responsible AI

4 questions
A written policy staff know, a review step for AI output, a list of AI tools and uses, and who approves new uses.
Research behind it: Moderate for adoption, thin for value

Culture & Change Readiness

4 questions
The share of staff using AI weekly, who designs changes to the work, whether people talk openly about their AI use, and a plan for people whose work changes.
Research behind it: Moderate

Self-reported

You answer for your organization, about what it does today: owners, counts, written policies, not ratings. Nothing is verified against your systems, so the report is as accurate as the answers.

A profile, not a benchmark

You see your six areas side by side, not a rank against other companies. The overall score is a plain average of the six, each counting the same, because no study we found fits readiness weights to outcomes. Read the areas first.

Research where you’re lowest

For your two lowest areas, the report shows published findings, with who was studied, and a case from our own work.

For context: how other organizations rate themselves

Share of leaders who rate their organization highly prepared for broad AI adoption, by area. Other organizations, not a comparison with yours.
Technology infrastructure
Strategy
Data management
Risk and governance
Talent
Deloitte did not report culture as an area.
Source: Deloitte AI Institute, State of AI in the Enterprise: The untapped edge (2026). Studied: 3,235 director to C-suite respondents in 24 countries, surveyed August to September 2025.

What you get, and what happens to your answers

A question from the AI Readiness Assessment: four answer options, with the six-area profile and a two-axis map filling in as you answer

What you do

  • Give your work email, team size and revenue band. We fill in your company and website from the email domain.
  • Answer 24 multiple-choice questions (22 for teams of 50 or fewer), four options each, with nothing to write out.
  • If you like, name up to three workflows your team repeats, with how many people do them and for how many hours a week.

What you get

  • A score for each area and an overall score, on screen as soon as you finish, so you see your two lowest areas before the next AI budget is set.
  • Published research and a case from our work for your two lowest areas, so you see what others found there before you spend.
  • For each workflow you name, the hours AI could free a year, as a range from published studies of that type of task, and which one to start with.
  • A written analysis that reads your results against your company’s public context, as a PDF and an emailed copy you can forward.

What happens to your data

  • We save your contact details, answers, scores and any workflows you name. Your score comes only from your answers.
  • We read your company’s public website with Firecrawl. To write your analysis, an open-weight AI model gets that page, your company name, your answers and scores, any workflows you name, and signals about how the assessment was completed, such as the time taken. A second model checks the analysis. Both run only on providers that OpenRouter routes with zero data retention and no training on your data.
  • Records are removed when no longer needed. Ask us for your retention date at any time.
  • Up to four follow-up emails about your results over the next two weeks. Each one has an unsubscribe link that cancels the rest, and replying “stop” works too.
  • Email hello@vibrana.ai to access, correct, object to follow-up, or delete your data.
No obligation to buy anything. If you want help afterward, our build sprints start from $10,000. Read the privacy policy

In their words

“The biggest barrier to success is leadership.”
“Adoption tells you that people use AI, not whether it pays off.”
“The biggest wins rarely come from automating one task. They come from rethinking an entire workflow.”
“Impressed and at the same time scared that you have a very similar understanding of what my perception is.”
Leader of a business unit, Orion Tech
“It’s all about coworking.”
An engineer at Orion Tech, after redirecting an AI agent that was going in circles
Our case studies use pen names. To speak with a client reference, email hello@vibrana.ai.

Start with your own profile

Know your two weakest areas before the next AI budget is set.

24 questions. Your profile on screen when you finish, with the research behind your two lowest areas and the hours AI could free in the workflows you name. If it’s useful, book a 30-minute call to walk through it with us.

Take the assessmentBook a 30-min call
A question from the AI Readiness Assessment: four answer options, with the six-area profile and a two-axis map filling in as you answer
Every figure on this page links to its source. Findings come from our research library, checked at source; client numbers are the ones each client approved.