Asahi Kasei
What we did
Reduction in search time
AMS capacity shifted to strategic projects
Research time per question
Asahi Kasei
Global chemical and materials innovator with ~50,000 employees across 20+ countries
Industry
Chemical
head count
49,295
team size
4
At a glance
The Challenge
Asahi Kasei isn't a company where the Workday team keeps the lights on. With 49,000+ employees across 20+ countries, the four-person Workday team operates more like an internal product org: running a live Workday-to-ServiceNow core connector integration, optimizing payroll, tightening security audits, evaluating new modules like Engagement Builder and Skills Cloud. Christian Delcid, a senior Workday analyst, splits his time roughly 50/50 between break-fix support and strategic work. That's the job. The problem was that even simple questions came with enormous overhead.
Christian estimated he was losing 30 to 90 minutes per day to research. Not complex analysis. Not deep architectural thinking. Just finding answers. The workflow looked the same every time: start in Workday Community, hit a dead end on deprecated threads and reorganized documentation, pivot to SharePoint, check the internal wiki, skim a Slack thread from four months ago, open fifteen browser tabs, and eventually piece together something that might be correct. On top of that, another 30 minutes each morning reviewing Community digest emails just to stay current.
"I'd go in for one question and get stuck in rabbit holes. An hour and a half later, I'd have 10 tabs open and no idea what my original goal was."
The friction wasn't just eating hours — it was eating ambition. When validating a configuration question costs an afternoon, you raise the bar for which questions are worth asking. Ideas sit in notebooks. Proposals don't get made. The platform stays where it is — not because the team lacks vision, but because research overhead eats the bandwidth to act on it.
And then there was the AMS contract. Asahi Kasei was paying $12,000-$13,000 per month to their AMS partner for 10 active ticket slots. Not 10 resolutions per month, but 10 concurrent tickets in flight. The model created a rationing dynamic: every issue required a judgment call about whether it was worth burning a slot. Routine documentation-level questions competed for the same finite pool as deep architectural decisions, and the routine ones were winning.
What Changed
All four admins got access to Mando. Christian started using it for quick validation on configuration questions. The rest of the team followed. Not because they were told to or forced by a process…but because the alternative was inherently painstaking and time consuming.
Instead of bouncing between Community threads, buried documentation, and half-remembered email threads, the team started every question in Mando. Answers came back grounded in Workday admin guides, real-world implementation patterns, and validated customer examples — with citations they could verify. The research loop that used to consume 30 to 90 minutes collapsed to a few minutes. But the real shift wasn't speed. It was trust.
"If Mando's not bringing back what I'm looking for, it's nowhere else."
When the absence of an answer carries that kind of authority, the entire research behavior changes. You don't open fifteen tabs to cross-reference. You ask once, you get a grounded answer, and you move on to the work that answer enables.
The AMS equation shifted in parallel. The documentation-level questions that used to consume ticket slots — configuration checks, "is this a bug or expected behavior" lookups, calculated field interactions — those get handled in Mando now. 7 of the team's 10 slots sit available at any given time, preserved for the projects that genuinely require hands-on tenant expertise. With a public cloud migration scheduled for July and a potential multi-year security group rebuild on the horizon, that capacity is about to matter more than ever.
"We're not buying into something that is going to stay stagnant. This is the complete opposite. I'm seeing this thing get better in real time."
The Results
Since adopting Mando, Asahi Kasei's Workday team has shifted how they spend their time, their AMS budget, and their strategic energy:
>70% reduction in time spent searching: from 30–90 minutes per day to roughly a third of that. The fifteen-tab, multi-platform research spiral replaced by a single query.
~$100K/yr in AMS capacity redirected to strategic projects: seven of ten ticket slots consistently available — banked for the public cloud migration, the security group rebuild, and the initiatives that actually require outside expertise.
90-minute research loops resolved in minutes: questions that used to trigger an afternoon of cross-referencing now get answered in a single Mando query, with citations the team can verify.
The team didn't abandon their AMS partner. They elevated the relationship. Mando handles the first line of inquiry. AMS handles the exceptions. We expect more teams to move toward this model as AI continues to lower the barrier to action — where the admin isn't waiting on an outside team to answer a question they could answer themselves in two minutes, and the AMS relationship is freed up to do genuinely high-value architectural work.
"This is the floor. And I know it's only going up from here."
Use case
01
From Rabbit Holes to Real Answers
Every Workday admin knows the feeling. A straightforward question on how to adjust a calculated field, whether a security group change will cascade, what the interaction is between two business process steps turns into an afternoon. Not because the question is hard. Because the information is scattered across a dozen surfaces that were never designed to work together.
Before Mando
Christian's team ran the same loop on every question. Start with Workday Community, the obvious first stop, and often the worst. Community search is notoriously unreliable: you get threads from 2019, answers that reference deprecated functionality, links to documentation that's been reorganized twice since the post was written. If Community doesn't yield a clear answer in ten minutes, you pivot. SharePoint folders next. Then the internal wiki. Then Slack, searching for that one thread where someone mentioned this exact issue four months ago.
An hour and a half later, you have fifteen browser tabs open, three partially relevant answers, and no certainty about any of them. So you test in sandbox. Another thirty minutes. And if the sandbox results are ambiguous, you escalate: file an AMS ticket and wait.
This wasn't an occasional frustration. It was the daily reality for all four admins. Every day, each person lost 30 to 90 minutes to this cycle. Across the team, that was 10 to 15 hours per week spent not on configuration, not on strategy, not on moving the platform forward - on searching.
"I'd go in for one question and get stuck in rabbit holes. An hour and a half later, I'd have 10 tabs open and no idea what my original goal was."
The cost of this wasn't just time. It was cognitive. Every context switch, from Community to SharePoint to Slack to sandbox, carried a mental tax. Admins weren't just losing minutes; they were losing focus. The deep work of actually configuring and improving Workday kept getting interrupted by the shallow work of finding information about how to configure and improve Workday.
And the organizational cost was real too. When it takes an hour to validate a simple question, you stop asking questions. You stick with what you know. You don't propose the optimization you were thinking about because you can't afford the research time to confirm it's viable. Innovation doesn't get killed by complexity, it gets killed by friction.
After Mando
Mando replaced the entire research loop with a single starting point. Instead of Community → SharePoint → wiki → Slack → sandbox, the team now starts in Mando. Asks the question. Gets an answer grounded in Workday admin guides, validated implementation patterns, and real-world customer examples.
The answer isn't always complete: sometimes it's a pointer, sometimes it's a narrowed-down starting point. But even in the worst case, it eliminates the first hour of searching. In the best case, it resolves the question in under two minutes. Either way, the fifteen-tab rabbit hole is gone.
Daily research time dropped from 30–90 minutes per admin to 10–15 minutes. Often less. The team collectively reclaimed over 10 hours per week, time that went directly back into the configuration and strategic work they were hired to do.
The Results
Search time dropped by more than 70%. All four admins independently adopted Mando as their default starting point, not because they were told to, but because the alternative was obviously worse. The rabbit holes didn't disappear entirely. But they went from being the daily norm to a rare exception, triggered only when a question genuinely had no documented answer. For everything else, there was Mando.
Use case
02
Putting a $156K/Year AMS Contract Back to Work
Application Management Services contracts are the safety net of enterprise Workday teams. You pay a monthly fee, you get a fixed number of active tickets, and when something breaks or you need expertise you don't have in-house, you file a ticket and wait. It's insurance. And like all insurance, the economics only work if you're not using it for every little thing.
The problem is that most teams do.
Before Mando
Asahi Kasei was paying between $12,000 and $13,000 per month — roughly $156,000 per year — for their AMS contract. That bought them 10 active tickets at any given time. Not 10 resolutions per month, but 10 concurrent tickets in flight. The team started at 5 slots for $6,000 and recently upgraded. Every issue that got escalated occupied a slot. Every slot occupied was a slot unavailable for the next issue.
This created a rationing dynamic that distorted how the team worked. Before filing a ticket, admins had to make a judgment call: is this problem worth one of our ten slots? Can I solve this faster myself than waiting for the AMS team to pick it up? If I use a slot on this, will we have capacity if something bigger breaks tomorrow?
Christian was blunt:
"You didn't ask for this, but I'm going to tell you — not a big fan of this approach."
His frustration was structural. The AMS partner doesn't know Asahi Kasei's tenant setup. There's constant consultant turnover, which means constant re-onboarding — every new resource needs implementer access provisioned, and when someone leaves the partner's side, Asahi Kasei is responsible for cleaning up credentials. The overhead of managing the relationship sometimes rivaled the overhead of just solving the problem internally.
The result was a team spending significant time trying to avoid using the support they were paying $156K a year for — burning hours on Community searches and sandbox testing before reluctantly filing a ticket. And the mix was wrong. The full spectrum of questions competed for those 10 slots: deep architectural decisions that genuinely required outside expertise alongside configuration checks and documentation lookups the team could have handled themselves if they could just find the answer fast enough. Routine questions crowded out strategic ones.
The AMS Model Isn't Broken — It's Misallocated
The AMS model assumes the primary bottleneck is expertise: you don't have the knowledge in-house, so you rent it from a partner. For genuinely complex, architecture-level decisions, that's still true.
But most AMS tickets aren't architecture-level. They're documentation-level. "How does this calculated field interact with that earning?" "What's the correct security group for this business process?" "Is this behavior a bug or expected?" These are questions with known answers buried in admin guides, community threads, and release notes. The knowledge exists. The team just can't find it fast enough to avoid filing a ticket.
So the AMS contract becomes a tax on findability. You're not paying for expertise — you're paying for someone else to search for answers you could have found yourself, if the search experience weren't so broken.
After Mando
Mando became the team's first line of inquiry for the documentation-level questions, configuration checks, and "is this a bug or expected behavior" lookups that used to consume AMS tickets. When an admin hits a question, they check Mando first. If Mando answers it — and it usually does — no ticket is filed. No slot is occupied. No latency is introduced. The question goes from open to resolved in minutes instead of days.
The team still uses AMS for genuinely complex issues — the kind that require deep architectural expertise and hands-on configuration inside the tenant. But those are the exception, not the default. The ratio flipped. What used to be "file a ticket for everything, self-serve when you can" became "self-serve for everything, file a ticket when you must."
The Results
At any given time, the team has 7 of their 10 ticket slots available — roughly $100K per year in AMS capacity that used to get consumed by routine questions, now preserved for work that genuinely requires outside expertise.
That reallocation is about to matter more than ever. Asahi Kasei has a public cloud migration scheduled for July. They're evaluating a security group rebuild that could span years. These are projects that demand sustained, focused AMS engagement — the kind of engagement that's now possible because routine questions aren't eating the allocation. Instead of arriving at their biggest initiatives having already burned through capacity on break-fix, the team can direct the full weight of their AMS contract toward the work that defines the next phase of their platform.
The team didn't abandon their AMS partner. They elevated the relationship. Mando handles the first line of inquiry. AMS handles the exceptions. We expect more teams to move toward this model as AI continues to lower the barrier to action — where the admin isn't waiting on an outside team to answer a question they could answer themselves in two minutes, and the AMS relationship is freed up to do genuinely high-value architectural work instead of serving as an expensive search engine.
Use case
03
From Reactive Admin to Strategic Operator
There's a version of the Workday admin role that's purely operational. Tickets come in. You fix things. You maintain configurations. You keep the system running. It's important work, but it's reactive by definition. The system dictates your priorities. Your calendar is shaped by what breaks.
There's another version of the role — the one that actually moves the business forward. You evaluate new modules. You optimize existing workflows. You bring proposals to leadership that make the platform more valuable. But that version requires something most admin teams don't have: the bandwidth to validate whether their ideas are feasible before putting them on the table.
Before Mando
Christian wasn't short on ideas. He had a running list of optimizations, feature evaluations, and workflow improvements he wanted to pursue: Engagement Builder for learning campaigns, Skills Cloud, payroll streamlining, a potential multi-year security group rebuild. His HR leadership recently made it a goal to identify and document critical payroll processes for the fiscal year. The ambition existed at every level of the organization.
The problem was the distance between having an idea and being able to stand behind it.
Every idea required validation before it could become a proposal. Does the module work the way I think it does? Has anyone implemented this at a similar organization? What are the downstream implications? Before Mando, answering those questions meant hours of research — hours that competed directly with break-fix work, AMS ticket management, and every other demand on a four-person team running a global enterprise platform.
So the ideas sat. Not because they were bad. Not because leadership wouldn't have been receptive. Because the time required to pressure-test them didn't exist in a schedule already consumed by daily operations. The strategic backlog grew while the team stayed pinned to the operational treadmill.
This is the hidden cost that never shows up in a productivity metric. It's not about hours lost — it's about proposals that never get made. Improvements that never get surfaced. A platform that stays static not because the team lacks vision, but because they lack bandwidth to convert vision into something they can walk into a meeting and defend.
Meanwhile, internal documentation coverage sat at roughly 3%. Christian described it as "virtually non-existent." Not because he doesn't value documentation — he actually enjoys writing it and sees it as a way to force clarity on definitions. But the team barely has time to build, let alone go back and document what they've built.
After Mando
When validation drops from an hour to two minutes, the threshold for what's worth exploring changes fundamentally.
Christian doesn't shelve the Engagement Builder idea because he can't afford the research afternoon. He doesn't wait for the next quarterly planning cycle to float the payroll optimization. He verifies feasibility in Mando — in minutes — and shows up to the conversation having already confirmed that the approach is grounded in real implementation patterns. The proposal arrives pressure-tested instead of tentative. The meeting becomes a decision point instead of a scoping session.
This is the pattern we see forming across our customer base. The best Workday teams already have the strategic instincts. What they don't have is a low-friction way to act on them. When you give a sharp admin a tool that answers in seconds what used to take an hour, you don't just save time — you unlock a version of the role that was always there but couldn't surface through the research overhead. The admin becomes an operator. The platform starts compounding.
And the 3% documentation gap — the one that leadership has identified as a fiscal year priority — now has a path to resolution. Mando's upcoming auto-documentation capabilities are built for exactly this problem: a browser extension that captures configuration workflows as they happen, generates documentation automatically, and makes that documentation searchable and retrievable through the same interface the team already trusts. For a team that barely has time to build, let alone document, the difference between "documentation is a goal" and "documentation happens for free" is the difference between a permanent backlog and a solvable problem.
When asked whether Mando is worth $250 a month, Christian didn't hesitate:
"For myself, absolutely. For my seat, I think we're maximizing the most that we can from that."
He drew the contrast with the AMS model unprompted. A traditional support partner will, at best, stay the same quality over time — and often degrades as they scale and arbitrage the resources they put in front of you. A software product built on improving AI models is the opposite.
"We're not buying into something that is going to stay stagnant. This is the complete opposite. I'm seeing this thing get better in real time."
The Results
The team's relationship to their own work changed. They went from keeping Workday running to pushing it forward. Ideas that had been sitting in backlogs because validation was too expensive started making it into leadership conversations — faster, better-supported, and with fewer open questions. The strategic backlog started shrinking, not because anyone deprioritized it, but because the team finally had the bandwidth to work through it.
That's the shift that matters most. Mando didn't just make Asahi Kasei's Workday team faster. It made them more valuable. And Christian knows it's just the beginning.
"This is the floor. And I know it's only going up from here."
