Illustrative case study · Private Equity

As AI takes the first line, seat money moves to AI add-ons.

Would AI assistants shrink the budget a seat-priced service desk tool relies on? That was the decision facing a private equity investor weighing a service desk software company. Interviews with 180 IT service staff and leaders went into a report delivered in 12 days.

Illustrative scenario

Samplen = 180

Illustrative report

AI Assistants in IT Service Desks

When AI assistants take over routine tickets, do budgets for service desk tools grow or shrink?

Study architecture

Map the workEight desk tasks placed on one-, three- and five-year horizons

Follow the budgetSeats, AI add-ons and total spend over three years

Compare the viewsWhat staff expect against what leaders plan

Scope

Fewer agents on the desk need not mean less spent on desk tools, if the money follows the work to AI.

Sample

180

Service desk staff and the IT leaders who own the desk’s tool budget.

Research by

Study typeCategory demand study

ScenarioIllustrative

Service desk AI demand · Executive reportExecutive summary

Executive summary

AI takes the routine tickets, seats fall as staff leave, and the money moves to AI add-ons

The 100 service desk staff and 80 IT leaders placed eight desk tasks on a three-horizon map and described their budget plans. AI takes status updates, article drafts and access requests first. Leaders expect 15% fewer agent seats in three years, shed as people leave, while AI add-ons grow to 23% of desk tool spend. Leaders expect more from AI than their staff do.

15%n = 80

fewer agent seats leaders expect in three years

23%n = 80

of desk tool spend on AI add-ons in three years

63%n = 80

of leaders would shed seats only as staff leave

65%n = 80

of leaders hold or raise total desk tool spend

Figures rounded. Base n = 180 staff and leaders unless stated.

Top lines

1AI takes status updates and article drafts first, and incidents stay with peopleEight desk tasks on three horizons03

2Seat money moves to AI add-ons, and 65% hold or raise the totalSeats, AI add-ons and budget direction04

3Seats go as people leave, so the shift runs at the pace of attritionWhat leaders do with a seat05

4Leaders expect AI to take more than the people who do the workStaff and leader views, task by task06

5Sample, screener and quality controlWho we interviewed and how07

AAppendixQuestion flow and data tables09

Sample

180 staff and leaders

Segments

100 staff, 80 leaders

Method

AI-moderated interviews, task horizons

Markets

US and Canada

Illustrative report. The company, scenario, and all figures are fictional.

Illustrative report02

Illustrative reportExecutive summary

Illustrative example. The company, scenario, and all figures are fictional. They show UserCue’s method and deliverable format. They do not come from any client study.

The engagement

Delivered in

days, with the report, the data file and a live readout.

Interviews

180

IT service desk staff and the leaders who fund the desk.

Audiences

100 IT service staff and 80 IT leaders.

Method

Forecast

Staff and leaders placed eight tasks on horizons; leaders split the budget.

The brief

Client

A private equity investor

Decision

Whether AI assistants shrink the budget a seat-priced desk tool relies on

Stage

Category demand study, before a deal

How we ran it

Approach

AI-moderated interviews with a task horizon map and a budget split

Measured

Task horizons, seat plans, budget direction, staff and leader views

Sample

Service desks at mid-size and large companies in the US and Canada

Headline figures

Budget65%

of IT leaders plan to hold or raise total desk tool spend over three years

Seats15%

fewer agent seats leaders expect in three years, mostly as staff leave

AI add-ons23%

of desk tool spend goes to AI add-ons in three years, up from 8% today

The question

If AI takes the first line, where does the desk budget go?

Seat-based pricing was the exposure. The service desk software company the investor was weighing charges per agent seat, and if AI assistants resolve the routine tickets, seat counts could fall and take revenue with them. Two views counted: the people who work the queue and the leaders who fund it.

Which desk work do staff and leaders expect AI assistants to take over, and by when?

Where are service desk tool budgets heading over the next three years?

Where do the people who work the desk and the people who fund it disagree?

What we did

How the study ran, start to finish.

1Kickoff
Seats on the line.
If AI assistants take over the first line, will budgets for service desk tools grow or shrink? The kickoff turned it into three questions: which desk work AI takes and when, where tool budgets head, and where the desk and its leaders disagree.
2Design
Two roles, one task map.
The screener admitted 100 staff who work the queue daily and 80 leaders who own the desk’s tool budget. Both placed the same eight tasks on one-, three- and five-year horizons, and leaders split their budget between seats and AI add-ons.
3Fielding
Each task, followed to its seat.
By voice or text, quant and qual in one conversation, for a median 34 minutes, each task handed to AI drew a follow-up on whether its seat would be refilled. The researcher read all 214 and removed 34 for mismatched roles or duplicates, leaving 180.
4Analysis
One tool budget, a new mix.
The researcher put the leaders’ three-year budget splits against today’s, and money moved from seats to AI add-ons inside one tool budget. AI add-ons grow from 8% to 23% of desk spend, and 63% of leaders would shed seats only as staff leave.
5Readout
AI add-ons grow to 23% of desk spend.
The researcher presented a 37-page report on task horizons, the staff and leader gap and the budget split to the deal team in a live readout. Now the revenue model replaces lost seats with add-on revenue at the pace of staff attrition.

What we found

Four findings from the report.

Summary

Staff and leaders expect AI assistants to take the routine tickets first, and most leaders plan to shed seats only as staff leave. AI add-ons grow from 8% to 23% of desk tool spend, and 65% of leaders hold or raise the total. Leaders expect more from AI than the staff who work the queue.

Service desk AI demand · Executive reportTop line 1

Top line 1 of 4

AI takes status updates and article drafts first, and incidents stay with people

¹ Q4. For each task, when do you expect an AI assistant to handle most of it on your desk? Four options, n = 180.

² Task scores by group are in the appendix tables.

Both groups placed eight desk tasks on a one-, three- and five-year horizon. Within three years, 68% expect AI to send status updates and 55% to draft knowledge articles¹; 52% expect it to resolve password and access requests. Judgment work stays with people: 20% expect AI to diagnose hardware faults and 15% to handle incident escalation in that time.

68%n = 180

expect AI to send status updates within three years

55%n = 180

expect AI to draft knowledge articles

52%n = 180

expect AI to resolve password and access requests

15%n = 180

expect AI to handle incident escalation

Desk tasks AI will handle within three years

Expect AI to handle the task within three years (n = 180)

When AI will resolve access requests (n = 180)

Within a yearOne to three yearsThree to five yearsLater or never

Expect AI to triage tickets within three years

Eight tasks placed at one, three or five years, or later; shares show within three years. Major incident coordination scored lowest, at 6%.

Supporting evidence

Staff who already use an assistant rate access requests as the easiest task to hand over. § 3.2
Staff and leaders name outage calls as the work they least expect to hand to an assistant. § 3.3²

Illustrative report03

Service desk AI demand · Executive reportTop line 2

Top line 2 of 4

Seat money moves to AI add-ons, and 65% hold or raise the total

¹ Q6. Split 100 points of three-year desk tool spend between agent seats and AI add-ons, n = 80.

² Q7. Who would you buy an AI add-on for the desk from? Single select, n = 80.

Leaders split their three-year desk tool budget between agent seats and AI add-ons. AI add-ons grow from 8% of desk tool spend today to 23% in three years¹, while leaders expect 15% fewer agent seats. 65% hold or raise the total: 42% plan to raise spend on desk tools and 23% to hold it, while 35% cut it.

23%n = 80

of desk tool spend on AI add-ons in three years

8%n = 80

of desk tool spend on AI add-ons today

15%n = 80

fewer agent seats expected in three years

65%n = 80

hold or raise total desk tool spend

Where the desk tool budget goes, today and in three years

Share of desk tool spend, leaders’ plans (n = 80)

Plans for total desk tool spend over three years (n = 80)

RaiseHoldCut

Where seat savings go (n = 80)

Leaders split 100 points of three-year desk tool spend between seats and AI add-ons. Budget direction is single select.

Supporting evidence

Leaders who expect the steepest seat cuts also plan the largest AI add-on budgets. § 4.2
61% of leaders would buy the AI add-on from their current desk vendor; 34% would add a separate assistant. § 4.3²

Illustrative report04

Service desk AI demand · Executive reportTop line 3

Top line 3 of 4

Seats go as people leave, so the shift runs at the pace of attrition

¹ Q8. When an AI assistant takes most of a role’s tasks, what happens to that seat? Single select, n = 80.

² Q9. How do you expect your own role to change over three years? Single select, n = 100.

When a task moves to AI, most leaders keep the person and let the seat go later. 63% would shed a seat only by leaving the role unfilled when its holder moves on¹; 20% would cut seats once the assistant is live, and 17% would move staff to other IT work. That sets the pace of the 15% seat fall leaders expect.

63%n = 80

would leave roles unfilled as staff move on

20%n = 80

would cut seats once the assistant is live

17%n = 80

would move staff to other IT work

74%n = 80

expect fewer agent seats in three years

What leaders do with a seat when AI takes its tasks

Plan for the seat when AI takes its main tasks (n = 80)

Agent seats expected in three years (n = 80)

FewerAbout the sameMore

How staff expect their own role to change (n = 100)

Single select. Seat change is each leader’s estimate for their own desk.

Supporting evidence

Leaders of the largest desks are the most likely to cut seats as soon as the assistant goes live. § 5.2
Knowledge work and assistant upkeep top the new duties staff expect. § 5.3²

Illustrative report05

Service desk AI demand · Executive reportTop line 4

Top line 4 of 4

Leaders expect AI to take more than the people who do the work

¹ Q4 by group. Access requests and triage: n = 80 leaders, n = 100 staff.

² Assistant users against non-users is directional; see the appendix tables.

Staff and leaders rated the same eight tasks. 70% of leaders expect AI to resolve password and access requests within three years, against 37% of the staff who handle them¹. The gap holds on triage and narrows on status updates. Staff point to the edge cases: locked accounts, shared mailboxes and approvals that need a manager.

70%n = 80

of leaders expect AI to resolve access requests

37%n = 100

of staff expect the same

63%n = 80

of leaders expect AI to triage tickets

39%n = 100

of staff expect AI to triage tickets

Expect AI to handle the task within three years, by group

Leaders against staff, within three years

Top three for AI, IT leaders (n = 80)

1Status updates to users
2Password and access requests
3Knowledge article drafts

Top three for AI, IT service staff (n = 100)

1Status updates to users
2Knowledge article drafts
3Ticket triage and routing

The same eight tasks and horizons for both groups; shares show within three years.

Supporting evidence

On access requests, staff point to the approvals an assistant cannot give. § 6.2
Staff who use an assistant today sit closer to their leaders’ view. § 6.3²

Illustrative report06

Top line 1 of 4 · page 03Illustrative report · top line 1 of 4 · swipeIllustrative report

Question 1 · The work

Routine requests go to AI first, and incidents stay with people.

Within three years, 68% expect AI to send status updates to users and 55% expect it to draft knowledge articles. Only 15% expect it to handle incident escalation in that time.

Status updates to users68%

Knowledge article drafts55%

Password and access requests52%

Incident escalation15%

Expect AI to handle the task within three years · n = 180

Question 2 · The budget

Seat money moves to AI add-ons, and 65% hold or raise the total.

Leaders expect 15% fewer agent seats in three years, while AI add-ons grow from 8% to 23% of desk tool spend. 65% plan to hold or raise total spend on desk tools: 42% raise it, 23% hold it and 35% cut.

23%AI add-ons in three years

8%AI add-ons today

AI add-ons’ share of desk tool spend · IT leaders · n = 80

Question 2 · The pace

Seats go as people leave, so attrition sets the pace.

63% of leaders would shed a seat only by leaving the role unfilled when its holder moves on. 20% would cut seats once the assistant is live, and 17% would move staff to other IT work.

Leave the role unfilled63%

Cut the seat at go-live20%

Move the person to other work17%

Plan for a seat when AI takes its main tasks · IT leaders · n = 80

Question 3 · The gap

Leaders expect AI to take more than the people who do the work.

70% of leaders expect AI to resolve password and access requests within three years, against 37% of the staff who handle them. The gap holds on ticket triage, 63% against 39%.

70%IT leaders

37%IT service staff

Expect AI to resolve access requests in three years · n = 80 vs n = 100

“Half my tickets are password resets. The other half are the ones where the reset didn’t work, and that’s where I earn my pay.”

Service desk analyst · Illustrative quote

Illustrative quotes

Three themes behind the numbers.

Routine work

“The assistant already clears most of the access requests overnight. By the time I log in, the queue is the hard stuff.”

Service desk analyst

Seats

“I won’t lay anyone off for this. When someone leaves, I won’t backfill the seat, and the license goes with them.”

IT service director

Budget

“The money stays with the desk. What we save on seats goes into the AI add-on, and I would rather pay for that than for overtime.”

Head of IT operations

What the investor did

Three moves made with the report.

1The model · from finding 2

Modeled the assistant as a desk add-on.

The revenue model grows AI add-ons to 23% of desk tool spend as seat revenue falls; 61% of leaders would buy the add-on from the desk vendor they use today.

2The timing · from finding 3

Let attrition pace the seat losses.

Seats fall 15% over three years in the model, mostly as roles go unfilled, plus a smaller step for the 20% of leaders who cut seats once the assistant is live.

3The roadmap · from finding 4

Put the edge cases first on the roadmap.

Locked accounts and manager approvals, where staff doubt AI most, lead the target’s assistant roadmap.

Illustrative example. The company, scenario, and all figures are fictional. They show UserCue’s method and deliverable format. They do not come from any client study.

Bring us your next diligence question.

A senior researcher designs and presents every study.

Talk to our team

More private equity studies

Renewal pricing study

If AI takes the first line, where does the desk budget go?

How the study ran, start to finish.

1Kickoff

Seats on the line.

If AI assistants take over the first line, will budgets for service desk tools grow or shrink? The kickoff turned it into three questions: which desk work AI takes and when, where tool budgets head, and where the desk and its leaders disagree.

2Design

Two roles, one task map.

The screener admitted 100 staff who work the queue daily and 80 leaders who own the desk’s tool budget. Both placed the same eight tasks on one-, three- and five-year horizons, and leaders split their budget between seats and AI add-ons.

3Fielding

Each task, followed to its seat.

By voice or text, quant and qual in one conversation, for a median 34 minutes, each task handed to AI drew a follow-up on whether its seat would be refilled. The researcher read all 214 and removed 34 for mismatched roles or duplicates, leaving 180.

4Analysis

One tool budget, a new mix.

The researcher put the leaders’ three-year budget splits against today’s, and money moved from seats to AI add-ons inside one tool budget. AI add-ons grow from 8% to 23% of desk spend, and 63% of leaders would shed seats only as staff leave.

5Readout

AI add-ons grow to 23% of desk spend.

The researcher presented a 37-page report on task horizons, the staff and leader gap and the budget split to the deal team in a live readout. Now the revenue model replaces lost seats with add-on revenue at the pace of staff attrition.

Four findings from the report.

Summary

Service desk AI demand · Executive reportTop line 1

Top line 1 of 4

AI takes status updates and article drafts first, and incidents stay with people

¹ Q4. For each task, when do you expect an AI assistant to handle most of it on your desk? Four options, n = 180.

² Task scores by group are in the appendix tables.

68%n = 180

expect AI to send status updates within three years

55%n = 180

expect AI to draft knowledge articles

52%n = 180

expect AI to resolve password and access requests

15%n = 180

expect AI to handle incident escalation

Desk tasks AI will handle within three years

Expect AI to handle the task within three years (n = 180)

When AI will resolve access requests (n = 180)

Within a yearOne to three yearsThree to five yearsLater or never

Expect AI to triage tickets within three years

Eight tasks placed at one, three or five years, or later; shares show within three years. Major incident coordination scored lowest, at 6%.

Supporting evidence

Staff who already use an assistant rate access requests as the easiest task to hand over. § 3.2
Staff and leaders name outage calls as the work they least expect to hand to an assistant. § 3.3²

Illustrative report03

Service desk AI demand · Executive reportTop line 2

Top line 2 of 4

Seat money moves to AI add-ons, and 65% hold or raise the total

¹ Q6. Split 100 points of three-year desk tool spend between agent seats and AI add-ons, n = 80.

² Q7. Who would you buy an AI add-on for the desk from? Single select, n = 80.

23%n = 80

of desk tool spend on AI add-ons in three years

8%n = 80

of desk tool spend on AI add-ons today

15%n = 80

fewer agent seats expected in three years

65%n = 80

hold or raise total desk tool spend

Where the desk tool budget goes, today and in three years

Share of desk tool spend, leaders’ plans (n = 80)

Plans for total desk tool spend over three years (n = 80)

RaiseHoldCut

Where seat savings go (n = 80)

Leaders split 100 points of three-year desk tool spend between seats and AI add-ons. Budget direction is single select.

Supporting evidence

Leaders who expect the steepest seat cuts also plan the largest AI add-on budgets. § 4.2
61% of leaders would buy the AI add-on from their current desk vendor; 34% would add a separate assistant. § 4.3²

Illustrative report04

Service desk AI demand · Executive reportTop line 3

Top line 3 of 4

Seats go as people leave, so the shift runs at the pace of attrition

¹ Q8. When an AI assistant takes most of a role’s tasks, what happens to that seat? Single select, n = 80.

² Q9. How do you expect your own role to change over three years? Single select, n = 100.

63%n = 80

would leave roles unfilled as staff move on

20%n = 80

would cut seats once the assistant is live

17%n = 80

would move staff to other IT work

74%n = 80

expect fewer agent seats in three years

What leaders do with a seat when AI takes its tasks

Plan for the seat when AI takes its main tasks (n = 80)

Agent seats expected in three years (n = 80)

FewerAbout the sameMore

How staff expect their own role to change (n = 100)

Single select. Seat change is each leader’s estimate for their own desk.

Supporting evidence

Leaders of the largest desks are the most likely to cut seats as soon as the assistant goes live. § 5.2
Knowledge work and assistant upkeep top the new duties staff expect. § 5.3²

Illustrative report05

Service desk AI demand · Executive reportTop line 4

Top line 4 of 4

Leaders expect AI to take more than the people who do the work

¹ Q4 by group. Access requests and triage: n = 80 leaders, n = 100 staff.

² Assistant users against non-users is directional; see the appendix tables.

70%n = 80

of leaders expect AI to resolve access requests

37%n = 100

of staff expect the same

63%n = 80

of leaders expect AI to triage tickets

39%n = 100

of staff expect AI to triage tickets

Expect AI to handle the task within three years, by group

Leaders against staff, within three years

Top three for AI, IT leaders (n = 80)

1Status updates to users
2Password and access requests
3Knowledge article drafts

Top three for AI, IT service staff (n = 100)

1Status updates to users
2Knowledge article drafts
3Ticket triage and routing

The same eight tasks and horizons for both groups; shares show within three years.

Supporting evidence

On access requests, staff point to the approvals an assistant cannot give. § 6.2
Staff who use an assistant today sit closer to their leaders’ view. § 6.3²

Illustrative report06

Top line 1 of 4 · page 03Illustrative report · top line 1 of 4 · swipeIllustrative report

Question 1 · The work

Routine requests go to AI first, and incidents stay with people.

Within three years, 68% expect AI to send status updates to users and 55% expect it to draft knowledge articles. Only 15% expect it to handle incident escalation in that time.

Status updates to users68%

Knowledge article drafts55%

Password and access requests52%

Incident escalation15%

Expect AI to handle the task within three years · n = 180

Question 2 · The budget

Seat money moves to AI add-ons, and 65% hold or raise the total.

23%AI add-ons in three years

8%AI add-ons today

AI add-ons’ share of desk tool spend · IT leaders · n = 80

Question 2 · The pace

Seats go as people leave, so attrition sets the pace.

63% of leaders would shed a seat only by leaving the role unfilled when its holder moves on. 20% would cut seats once the assistant is live, and 17% would move staff to other IT work.

Leave the role unfilled63%

Cut the seat at go-live20%

Move the person to other work17%

Plan for a seat when AI takes its main tasks · IT leaders · n = 80

Question 3 · The gap

Leaders expect AI to take more than the people who do the work.

70% of leaders expect AI to resolve password and access requests within three years, against 37% of the staff who handle them. The gap holds on ticket triage, 63% against 39%.

70%IT leaders

37%IT service staff

Expect AI to resolve access requests in three years · n = 80 vs n = 100

Three themes behind the numbers.

Routine work

“The assistant already clears most of the access requests overnight. By the time I log in, the queue is the hard stuff.”

Service desk analyst

Seats

“I won’t lay anyone off for this. When someone leaves, I won’t backfill the seat, and the license goes with them.”

IT service director

Budget

“The money stays with the desk. What we save on seats goes into the AI add-on, and I would rather pay for that than for overtime.”

Head of IT operations

Three moves made with the report.

1The model · from finding 2

Modeled the assistant as a desk add-on.

The revenue model grows AI add-ons to 23% of desk tool spend as seat revenue falls; 61% of leaders would buy the add-on from the desk vendor they use today.

2The timing · from finding 3

Let attrition pace the seat losses.

Seats fall 15% over three years in the model, mostly as roles go unfilled, plus a smaller step for the 20% of leaders who cut seats once the assistant is live.

3The roadmap · from finding 4

Put the edge cases first on the roadmap.

Locked accounts and manager approvals, where staff doubt AI most, lead the target’s assistant roadmap.

Illustrative example. The company, scenario, and all figures are fictional. They show UserCue’s method and deliverable format. They do not come from any client study.

If AI takes the first line, where does the desk budget go?

Which desk work do staff and leaders expect AI assistants to take over, and by when?

Where are service desk tool budgets heading over the next three years?

Where do the people who work the desk and the people who fund it disagree?

How the study ran, start to finish.

Seats on the line.

Two roles, one task map.

Each task, followed to its seat.

One tool budget, a new mix.

AI add-ons grow to 23% of desk spend.

Four findings from the report.

Routine requests go to AI first, and incidents stay with people.

Seat money moves to AI add-ons, and 65% hold or raise the total.

Seats go as people leave, so attrition sets the pace.

Leaders expect AI to take more than the people who do the work.

Three themes behind the numbers.

Three moves made with the report.

Modeled the assistant as a desk add-on.

Let attrition pace the seat losses.

Put the edge cases first on the roadmap.

Bring us your next diligence question.

Tax Compliance Software VoC

Dental Lab Contract Manufacturing Outlook

Wound Care Product Evaluation

If AI takes the first line, where does the desk budget go?

Which desk work do staff and leaders expect AI assistants to take over, and by when?

Where are service desk tool budgets heading over the next three years?

Where do the people who work the desk and the people who fund it disagree?

How the study ran, start to finish.

Seats on the line.

Two roles, one task map.

Each task, followed to its seat.

One tool budget, a new mix.

AI add-ons grow to 23% of desk spend.

Four findings from the report.

Routine requests go to AI first, and incidents stay with people.

Seat money moves to AI add-ons, and 65% hold or raise the total.

Seats go as people leave, so attrition sets the pace.

Leaders expect AI to take more than the people who do the work.

Three themes behind the numbers.

Three moves made with the report.

Modeled the assistant as a desk add-on.

Let attrition pace the seat losses.

Put the edge cases first on the roadmap.

Bring us your next diligence question.

Tax Compliance Software VoC

Dental Lab Contract Manufacturing Outlook

Wound Care Product Evaluation