Positioning With Customers
About how we open, navigate and close AI conversations with the buyers and influencers in the customer's room.
Welcome
AI conversations with customers fail for one reason more than any other: we answer the question we wanted them to ask, not the one they're actually asking. A CFO asking about AI is not asking the same thing as a CHRO asking about AI. Same words, different conversation.
This module is about reading the room. Who's sat opposite you, what they care about, what they fear, and the opening line that earns the next twenty minutes. It is aimed at anyone who walks into a customer meeting where AI will come up - which is now every meeting.
Learning objectives
Who's in the room
Four personas drive almost every Workday AI conversation. The same slide lands differently with each of them. Know the fear before you reach for the hook.
CHRO
What they care about
Employee experience, manager productivity, talent retention, and being seen as a modern HR function at board level.
What they fear
Bias in hiring or performance decisions. Union or works-council pushback. Being the executive who rolled out AI that made the Financial Times for the wrong reason.
"Workday ships the AI inside the same tenant your HR team already trusts - so the governance conversation is one you've largely already had. Let's talk about where it takes real work off your managers."
CFO
What they care about
Unit economics, ROI timelines, and whether AI spend is operating cost or transformation. Control over contracting and renewal exposure.
What they fear
Paying twice - once for Workday, again for AI on top. Runaway consumption bills. An AI initiative that can't be defended in the next audit committee.
"Your AI capacity comes through flex credits you already own. The question isn't 'how much more do we spend?' - it's 'where do we point the credits we've got for the biggest return?'"
CIO / CDO
What they care about
Architecture coherence, data boundaries, a defensible AI policy, and not being the team that has to clean up a shadow AI estate in two years.
What they fear
Another fragmented platform story. Data leaving the tenant. Being bypassed by the business because "HR already bought it".
"Workday agents inherit your existing tenant security - they don't create a new surface. Where it gets interesting architecturally is deciding when to Use, when to Adapt, and when to Build."
HRBP or business leader
What they care about
Their team's day-to-day - time spent on admin, case backlogs, manager enablement, getting through open enrolment without it eating a month.
What they fear
Another "transformation" that lands on them. Being told to adopt something that doesn't fit how the work actually happens. Job security for their team.
"Forget the strategy deck for a moment. Where in your week does your team lose hours to work the system should already be doing? That's where we start."
Discovery questions
Pick two or three per meeting - not the whole list. The goal is to hear the customer describe their own problem in their own words before we reach for an answer.
For the CHRO
Where are your managers losing the most time inside Workday today? Which HR decisions are the ones you most want to make more consistent? What's your works-council posture on AI in people decisions?
For the CFO
How are you thinking about AI as operating cost versus transformation investment? Are flex credits currently on the table in your Workday renewal? What would an AI business case need to look like to clear your hurdle rate?
For the CIO / CDO
Where does Workday sit in your AI policy today - in scope, out of scope, or undecided? Which data boundaries are non-negotiable for you? Where have you already seen shadow AI creep in from the business?
For the HRBP / business leader
Walk me through a week - where does the system get in your way? Which tasks do you wish somebody else was doing before they hit your team? What's one process you'd rip up tomorrow if you could?
Resist the urge to answer your own question. If a customer pauses for five seconds, let them. The answer you get after the pause is almost always the real one.
The value narrative - 90 seconds
When a customer asks "so what's your point of view on AI in Workday?", you need a short answer. Not a deck. Ninety seconds, three moves, and a clear next step.
Use
Start with the Workday agents that are GA today. Fastest path to value, lowest risk, already paid for.
Adapt
Configure the GA agents and the supporting processes around the customer's operating model.
Build
Where there's a genuine gap and the value is there, build custom agents on Sana Agent Builder.
The 90-second script
"Most customers we work with start in the same place - they know AI is going to matter, and they're not sure where to point it first. Our approach, the Kainos AI Navigator, is three moves. First, Use - the agents Workday already ships, generally available today, that you're effectively already paying for. Second, Adapt - shaping those agents and the processes around them to your operating model, because an out-of-the-box agent in the wrong process still wastes people's time. Third, Build - where there's a real gap that matters, we build a custom agent on Sana. Most customers get 70% of their outcome from the first two. The skill is knowing when to reach for the third."
Use, Adapt, or Build. Never "Extend". Workday ships a portfolio of agents - not a product called the Workday AI Navigator. Kainos AI Navigator is our approach, not a Workday feature.
Objection handling
The same five objections come up in almost every conversation. You don't need a clever answer - you need a clear one. Acknowledge, reframe, offer the next step.
"We can't let AI near our people data - security and data residency are a hard no."
Acknowledge it's the right question to ask - then walk them through the four guardrail answers from Module 2. Workday agents inherit tenant security; they only see what the invoking user can already see. Data stays within Workday's trust model. Actions are auditable and human-in-the-loop. Offer to bring the CoE AI and their security team into a single working session - most security objections dissolve once the risk team sees the actual architecture rather than a generic AI scare story.
"This sounds expensive. What's it going to cost us?"
For most customers, the answer is: less than they think, because they likely already have flex credits as part of their Workday agreement. Reframe the conversation from "new spend" to "capacity you already own". Don't quote numbers on the spot - promise a flex credit baseline review and route into the CoE AI to run it properly. That becomes a legitimate reason for the next meeting.
"We're not ready. Our data is a mess, our processes are a mess, and we've just finished deploying Workday."
Agree with them - and then push gently. Readiness is not a pre-condition; readiness is built by starting with a GA agent on a bounded use case where the data is already good enough. Point at the Use leg of the approach. "You don't need to fix everything to start - you need to start somewhere you can win, and use that to earn the right to do the harder things."
"We've been burned by AI already. The last thing our business needs is another AI pilot that goes nowhere."
Take the scepticism seriously - it's usually earned. The problem in most failed AI pilots was not the technology; it was a solution in search of a problem, no owner, and no adoption plan. Our answer is the opposite: start from a job to be done, use agents that are already GA, and plan for adoption from day one (point at Modules 9 and 10). Offer a scoped, time-boxed first move rather than another strategy exercise.
"Doesn't this lock us further into Workday?"
It's a fair challenge. The honest answer: Workday AI is tightly coupled to Workday data and process - that's what makes it work, and it's also what the customer is buying. We don't try to argue lock-in away. Instead, reframe: the lock-in question is already answered the day they chose Workday as their system of record. The AI layer is about getting more value out of a platform they've already committed to, not committing to a new one. If they want portable AI capability for processes that sit outside Workday, that's a different conversation - and an honest one we should have.
Routing qualified interest
When a customer says "yes, let's go further", the worst outcome is a cold hand-off. The CoE AI picks up the thread only if you hand it over warm.
What qualifies as warm
A named sponsor, a specific use case or pain point described in their words, an indication of flex credit position, and a date in the diary for the next conversation.
What to capture before handing off
Persona and role of the customer contact. The discovery questions you asked and the answers you heard. Any objections raised and how you responded. Anything you promised to come back on.
How to bring in the CoE AI
One Teams message to the CoE AI Lead with the four items above. Don't attach a 40-slide pack - attach the two sentences that matter. The CoE will route to the right architect, change lead, or delivery lead within 48 hours.
When not to route yet
If the conversation is still exploratory, keep it with you. Bringing the CoE in too early burns their time and signals to the customer you've escalated out of your depth. Route when there's a real thread to pull.
A good hand-off is a paragraph the CoE can read in 30 seconds and act on in the next meeting. If you need a deck to brief them, you're not ready to hand off.
Three common traps
1. Being too technical, too early
The quickest way to lose a CHRO is to open with agent anatomy, tenant security posture, and the Sana Agent Builder SDK. Save the architecture for the architect. With business leaders, lead with the job to be done and the time it gives back - then let them pull you deeper if they want to.
2. Pitching custom build when GA already fits
Customers often ask for a "custom AI agent" because that's the language the market has taught them. Most of the time, a GA agent plus sensible configuration solves 70-80% of the problem at a fraction of the risk. Default to Use, then Adapt. Build is the answer when the value is big enough to justify the lifecycle - not the opening move.
3. Promising the roadmap
Never say "that'll be available next quarter". Workday dates move, and a promised date you can't keep costs you more credibility than the original gap ever would. Say "it's on the current Workday roadmap" and set expectations that you'll confirm timing on the call with the CoE AI. Trust is harder to rebuild than expectations are to reset.
Concrete beats clever. A customer remembers the one thing you said that matched their week - not the five things you said about the platform.
Check your understanding
Three questions. Each explains why every answer is right or wrong - the reasoning matters more than the score.
Next steps
- Architects: Module 5 - Flex credits & AI architecture. The technical backbone of every commercial conversation you'll have after this one.
- Customer Success and delivery leads: Module 9 - Change management for AI. The conversation you'll be having the moment the customer says yes.
- Everyone customer-facing: Module 10 - Adoption, not activation. Why the deal isn't done when the agent is switched on.
- Practice: Pick a live account. Write down which persona you're meeting next, which two discovery questions you'd open with, and which objection you're most worried about. Bring it to your next 1:1 with the CoE AI Lead.