CASE STUDY · INFORMATION SERVICES
At EBSCOlearning, AI tool use rose 77% and every non-user started
A division of EBSCO Information Services trained people in every function with Mindstone, each starting from their own work.
The result
33%
became daily AI users
people in the Enterprise AI Academy, after the programme
-
19 hours a week
freed for a two-person marketing team
with assistants one marketer set up
-
4.1 / 5
average readiness to use AI at work
people's own score at the end of the programme
-
3,000+
employees in the parent company, EBSCO Information Services
Company figure.
The challenge
One programme for very different jobs
EBSCOlearning provides learning, education and career development products to businesses, schools and workforce development organisations. As AI tools spread into everyday work, its leadership decided that knowing how to use them was no longer optional for anyone, not only the product team.
The path was harder than the goal. People do very different jobs, from engineering to product to marketing, and generic training would not fit them all. Any programme had to suit each person's role and EBSCOlearning's own business, and it had to show results quickly. The team looked at several options before choosing one to pilot.
We needed to level up our teams' understanding of AI tools, their capabilities, strengths and limitations to identify the right tools for specific use cases and inform our product and business strategy.
Michael LaddinSenior Vice President & General Manager, EBSCOlearning
What they did
Learn on your own work, then keep showing each other
EBSCOlearning ran Mindstone's Enterprise AI Academy from the bottom up: each person looked for the jobs in their own role where AI could help.
-
Step 1: Own work
Each person starts from tasks in their own role, not a set list.
-
Step 2: Platform
Training shaped to the role and the company, with instant feedback on what they ask the AI.
-
Step 3: Live help
Sessions with Mindstone experts on each team's specific problems.
-
Step 4: Meetups
Teams keep meeting to show what they have built.
The first result
A software architect's build cycle, before and after
Before
Days of work
Turning a product idea into a working version to test took Kishore Karavadi about 31 hours over two to three days.
After
About an hour
He now uses AI to write the feature descriptions, the specification and the application, so a first version to test is out within a day.
What changed, person by person
Different jobs, different gains
-
Product management
Maria Rioux Nagy set up reusable AI assistants for the constant switching between products, partner projects and teams.
-
Marketing
Maggie Woodroof built assistants for content, email and data work, connected to HubSpot and Google Sheets. Team capacity used fell from 85% to 61%.
-
Engineering
Ashley Garcia-Arellano set up an assistant to stop stray processes on her machine, saving 90 minutes a week.
Where it went next
Nobody was told what to automate
Each person found a different job to hand to AI, so the gains looked different in every role. Some cut the time of a single task. Others freed hours across the week.
The internal meetups keep it going. They act as a regular check-in, keep AI on everyone's agenda and show each team what the others are doing, so ideas travel. Colleagues have started using what each other build, too.
With all the context shifting I do - overseeing multiple products, partner projects, and new innovations - using AI efficiently has been a game changer. What I've learned through the program has allowed me to save at least 10 hours worth of work a week. A colleague even asked me, 'I don't know how you get everything done that you get done.' And it's because I've been leveraging AI better.
YOUR TURN
Every role has a different first use for AI
At EBSCOlearning an architect, a product manager, a marketer and an engineer each found their own. Find where yours pays off first.
- Where AI pays off first, role by role
- What people in roles like yours fixed first
- A straight answer on the return
About 10 minutes
30 minutes, 3 straight answers
Not ready to talk yet? Explore 58,733 AI use cases by department
Sources and notes
- 77%: increase in AI tool usage after the Enterprise AI Academy; all non-users increased usage. Measured; base (number of people, and what was counted) not stated in the source. Source: p.1.
- 33%: share of people who became daily AI users after the programme. Measured; base not stated.
- 4.1 / 5: average AI adoption readiness score at the end of the programme. Self-assessed.
- 19 hours a week: time freed for a two-person marketing team by two AI assistants; capacity used fell from 85% to 61% (68 of 80 hours before). Self-reported. Source: p.11.
- About 31 hours over two to three days, to about one hour: one software architect's idea-to-test build cycle, before and after. Self-reported time comparison. Source: pp.6 to 7.
- At least 10 hours a week (inside the quote): self-reported by Maria Rioux Nagy.
- 90 minutes a week: one engineering task, about 2 hours to about half an hour a week. Self-reported. Source: p.8.
- 3,000+: employees in the parent company, EBSCO Information Services. Company fact; Source: p.1.
- Dollar values of time saved and return multiples in the source are estimates based on assumed salaries and are not used on this page.
- Michael Laddin quote: source p.2.
- Company size band 2,000-10,000: parent company, from 3,000+.