What are your thoughts on Windows Copilot? Are you considering it in your org?
Excited! I see plenty of use cases for our team21%
It looks interesting, but I’m not sure it’ll fit in our tool stack42%
It doesn’t look useful for us16%
I have concerns about it11%
I have concerns about AI tools in general4%
I haven’t looked into it yet6%
I would love to hear an update after five months before I decide if I want to follow in your footsteps or whether you advise me to avoid at all costs!
Sure Mark - it was a successful pilot and we are now planning a production project. Here are a few key learnings and outcomes:
Benefits Definition - we defined the following as key measures
1. ROI – how many hours saved using Copilot
2. A usefulness rating for the most common Copilot use cases
3. Positive impact to productivity, wellbeing and/or work life balance
4. Any potential duplicate functionality/licenses that could be removed
5. Improves work life for people with a disability or where English is their second language
#1 ROI – how many hours saved using Copilot:
- Low value: 20% of staff (n=230) saved 5-10 mins per day. Executive assistants, Comms, L&D, Change and administrators topped this category.
- Medium value: 62% saved 11mins to 1 hour per day. Executives, managers, and people with a disability topped this list.
- High value: 16.5% saved 1+ hours per day. Project managers, people with English as a second language and people leaders topped this list.
Based on these results, we removed the low value category and calculated an average of $10,483 per person per annum. We used actual salaries.
#2 A usefulness rating for the most common Copilot use cases
In order of most useful: Teams Recap, Word, Outlook, Teams In-meeting prompting, Teams Chat, PowerPoint
Least useful: Excel (too immature for advanced Excel users but will likely improve over time), Loop (just because it is new and not integrated into people's workflow but I think it has great potential for team-based collaboration generally - better than OneNote put it that way), OneNote and Whiteboard
#3 Positive impact to productivity, wellbeing and/or work life balance
BACKGROUND
34% of staff are regularly working overtime – Executives, Managers and Project Managers top the list
46% of staff have to work overtime on certain occasions – Project Managers top the list, with Managers, Communications/ Change/Education next
Not easily finding information is the top obstacle to productivity with having inefficient meetings coming in second
47% of staff spend 3-4 hours in meetings per day. The more time spent in meetings, the less time spent on actionable activities.
SURVEY
- Staff opinions were strong towards adopting AI for both productivity and wellbeing. Staff agree or strongly agree:
88% - AI needs to be adopted in the department to improve productivity
76% - AI needs to be adopted in the department to improve wellbeing
13% - AI is going to take my job in the future
- Broadly, Copilot had a positive impact on productivity and reduced time to execute certain tasks, especially drafting Microsoft Office-based communications
Department Results (n=230)
Microsoft Global Baseline
helps get to a good first draft faster
92%
85%
helps improve productivity
87%
80%
wish it could do more for them
73%
-
saves time on mundane tasks
73%
71%
helps make work more enjoyable
65%
-
helps focus with important work
64%
67%
- Copilot had a positive impact on creativeness and helped more than 80% of participants to generate new ideas and jump start their creative processes
- Although less of an impact in the work-life balance area, Copilot still produced a positive outcome in reducing workloads and over 60% agreed the ability to use AI would be a factor in their future workplace choice. Staff agree or strongly agree:
61% - the ability to use AI technologies would affect their choice of future workplace
56% - helps reduce their workload
35.7% - improved work life balance
# 4 Any potential duplicate functionality/licenses that could be removed
This will vary per organisation but recommend looking at overlap between existing paid tools e.g. Miro Vs MS Whiteboard.
# 5 Improves work life for people with a disability or where English is their second language
24 participants speak English as a second language
The vast majority of them saved at least 30 minutes of time per day with Copilot
The average annual productivity saving across this group is $9024 (per annum)
ENGLISH AS A SECOND LANGUAGE
Most benefited from:
helps focus on important work
spending less time processing email
getting first drafts faster
saves time on mundane tasks
being more productive overall
DISABILITY
16 participants have a disability
The vast majority of them saved at least 30 minutes of time per day with Copilot
The average annual productivity saving across this group is $8223 (per annum)
Most benefited from:
getting first drafts faster
helping to generate ideas for writing
jump starting the creative process
being more productive overall
Key Learnings:
- We made each participant go through 5 hours of mandatory training. It appeared excessive to some individuals at the start but they changed their mind by the end.
- Data security is a problem. Copilot does not allow access to more files but if you have historically poor processes around granting data access then you'll have more cleanup to do prior - and determining the inflection point for risk mitigation Vs acceptance.
- Microsoft Copilot for Office is built into Microsoft Office. Staff who do not use Office regularly e.g. they specialise in a business application (e.g. spatial) or are rarely in front of a computer (e.g. field staff) will gain limited value from this technology and are unlikely to return reasonable productivity or wellbeing benefits, regardless of their role or function. This is because not only do you need to be using the Office products regularly, but you also need to make a mindset change that gets embedded into a regular routine.
- At this stage Excel Copilot functionality is limited and verbal and training feedback indicated it would be one of the least useful functions. This subsequently indicates staff who use Excel most of the day, are unlikely to see a major benefit until the product is improved.
* Executive Assistants, Executive Officers and administrative support staff
* An executive pair i.e. an executive (SEB 1/2/3) and their assistant. It is likely to be more beneficial if both participate to cross-collaborate on the new features.
* Individuals who work with financial models in Excel
* Individuals with accessibility needs e.g. limited use of limbs, dyslexia
* Staff involved in daily formal communications, marketing, events, knowledge management and learning development
* Staff involved in daily communication with citizens and customers
* Individuals whose primary role is to develop and/or curate content e.g. slides, presentations, reports, briefings, scientific research etc
* Individuals who perform project management, business analysis and/or project coordination functions daily
* Individuals who maintain formal appointments scheduling
We also enforced a mandatory 5-hour training regime for staff in the POC before their license was assigned, and they had to sign an individual commitment around notifying us immediately of any data privacy issues and commiting to reporting back once a week (a 2 min survey) on productivity improvements or issues.