The State of AI+BI Analytics Global 2025 Report PDF Free Download
1 / 46/46
100%
The State of AI+BI
Analytics Global
2025 Report
Report Author:
Brett Sheppard, Dúnedain Research
June 30, 2025
We are witnessing a fundamental shift in how organizations harness their data.
The early days of analytics were marked by what I call Type 1 Intelligence—departmental
insights born from tools like Excel, Tableau, and Google Sheets. These tools gave us our “first
light.” It was a thrilling beginning—data finally being used, questions being asked, answers
starting to emerge. But this light was fragmented. Each department lit up its own corner of the
enterprise, disconnected from the rest. Data lived in silos. Insight didn’t scale. The organization
remained blind to the bigger picture.
Today, the most forward-thinking companies are pushing into Type 2 Intelligence—Enterprise
Intelligence. They’re not just collecting data; they’re connecting it. Across clouds, across
applications, and across business units. They’re building semantic, governed, and open
intelligence layers that represent how their businesses truly operate. These leaders are moving
past pilots and proofs of concept. They’re deploying intelligence in production, at scale, where
it can drive real outcomes.
But this is not the end goal. It’s the foundation for Type 3 Intelligence—Autonomous
Intelligence. This is where AI not only augments human decisions but takes initiative: modeling
data, finding patterns, surfacing insights proactively, and even answering questions from both
structured and unstructured sources. The intelligence becomes active, contextual, and
continuous.
Underlying all of this is a simple but powerful principle: Freedom as a Service. If your
intelligence layer is locked into a database or cloud platform, it’s not really yours. True
intelligence must be portable, open, and sovereign—because your ability to move, scale, and
adapt is what determines your competitive edge.
This report explores where companies stand across this spectrum—and why the next leap
forward starts with connecting what you already have.
Foreword
Saurabh
Abhyankar
Executive Vice President &
Chief Product Officer,
Strategy
Key Findings
As organizations race to unlock the value of their data, AI-powered
analytics is moving from hype to habit. But adoption isn’t just a matter of
flipping a switch—it’s a complex shift in how decisions are made, who gets
to make them, and what infrastructure supports it all. The survey results
reveal how organizations are navigating that transformation—what’s driving
momentum, where friction remains, and how the landscape is evolving.
Who’s Using AI
AI-powered analytics is being used
across executive teams, data
science, IT, ops, finance, and even
customer support.
Why It Matters
Top goals for adopting AI analytics:
better decisions, lower costs, and a
sharper competitive edge.
From Pilot Programs
to Production
43%
are now using AI analytics in production—
not just pilots. Of these, a third are scaling
across departments.
Nearly of organizations
Scaling Up Fast
Everyday Use Cases
What’s Holding AI Back
Organizations focus on the twin use
cases of boosting data expert efficiency
and enabling more employees to
explore data on their own.
High implementation costs, unclear ROI,
poor semantic data layers, and weak
governance structures.
access to AI analytics to at least 30% of
employees within the next year.
of orgs plan to roll out
25%
Trust Gap
Top technical concern: AI tools delivering
inaccurate or inconsistent answers—
undermining user trust at scale.
Understanding AI-Powered Analytics
7
AI Adoption Trends
8
From Limited Access to “Intelligence Everywhere”
9
Adoption Outlook
10
Stages of Adoption
12
Priority Investments in AI-Powered Analytics
13
Adoption of AI Agents and Bots
14
Organizational Models for Analytics
15
Organizational Approaches to Analytics and BI by Size of Organization
17
Case Study: The Warehouse Group. Powering Retail Agility with Data and AI
18
Business Outcomes
19
Business Outcomes by Category
20
Goals for the Next 12–18 Months
22
Departmental Impact
23
Challenges and Risks
24
Confidence in AI-Powered Analytics
26
What Shapes Confidence?
27
Compliance and Organizational Barriers
28
From Pilots to Production: Enabling Enterprise Adoption
29
Empowering Data Experts Through AI-Driven Automation
29
Expanding Access Through Natural Language Interfaces
30
Measuring What Matters: Goals, Services, and Strategic Choices
32
Setting SMART Goals to Improve ROI and Justify Investment
32
Prioritizing AI-Powered Analytics Services
34
Analytics and BI Vendor Landscape
35
Cloud Infrastructure and Vendor Lock-in
37
Total Cost of Ownership (TCO) Considerations
38
Author’s Note: Usability Is the Missing Link
39
Conclusion
40
Appendix: Survey Information
41
Table of Contents
The State of AI+BI Analytics Global 2025 Report 6
Why AI+BI is Different from Generic GenAI
Generic GenAI tools, such as ChatGPT or other large language models (LLMs), can
generate fluent responses but often struggle with math, statistics, or logic. This is due
to how they’re trained—by mimicking patterns in text rather than performing true
calculations.
AI-powered analytics addresses this by splitting responsibilities:
The LLM handles natural language input and explanation.
The analytics platform performs the actual calculations and ensures results are
accurate.
This structure reduces the risk of “hallucinated” answers and increases user trust in the
output.
Understanding AI-Powered Analytics
This report uses the terms “AI-powered analytics,” “AI-powered business intelligence,” “AI
analytics,” and “AI+BI” to describe systems that use AI to:
Accelerate workflows for data experts
Help less technical staff answer data questions
Support decision-making across the organization
The focus is specifically on data platforms that integrate generative AI (GenAI) with enterprise
analytics tools. These platforms allow users to ask data-related questions in natural language.
The GenAI then converts the question into structured queries, which are processed by the
analytics engine. The engine returns the results, and GenAI explains them in plain language.
The State of AI+BI Analytics Global 2025 Report 7
Why This Matters
By combining GenAI with enterprise-grade analytics engines, organizations aim to make data
access more intuitive and inclusive—especially for employees without technical expertise. This
hybrid approach also helps reduce errors in math, statistics, and visual interpretation that are
more common with standalone GenAI tools.
Gartner, Predicts 2025, February 26, 2025
AI-Powered analytics will revolutionize
decision making.”
“
AI Adoption Trends
Organizations are increasingly turning to AI-powered analytics to enhance how they work with
data. The top reason cited in the survey is to improve decision-making, followed closely by
goals related to cost savings and efficiency.
Respondents’ priorities include:
These results show that the push toward AI+BI is not driven by novelty—it’s a strategic
response to performance and efficiency pressures.
The State of AI+BI Analytics Global 2025 Report 8
Improving decision-
making
56.2%
Gaining competitive
advantage
42.6%
Empowering frontline workers
and meeting executive
mandates
each
32.3%
Enhancing operational
efficiency
55.7%
Reducing costs
50.2%
132
118
100
131
76
76
38
30
10
The State of AI+BI Analytics Global 2025 Report
Figure 1
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
26%
30–40%
10%
That’s starting to change.
Survey question:
Note: As of H1 2025, adoption planning is occurring in the context of macroeconomic
uncertainty. Many organizations are adjusting expectations in light of possible market
instability, including recession or higher inflation. This may affect spending on new
technologies, including AI-powered analytics.
The State of AI+BI Analytics Global 2025 Report 10
Expanding Access Beyond Advanced License Holders
Traditionally, only a small group of employees—those with advanced analytics licenses
—have been able to create dashboards, explore data visually, and publish reports. This
limited access kept most users dependent on specialists.
Newer AI+BI platforms, such as Strategy One, now support these capabilities under
more accessible data consumer licenses. These options are designed for everyday
users who need to ask questions, generate charts, and interpret results—without
advanced technical training.
By reducing licensing costs and lowering the skills barrier, organizations can open up
analytics access to more employees and scale data use across teams.
Adoption Outlook
One in four survey respondents (24.3%) say their goal is to give 30% or more of their
workforce access to AI-powered analytics within 12 months—three times the current average.
This signals momentum. While many organizations are still in early stages, the goal of broader
adoption aligns with a growing vision of “intelligence everywhere”—where data is accessible
and actionable across all levels of the organization.
The State of AI+BI Analytics Global 2025 Report 11
Figure 2
Current (top) and Planned (bottom) Adoption of AI-Powered
Analytics
0 10 20 30 40 50 60
Skipped question
Less than 1%
1%–5%
6%–10%
11%–20%
21%–30%
31% or more
13
59
40
53
34
17
19
(5.5% of the total of 235 surveyed)
(25.1%)
(17.0%)
(22.6%)
(14.5%)
(7.2%)
(8.1%)
0 10 20 30 40 50 60
Skipped question
Less than 1%
1%–5%
6%–10%
11%–20%
21%–30%
31% or more
13
10
18
43
49
45
57
Compared to today, 3x
the number of surveyed
organizations would like
to achieve 30% adoption
of AI-powered analytics
by their workforce within
the next 12 months
(5.5% of the total of 235 surveyed)
(4.3%)
(7.7%)
(18.3%)
(20.9%)
(19.1%)
(24.3%)
(24.3%)
Number and percent of responders selecting each choice
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Survey question: What percent of your workforce is actively using AI-powered analytics
today (e.g., asking natural language questions of data or AI automation of dashboard &
report creation)? (Select one)
Survey question: Where would you like your organization to be 12 months from now for
the percent of your workforce actively using AI-powered analytics? (Select one)
The State of AI+BI Analytics Global 2025 Report 12
Stages of Adoption
From Pilots to Production
Organizations are at different stages in their journey with AI-powered analytics. Many have
moved beyond siloed, departmental data use to adopt more integrated practices—such as
enterprise-wide data governance, cross-departmental sharing, and advanced analytics.
The most forward-looking organizations are now evaluating how responsible AI can
proactively drive decisions across the enterprise. While this level of maturity is still rare, it
marks the direction many aim to move toward in the coming years.
Adoption Stages of Enterprise Analytics
Number of
Responses
Percent of
Responses
We employ business intelligence software with advanced
analytics.
138
58.7%
We harness data in departments.
112
47.7%
We standardize and integrate data across departments.
123
52.3%
We have a corporate data strategy and enterprise-wide data
governance.
107
45.5%
Data experts leverage AI to expedite their workflows.
56
23.8%
Our employees can use natural language to interact with data.
47
20.0%
Responsible AI proactively drives data and decisions across
the enterprise.
38
16.2%
Other / None
19
8.1%
Figure 3
Current and Planned Adoption of AI-Powered Analytics
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
A defining insight from the survey is the near-equal pace of adoption between two user
groups: data experts and general business users. AI-powered analytics is helping to automate
workflows of data experts so they spend more time on analysis and sharing insights than on
manual data wrangling. Meanwhile, the accessibility of analytics via natural language queries is
helping organizations expand AI+BI usage beyond traditional “power users.”
Survey question: Where is your organization today in your data journey? (Select all that
apply)
The State of AI+BI Analytics Global 2025 Report
AI-powered analytics
tools
60.9%
Dashboards and self-
service capabilities
51.5%
Data governance and
semantic layers
38.3%
143
121
84 90
58
49
59
42
19
(60.9%) (51.5%) (35.7%) (38.3%) (24.7%) (20.9%) (25.1%) (17.9%) (8.1%)
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Figure 4
Survey question:
McKinsey & Company, The State of AI, March 12, 2025
Organizations’ use of AI has accelerated
markedly in the past year, after years of
little meaningful change.”
“
The State of AI+BI Analytics Global 2025 Report 14
Changing Roles in Analytics Teams
As AI helps automate data discovery and dashboard creation, organizations are
rethinking their hiring priorities. According to McKinsey:
Demand for data visualization and design specialists is declining
Demand for AI data scientists remains strong
In the 2025 McKinsey survey, half of organizations using AI said they expect to hire
more data scientists this year to support expanding AI capabilities. This shift suggests
a growing emphasis on analytical depth over visual presentation in AI-driven
environments.
Adoption of AI Agents and Bots
Most organizations are either evaluating or piloting the use of AI agents or bots as part of their
analytics strategy. In the context of AI-powered analytics:
This distribution suggests that while few organizations have scaled AI agents across multiple
functions, a combined 70% are either already deploying or preparing to deploy them.
Adoption is still in its early stages but moving steadily toward broader production use and
integration with existing applications and systems.
An AI agent is an intelligent system that
can analyze data, generate insights,
recommend actions, and—in some
cases—automatically execute decisions
aligned with business goals and KPIs.
An AI bot is designed to respond to user
questions, helping non-experts quickly
find and understand data relevant to
their daily work.
The State of AI+BI Analytics Global 2025 Report 15
Figure 5
Adoption of AI Agents or Bots
Survey question: To what degree is your organization creating and deploying AI agents or
bots that serve specific functions? (Select the answer that best applies to your
organization)
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Organizational Models for Analytics
Organizations take different architectural approaches to deploying analytics and business
intelligence (BI), based on their size, structure, and internal capabilities.
The most common models include:
39.6%
Hybrid “hub-and-spoke”
In this setup, a centralized expert
team (often a center of
excellence) supports analysts
embedded in business units, who
maintain some flexibility to lead
their own initiatives.
30.2%
Centralized
A corporate sponsor
or a centralized
team owns and
manages many or
all BI capabilities.
19.6%
Decentralized
Business units
independently
manage their
own tools and
processes.
We do that well today,
with multiple different Al
agents or bots in
production in several
different departments or
lines of business, and
more on the way this year.
Number of
Responses
Percent of
Responses
We have deployed
one or more Al
agents or bots for
specific business
functions.
We are piloting Al
agents or bots
and plan to move
into production
this year.
We are evaluating
options for Al
agents or bots
without a clear
timetable for
production.
None of
the above Skipped
question
24
10,2%
60
25,5%
48
20,4%
48
20,4%
13
5,5%
13
5,5%
The State of AI+BI Analytics Global 2025 Report 16
A small number of respondents (3.8%) selected “Service provider,” wherein a consultancy or a
managed service provider leads the BI function, and 5.1% selected “Other,” with one comment
noting a “disparate/shadow IT” model where overlapping roles between IT and business units
lead to inefficiencies.
Organizational Approaches to Analytics and BI
Number of
Responses
Percent of
Responses
Centralized – Core capabilities organized by a corporate
sponsor or center of excellence for use organization-wide
71
30.2%
Decentralized – Stakeholders & analysts working in business
units & departments have the flexibility to experiment and
adopt the analytics & business intelligence which best fit their
group’s unique needs
46
19.6%
Hybrid (“Hub and spoke”) – Combination of an expert
centralized function with analysts distributed to lines of
business and departments with some flexibility to undertake
their own pilot projects
93
39.6%
Service provider – We have hired a consultancy service
provider to take the lead
9
3.8%
None / Other / Skipped
16
6.8%
Total
235
100.0%
Figure 6
Approaches to Analytics and BI
Survey question: How would you describe your organizational approach for analytics and
business intelligence? (Select one)
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
13
8
10
8
6
19
8
13
5
3 3
8
25
9
14 14
8
17
1 1
7
2
1
2
1 1
8
The State of AI+BI Analytics Global 2025 Report
72 (30.6%)
46 (19.6%)
39 (39.6%)
9 (3.8%) 15 (6.4%)
Mid-size organizations
decentralizedhybrid
Large organizations
centralizedhybrid
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Figure 7
Survey questions:
The State of AI+BI Analytics Global 2025 Report 18
Case Study
The Warehouse Group: Powering Retail Agility with Data and AI
Amid economic headwinds, The Warehouse Group (TWG), New Zealand’s largest non-food
retailer, turned to data and AI to stay resilient and drive growth. Facing reduced consumer
spending and traditional organizational silos, TWG unified its data strategy using the
Strategy platform to enable self-service, governed insights across every function—from
merchandising and supply chain to strategic planning and in-store experience.
By democratizing data access and empowering teams with AI tools, TWG created a more
agile, insight-driven business culture. A key success factor was the creation of a Data
Domain team, which bridged skill gaps and ensured cost-effective delivery of analytics
across the enterprise. The addition of Strategy AI allowed users to accelerate decision-
making and explore generative AI in practical, impactful ways—from prompt tuning to
chatbot deployment.
TWG’s transformation offers a blueprint for other retailers navigating similar challenges. By
embedding trust, governance, and curiosity into its data culture, the company not only
broke down technical and organizational barriers but also laid the groundwork for
continuous innovation.
Read the case study to learn more >
Keryn McKenzie, Chapter Area Lead, Data, Insights & Services
Data is the lifeblood of our business—it’s
the oxygen fueling everything we do.
Strategy democratizes our data in a
safe, governed, consistent, and well-
managed way.”
“
The State of AI+BI Analytics Global 2025 Report 19
Business Outcomes
Competitive advantage, cost savings, and better decision-making
Organizations are beginning to see tangible business results from AI-powered analytics.
61.7%
of respondents see a “moderate positive
impact” or “significant positive impact” of AI-
powered analytics on their current or expected
business outcomes.
Only two in 235
respondents chose the
option “Negative impact,
e.g., we’ve wasted time
and money”. Additionally,
a combined 32.7% see “no
impact yet” or “minimal
impact”.
Figure 8
Business Outcome Sentiment
Survey question: How would you rate the impact of AI-powered analytics on your
business outcomes? (Select one)
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
1 2 3 4 5 6
0
20
40
60
80
100
Number of responses
2
11
36
41
100
45
0.9%
4.7%
17.5%
42.5%
19.2.%
15.2.%
Negative
impact Not Applicable
(e.g., not planned) Minimal impact Moderate
positive impact Significant
positive impact
No impact
yet
At healthcare conglomerate Johnson & Johnson, CIO Jim Swanson and his team
evaluated 900 internal use cases for GenAI, AI-powered data science, and intelligent
automation. They found that between 10% and 15% of the use cases were driving about
80% of the total economic value. They transitioned from experimentation with a
“thousand flowers” of bottom-up experimentation to prioritize investments on the most
promising use cases. (The Wall Street Journal, Isabelle Bousquette, “Johnson & Johnson
Pivots its AI Strategy,” April 18, 2025).
The State of AI+BI Analytics Global 2025 Report 20
These results reflect a familiar adoption curve—from late adopters starting to see positive
results, to mainstream users experiencing moderate gains, to early adopters reporting
significant benefits. The trend suggests the possibility of a clear, predictable path to value for
organizations investing in AI-powered analytics, with the understanding that emerging
technologies will inevitably go from inflated expectations to what Gartner terms the “trough of
disillusionment” while working through the challenges before organizations experience steady
rise in consistent positive business outcomes.
Business Outcomes by Category
Productivity, cost savings, and faster decision-making lead the list
AI-powered analytics is delivering a broad range of benefits across organizations. The most
commonly reported outcomes include:
Increased employee
productivity
53.6%
Improved customer
satisfaction
46.8%
Competitive
advantage
42.6%
Cost
savings
48.9%
Faster decision-making
and innovation
48.1%
Cost-related outcomes also show up in areas such as operational efficiency, supply chain
optimization, and profitability—suggesting that organizations see savings not just in direct
expenses, but also through better resource allocation.
The State of AI+BI Analytics Global 2025 Report
Figure 9
Survey question:
53.6%
48.9%
48.1%
46.8%
42.6%
42.1%
35.7%
22.1%
21.3%
19.1%
19.1%
15.7%
15.7%
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Deloitte
Jason Girzadas, CEO, Deloitte, Fortune, January 29, 2025
There is a massive recognition of the need to train workforces, to
increase literacy around all things tech, all things AI. Within Deloitte
we talk about it in terms of tech literacy across our enterprise…
These are new tools, these are new technologies, and they’re
changing the way people do their work...”
The Role of Training in Adoption
“
Cost
savings
46.4%
Competitive
advantage
41.7%
The State of AI+BI Analytics Global 2025 Report 22
Goals for the Next 12–18 Months
From pilot projects to production priorities
To justify scaling from pilots to production, organizations are aligning their AI-powered
analytics strategies with high-impact business goals. According to survey respondents, the
top two priorities are:
These are followed closely by a cluster of priorities focused on decision-making, efficiency,
and customer experience—each selected by over one-third of respondents.
Figure 10
AI-Powered Analytics Goals: Next 12–18 Months
Survey question: What are your organization’s most important AI-powered analytics goals
for the next 12 to 18 months? (Select all that apply)
0% 10% 20% 30% 40% 50%
1
2
3
4
5
6
7
8
9
10
11
12
13
14
46.4%
41.7%
37.0%
36.6%
34.5%
34.0%
27.7%
27.7%
26.8%
26.4%
26.0%
25.1%
19.1%
13.2%
Cost savings
Competitive advantage
Improving decision-making
at all levels
Operational efficiency and
resource allocation
Enhancing customer
experience
Faster time to market
Revenue growth via
marketing and pricing
Strategic planning and
future trend prediction
Personalized learning and
development plans for employees
Reducing inefficiencies
Proactive risk management
Supply chain optimization
Improving profitability
Faster innovation
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Cost Considerations Under Uncertainty
As of H1 2025, many organizations are revising priorities due to macroeconomic
uncertainty. In this environment, new technologies like AI-powered analytics—often
funded by discretionary budgets or executive mandates—face added scrutiny.
Without a dedicated budget line item, these initiatives must clearly demonstrate value.
It’s no surprise, then, that “cost savings” has emerged as the top goal for the next 12–18
months.
The State of AI+BI Analytics Global 2025 Report 23
This data reflects a maturing approach to analytics investment. Organizations are looking
beyond experimental use cases and toward outcomes that directly affect cost structure,
strategic agility, and long-term competitiveness.
Departmental Impact
Widespread business value across technical and operational teams
AI-powered analytics is delivering measurable value across a wide range of business functions.
In the 2025 survey, the top beneficiaries identified by respondents include:
Data science
Executive leadership
IT / Technology
Finance and operations
Customer support
46.0%
42.1%
45.5%
41.7%
45.1%
Other departments such as marketing, human resources, and frontline teams also report
benefits. Open-text responses additionally mentioned governance, risk, compliance, audit,
and sales as departments benefiting from AI-powered analytics.
One respondent noted:
“Most things need to be quality assured (QA): this is still time consuming, as the QA
feeding back into the AI model in an automated way is still a challenge.”
The State of AI+BI Analytics Global 2025 Report 24
0% 10% 20% 30% 40% 50%
1
2
3
4
5
6
7
8
9
10
11
12
13
46.0%
45.5%
45.1%
42.1%
41.7%
41.7%
31.5%
24.7%
21.3%
17.4%
15.3%
2.6%
1.7%
Data Science
Executive Team
Human Resources
IT / Technology
Marketing
Finance
Supply Chain / Logistics /
Warehousing
Customer Support
Frontline Workers
Operations
R&D
None of the above / Other
Skipped question
Figure 11
Departments Seeing the Greatest Business Impact
Survey question: Which departments in your organization see (or will see) a significant
business impact from leveraging AI-powered analytics? (Select all that apply)
Challenges and Risks
Accuracy and implementation costs remain top concerns
While adoption is growing, organizations continue to face significant barriers in achieving
reliable and scalable outcomes with AI-powered analytics.
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
The most cited technical and operational challenge is inaccurate or
inconsistent answers, reported by 43.4% of respondents. These so-
called “hallucinations” undermine trust in AI-generated insights and often
require time-consuming quality assurance (QA) reviews.
The State of AI+BI Analytics Global 2025 Report 25
Cost-related concerns also ranked high. 41.7% of respondents said the
cost of implementation without a predictable return on investment is a
major issue.
Figure 12
Top Challenges in AI-Powered Analytics
Survey question: From technical and operational perspectives, which of the following
present problems for your AI-powered analytics? (Select all that apply)
0% 10% 20% 30% 40%
1
2
3
4
5
6
7
8
9
43.4%
41.7%
31.2%
21.3%
20.0%
18.3%
14.9%
8.1%
5.5%
Inaccurate or inconsistent answers
(“hallucinations”)
High implementation costs with
no clear ROI
Lack of an internal AI analytics
competency center or center of
excellence
Open-loop systems requiring
manual handoffs between tools
Inability to retain previous learning,
causing inconsistent responses
Unresponsive or resource-heavy
workloads
Inflexible infrastructure (e.g., no
containerized cloud scaling)
Other (please specify)
Skipped question
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
The State of AI+BI Analytics Global 2025 Report
Most organizations express trust—with caution
30.2%
very confident
extremely
confident.
35.7%
confident
5.5%
no
confidence
Figure 13
Survey question:
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
The State of AI+BI Analytics Global 2025 Report 27
The accuracy and
cleanliness of the data
The type of AI
model used
The nature of the
question being asked
What Shapes Confidence?
Several respondents added comments qualifying their confidence, noting that it depends on:
Many organizations—particularly in IT-intensive industries—take a “trust but verify”
approach. Confidence is earned over time through real-world use, especially in complex
environments using tools from multiple vendors.
This is reinforced by another data point: 23.0% of respondents cited “unrealistic expectations
from vendor demos” as a key challenge to adoption. Discrepancies between polished demos
and production realities can erode trust quickly—sometimes permanently. On the plus side,
76.6% of respondents did not cite this challenge, suggesting a more informed and grounded
approach to AI adoption.
Accenture Technology Vision 2025, January 7, 2025
People’s trust in AI—beyond any technical aspect, that it performs
justly and as expected—is essential to it having as broad and positive
an impact as anticipated.“
Why Trust Matters
“
The State of AI+BI Analytics Global 2025 Report 28
Compliance and Organizational Barriers
Regulatory readiness is now the top challenge to adoption
Compliance has overtaken cost as the most commonly cited challenge to enterprise adoption
of AI-powered analytics. When asked about the main challenges in the 2025 survey:
52.0%
of respondents reported
concerns about
regulatory risk, including
bias in AI models, data
privacy requirements,
and industry-specific
obligations.
49.0%
31.0%
cited cost-related
issues, including Total
Cost of Ownership
were concerned with
confusion around GenAI
vs. analytics-specific AI
41.5%
pointed to integration
challenges with existing
tools and systems.
0% 10% 20% 30% 40%
1
2
3
4
5
6
7
8
9
10
11
12
13
52.0%
49.0%
41.5%
31.0%
28.0%
24.5%
24.5%
22.5%
16.5%
12.0%
11.0%
10.5%
2.0%
Compliance (e.g., bias, data privacy, regulations)
Cost (e.g., high expense, unclear ROI)
Integrations (e.g., not embedded in current
systems)
Lack of corporate data strategy
Unrealistic expectations (e.g., demos vs.
production complexity)
Technology immaturity (e.g., missing enterprise-
grade features)
Point tools vs. unified platform
Vendor lock-in
Tech stack incompatibility
Limited availability on frontline devices
Other
Trust issues (e.g., hallucinations, inconsistent
results, lack of explainability)
GenAI confusion (e.g., unclear difference from
general-purpose models)
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Figure 14
Top Challenges to Adoption
Survey question: Which of these challenges has impacted or may impact your adoption of
AI-powered analytics across the organization? (Select all that apply)
Guidance: Start With the NIST AI Risk Management Framework
A practical step toward improving compliance is aligning internal processes and vendor
requirements with the NIST AI Risk Management Framework (AI RMF). This
framework, introduced by NIST in 2023, provides structured guidance for deploying AI
systems responsibly, including:
Governance and oversight
Risk identification and mitigation
Transparency and accountability
NIST also hosts a Trustworthy and Responsible AI Resource Center, which offers
real-world examples of how other organizations are applying the framework to meet
evolving regulatory expectations—including the EU AI Act.
The State of AI+BI Analytics Global 2025 Report 29
From Pilots to Production: Enabling Enterprise Adoption
Empowering Data Experts Through AI-Driven Automation
Among early adopters, of organizations report using AI-powered analytics
13.6%
to support data experts across multiple use cases and departments. These teams are
leveraging AI to speed up analysis, reduce manual effort, and improve the accuracy and
timeliness of insights.
Respondents also highlighted adoption nuances in open-text comments, citing:
A focus on pilots, not yet extending to analytics
Ongoing issues with data pipelines
Use of AI-powered analytics in R&D settings to develop products and services
These insights suggest that while many organizations are still in early stages, some are already
exploring how internal tools might evolve into customer-facing offerings.
AI to Reduce Analyst Workload at Ipsos
France-headquartered opinion poll provider Ipsos worked with Gemini 1.5 Pro & Flash to
build a self-service AI-powered data analysis tool for its teams of market researchers.
This tool reduces the burden on internal data analysts to answer time-consuming
requests. At the same time it creates monetization opportunities for Ipsos Data Labs
value-added services to upsell clients of the company’s market research.
Brett Sheppard, Five Ways to Elevate Analytics with
Generative AI (2024)
Asking questions is becoming how many of
us at work access and understand data.”
“
The State of AI+BI Analytics Global 2025 Report 30
Figure 15
Expediting Workflows for Data Experts
Survey question: Has your organization adopted AI-powered analytics to expedite the
workflows of data experts? (Select one)
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Yes, partially operational
for single use cases and
individual departments
28.9%
Yes, fully operational for
multiple use cases and
departments
13.6%
Considering adoption
within the next 12 months
20.0%
No plans to adopt in the
next 12 months
8.9%
None / Other
3.8%
In the pilot / testing phase
24.7%
Expanding Access Through Natural Language Interfaces
Supporting Non-Technical Staff With AI-Powered Queries
Organizations are increasingly using AI to help non-technical employees interact with data
using natural language. This approach lowers the barrier to analytics by allowing users to ask
questions conversationally, rather than relying on dashboards or technical reports.
The State of AI+BI Analytics Global 2025 Report 31
Figure 16
Adoption of Natural Language Interfaces
Survey question: Has your organization adopted AI-powered analytics to assist non-
technical staff in answering data questions? (Select one)
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Fully operational for
multiple use cases and
departments
15.7%
In pilot / testing phase
20.9%
No plans to adopt in the
next 12 months
12.3%
Partially operational for
single use cases and
individual departments
26.8%
Considering adoption
within the next 12 months
24.3%
This trend is growing in parallel with AI workflow automation for data experts. In fact, adoption
levels for both use cases are tracking closely—suggesting that organizations recognize the
value of expanding access across both technical and non-technical roles simultaneously.
The even uptake of natural language functionality and expert-facing automation is notable.
Traditionally, organizations focus on one area first—but in this case, many are implementing
both to support a broader set of users and use cases. This parallel investment reflects the
technology’s potential to improve decision-making, accessibility, and adoption across entire
organizations.
The State of AI+BI Analytics Global 2025 Report 32
Measuring What Matters: Goals, Services, and Strategic
Choices
Setting SMART Goals to Improve ROI and Justify Investment
For many organizations, the second biggest obstacle to adopting AI-powered analytics is the
perceived high cost of implementation without a predictable return.
Additional open-text responses pointed to:
Poor data quality
Blurred or undefined data boundaries
Industry-wide resistance to adopting new technologies in their specific market segment
These concerns underscore the need to link AI initiatives to clearly defined business
outcomes. One effective strategy is to use SMART goals—specific, measurable, achievable,
relevant, and time-bound—to evaluate success and guide decision-making.
Below are examples of SMART goals based on the business outcomes most frequently cited by
survey respondents.
of respondents cited this as a primary business concern.
42.0%
Outcome
% of
Respondents
Example SMART Goal
Employee productivity
53.6%
Increase productivity in customer support by 15%
within 6 months by automating ticket triaging and
enabling AI-generated responses.
Cost savings
48.9%
Achieve $50 million in annual savings in supply
chain operations by eliminating waste using AI
analytics by FY2025.
Faster decision-making /
innovation
48.1%
Cut strategic decision time in product planning by
30% by Q4 2025 through real-time predictive
dashboards.
Customer satisfaction
46.8%
Improve Net Promoter Score by 10 points in 9
months via AI-personalized customer engagement.
Competitive advantage
45.5%
Launch two new AI-augmented features with 10%
greater accuracy than competitor benchmarks by
December 2025.
Operational efficiency
42.1%
Boost manufacturing throughput by 5% by Q3 2025
using predictive maintenance and process
optimization tools.
Drive analytics adoption
35.7%
Increase usage of AI analytics tools across business
units from 35% to 75% by end of 2025 through
training and incentives.
Higher profitability
22.1%
Improve gross profit margin by 5% within 12 months
using AI-powered pricing and customer
segmentation.
Risk management
21.3%
Cut fraud detection time by 40% in 6 months
through deployment of anomaly detection models.
Time to market
19.1%
Reduce product lifecycle duration by 25% over 12
months using AI to accelerate user testing and
decisions.
Revenue growth
19.1%
Boost quarterly revenue by 10% by Q4 2025 using
AI analytics to identify upsell opportunities.
Supply chain optimization
15.7%
Improve forecast accuracy from 70% to 90% in 9
months by integrating real-time data across
procurement and logistics.
Improve mission readiness
15.7%
Raise readiness scores by 20% in 12 months by
deploying logistics analytics for defense operations.
The State of AI+BI Analytics Global 2025 Report 33
Source: Dúnedain Research, AI-Powered Analytics Survey 2025, Sponsored by Strategy
The State of AI+BI Analytics Global 2025 Report
Organizations are investing across a wide range of AI capabilities
no clear outlier at the top or bottom
multiple services in parallel
Figure 17
Survey question:
9.4%
22
7.7%
18
11.9%
28
17.0%
40 17.9%
42 25.5%
60 14.5%
34
25.1%
59
20.9%
49 31.5%
74 22.5%
53
34.5%
81 19.6%
46
19.2%
45
11.5%
27 26.0%
61 31.5%
74
37.0%
87 19.6%
46
17.9%
42
12.3%
29 22.1%
52 33.1%
78 15.3%
36
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
The State of AI+BI Analytics Global 2025 Report 35
While data discovery automation and decision support lead in high-priority rankings, even
more niche capabilities such as pixel-perfect reporting retain meaningful support. These
results reflect the experimental nature of the current adoption cycle, where organizations are
still defining which capabilities best align with their workflows and data strategies.
Analytics and BI Vendor Landscape
Diverse platforms and dual-provider strategies shape enterprise adoption
Survey respondents reported using more than 20 different vendors for analytics and BI — a
reflection of the fragmented, competitive landscape in this space.
The two most widely used solutions are:
These leading platforms are followed closely by providers such as:
No independent marketing share data exists yet for AI+BI. The results here may skew toward
Strategy given a possible greater inclination for existing Strategy customers to participate in a
survey branded “Sponsored by Strategy.”
Despite the growing integration of AI into traditional BI tools, many organizations still manage
multiple platforms in parallel — often a mix of legacy solutions and cloud-native services.
Microsoft Power BI with AI or Azure
Machine Learning
Strategy One
(formerly MicroStrategy ONE)
50.2%
27.2% 25.1% 19.6% 18.3%
27.2%
Vendor Diversity
Most organizations in this survey reported working with two to three vendors. In larger
enterprises, that number may be higher due to shadow IT purchases within
departments and lines of business.
The State of AI+BI Analytics Global 2025 Report 36
Figure 18
BI and Analytics Platforms Used by Survey Respondents
Survey question: What analytics or business intelligence software does your organization
use today? (Select all that apply)
0% 10% 20% 30% 40% 50%
1
2
3
4
5
6
7
8
9
50.2%
27.2%
27.2%
25.1%
19.6%
18.3%
12.3%
10.0%
11.1%
<10.0% each
Microsoft Power BI /
Azure ML
Strategy One (formerly
MicroStrategy)
Google (Looker, Sheets,
Vertex AI)
AWS (QuickSight,
SageMaker, AI/ML)
Salesforce / Tableau
SAP (BusinessObjects,
Analytics Cloud)
IBM Cognos / Watson
Others (e.g., Qlik, Oracle,
SAS, Zoho, TIBCO)
None of the above /
Skipped question
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
In Gartner’s most recent Magic Quadrant for Analytics and BI (June 2024), the report noted
ongoing challenges for Microsoft Power BI customers related to decentralized usage and lack
of governance, particularly in light of new Copilot features. This concern was echoed by
survey respondents in open comments, with several citing the absence of a semantic data
layer as a growing pain point.
While independent market sizing for AI-powered analytics remains limited, user surveys such
as this provide insight into how buyers perceive and use analytics platforms—including the use
of spreadsheet tools like Google Sheets as part of their BI stack.
The average organization in this survey works with between two and three analytics
and BI vendors. The actual figure for many large organizations is likely higher through
“shadow IT” departmental and line of business purchases.
The State of AI+BI Analytics Global 2025 Report 37
Cloud Infrastructure and Vendor Lock-in
Balancing integration, flexibility, and clarity in cloud-first environments
The 2025 survey shows that most organizations operate in multi-cloud environments,
averaging use of 1.5 hyperscale cloud providers alongside Software as a Service (SaaS)
platforms for CRM, ERP, and other enterprise applications.
Additional responses referenced private clouds, regional vendors (especially in China and
Germany), and services such as Snowflake and Render.
50.6% 38.3% 27.2%
Multi-Cloud is the Norm
On average, respondents use two to three analytics platforms and at least one to two
cloud providers. For large enterprises, the actual count is often higher due to
decentralized procurement and “shadow IT.”
Figure 19
Cloud Service Providers Used
Survey question: What cloud service providers does your organization currently use?
(Select all that apply)
0% 10% 20% 30% 40% 50%
1
2
3
4
5
6
7
8
9
50.6%
38.3%
27.2%
12.8%
11.9%
3.8%
0.4%
12.3%
1.7%
Microsoft Azure
AWS
Google Cloud
IBM Cloud
Oracle Cloud
Alibaba Cloud
OVHcloud
Other / Not listed
Skipped question
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Clarifying GenAI vs. AI-powered Analytics
In our survey, many respondents cited confusion between general-purpose GenAI (like
Microsoft Copilot or ChatGPT) and purpose-built AI-powered analytics as a barrier to
adoption. Over one-third (34.7%) selected this as a challenge, highlighting a need for
clearer education on use cases and tools.
“What were 2025 sales in French-speaking countries?” is a data-specific query best
handled by AI-powered analytics.
“List all countries where French is a primary language” is a general knowledge
question suited to GenAI.
The State of AI+BI Analytics Global 2025 Report 38
Despite growing cloud usage, vendor lock-in is not yet a major concern: only 13.6% of
respondents flagged it as a challenge. However, this may become more relevant as AI-
powered analytics platforms expand and organizations aim to preserve flexibility across multi-
cloud setups.
Total Cost of Ownership (TCO) Considerations
Pricing changes from major vendors like Microsoft have added
complexity. Power BI Pro and Premium pricing increased 25–40% in
April 2025, with bundled offerings like Fabric F64 presenting steep
three-year commitments.
While these bundles may offer value for highly integrated organizations, they are less suited to
firms that favor:
Best-of-breed architecture
Multi-cloud flexibility
Granular cost control
Analysts warn that many organizations fail to fully account for cloud compute costs linked to
increased Power BI usage. As one survey respondent noted, “Our finance team dreads
reviewing monthly cloud bills. Surprises are frequent — and expensive.”
This makes clear, predictable pricing models an emerging priority, especially as more
organizations scale AI-powered tools from pilot to production.
The State of AI+BI Analytics Global 2025 Report 39
Author’s Note: Usability Is the Missing Link
Despite increasing investment in AI-powered analytics, fewer than 10% of employees in most
organizations currently use tools more sophisticated than spreadsheets. While backend
components like semantic layers, data governance, and AI model orchestration continue to
evolve, true adoption hinges on something simpler: usability.
If the software is difficult to set up, learn, or navigate, it won’t matter how powerful the features
are. This is especially true for non-specialist users such as frontline employees, who often lack
time for formal training or the technical background to troubleshoot data formatting errors or
model behavior.
For AI-powered analytics to scale meaningfully, systems must:
In short, tools must meet users where they are, not expect users to become data engineers.
The organizations that succeed in making analytics accessible—not just available—will be the
ones that unlock its full business value at scale.
Guide users
intuitively through
core workflows.
Tolerate imperfect
inputs (like
spreadsheets with
extra headers).
Provide just-in-time
feedback and
embedded help.
Eliminate friction for
casual or first-time
users.
The next wave of enterprise analytics adoption will not be driven by
technical capability—it will be driven by user experience.
The State of AI+BI Analytics Global 2025 Report 40
Conclusion
AI-powered analytics is transitioning from a promising concept to an operational reality across
a growing number of organizations. This survey of 235 global respondents reveals that
adoption is no longer confined to isolated pilots. A significant portion of organizations are now
implementing AI-powered analytics to drive decision-making, reduce costs, and improve
operational efficiency.
At the same time, this shift comes with clear challenges. Concerns about compliance,
integration, and unpredictable ROI remain key barriers to broader adoption. The most widely
cited technical issue—generating incorrect or inconsistent results—underscores the
importance of trust and transparency in AI-assisted systems. Organizations are beginning to
address these concerns through improved governance frameworks, quality assurance
processes, and data literacy efforts.
A defining insight from the survey is the near-equal pace of adoption between two user
groups: data experts and generalist business users. The accessibility of analytics via natural
language queries is helping organizations expand usage beyond traditional “power users,”
bringing analytics closer to everyday decision-makers across departments.
But technology alone is not enough. Usability, training, and change management remain central
to realizing the full value of AI-powered analytics. Whether it’s a frontline employee uploading
store data or a finance director interpreting predictive insights, the software must be intuitive,
forgiving of imperfect inputs, and aligned with real-world workflows.
Looking ahead, successful adoption will require balancing innovation with clarity—choosing
tools that not only scale, but are also trusted and easy to use. With the right strategies in place,
AI-powered analytics can become the most widely adopted enterprise analytics model since
the rise of spreadsheets—unlocking a new era of data-informed decisions across the modern
workforce.
The State of AI+BI Analytics Global 2025 Report 41
Appendix: Survey Information
In H1 2025, the Strategy team surveyed leaders and practitioners in data, analytics, and
business intelligence across 235 organizations in 38 countries. Participants included
enterprises, governments, and nonprofits. Each respondent answered on behalf of their
organization. This research was commissioned by Strategy and conducted by Brett Sheppard
at Dúnedain Research, lead researcher and report author. All responses are anonymous and
not individually attributed.
To join future surveys, visit: strategysoftware.com/survey.
Organization Size
The 2025 survey reflects a balanced distribution across organization sizes, including:
Small organizations (<500 employees): 25.1%
Midsize organizations and larger (1,000+ employees): 66.8%
Very large organizations (20,000+ employees): 20.9%
Figure 20
Number of employees
Question: How many total employees does your organization have worldwide? (Select
one)
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
1 2 3 4 5 6 7
0
10
20
30
40
50
60
Number of responses
49
20
31 31
26
19
59
20.9%
8.5%
13.2% 13.2%
11.1%
8.1%
25.1%
20,000 or
more
employees
10,000 to
19,999 5,000 to
9,999 2,500 to
4,999 1,000 to
2,499 500 to 999 Fewer than
500
employees
The State of AI+BI Analytics Global 2025 Report 42
Roles in Evaluation and Deployment
Respondents often play multiple roles in the analytics adoption lifecycle. Both business and
technical stakeholders are well represented. Roles are defined below as they appeared in the
survey.
Functional Roles in Analytics Use
This question focused on how respondents work with analytics as part of their day-to-day
jobs. Nearly 40% manage teams of data professionals, while over one-third define
organizational needs for analytics data.
Figure 21
Figure 22
Roles in Analytics Adoption
Question: What is your role in the evaluation and purchase of enterprise analytics? (Select
all that apply)
Functional Roles
Question: What is your functional role in the use of enterprise analytics? (Select all that
apply)
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
Business
leader Department
head Data
scientist Data
practitioner Data
engineer Technical
evaluator Business
stakeholder Other
1 2 3 4 5 6 7 8
0
20
40
60
80
100
Number of responses
85
92
39
61
54
65
72
8
36.2%
39.1%
16.6%
26.0%
23.0%
27.7% 30.6%
3.4%
1 2 3 4 5
0
20
40
60
80
100
120
Number of responses
81
112 118
91
25
Economic
Buyer Sets
Requirements Technical
Evaluator Deployment Other
34.5%
47.7% 50.2%
38.7%
10.6%
The State of AI+BI Analytics Global 2025 Report
technology retail consulting banking
healthcare Public sector
12.3%
Figure 23
Question:
1.7%
1.3%
7.2%
0.9%
2.1%
2.6%
8.5%
2.1%
5.5%
5.1%
1.7%
5.5%
0.9%
6.4%
1.3%
3.8%
6.4%
1.3%
2.1%
8.9%
1.3%
12.3%
2.6%
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
The State of AI+BI Analytics Global 2025 Report
38 United States Germany the UK Australia
SpainNorth America43.8%Europe
36.2%
Figure 24
3.4%
4.3%
19.6%
4.3%
5.5%
40.9%
Source: Dúnedain Research, The State of AI+BI Analytics Global 2025 Report, Sponsored by Strategy
About the Research
This survey was commissioned by Strategy and conducted by Brett
Sheppard at Dúnedain Research as lead researcher and report author. Brett
Sheppard’s journey with data began as a U.S. military data engineer and
Gartner senior analyst. He has authored AI, data architecture and business
intelligence publications for Gartner, GigaOM and O’Reilly, and has been
quoted by Businessweek, Computerworld, Wired and other media. Contact
the author at dunedainresearch.com or linkedin.com/in/brettsheppard.
Participate in Future Surveys
Be part of shaping the future of AI-powered analytics.
To contribute to upcoming research:
Visit: strategysoftware.com/survey
Strategy Incorporated, 1850 Towers Crescent Plaza, Tysons Corner, VA 22182.
Copyright ©2025 . All Rights Reserved.
strategysoftware.com
