Case Study · Operating-Model Strategy × Systems Research
When disconnected workflows became an operating-model decision
How foundational research on a fragmented managed-service platform was reframed for leadership. Not as a UX problem, but as a margin and operating-model one. It turned a scatter of small fixes into a single decision the business could act on.
Role
Sole UX Researcher Strategy Partner
Context
Confidential ad-tech DSP managed service
Output
Vision & strategy Operating-model case
Impact
Leadership decision Phased roadmap
UX finding → margin problem
Fragmentation was reframed from a usability issue into a measurable tax on operator capacity, advertiser retention, and growth. This is the language leadership funds against.
Scattered fixes → one decision
A diagnosis spread across roles, tools, and handoffs was distilled into the single choice only leadership could make: name an end-to-end owner.
Strategy → shipped
The roadmap that followed has already moved early wins into production. It has cut weeks-long setup down to days and replaced redundant re-entry with connected workflow.
The research had already been done: personas, journey maps, a workflow analysis that showed a managed-service operation stitched together by hand. The question leadership was really asking wasn't how do we improve the experience. It was where is this costing us, and what do we do about it?
01 · The symptomOne outcome, fragmented across the whole operation
On the surface, the request was a familiar one: the workflows for delivering a managed advertising campaign were disconnected, and teams wanted them smoothed out. But the foundational research had already surfaced something bigger than a workflow problem. A single advertiser outcome, a campaign delivered well, was being produced by a relay of specialists, each working in their own tool, each handing context to the next by hand.
An account manager set the intent. A campaign manager built it. A trader optimized it. An analyst reported on it. Quality and operations caught what slipped. Between each of them sat a manual handoff and a different system of record: the DSP console, bidding tools, reporting and BI, CRM, and ticketing. No one owned the end-to-end outcome. The campaign goal was real, but the system serving it was a baton pass that dropped context at every exchange.
One advertiser outcome, fragmented across the relay
Abstracted · roles & tools generalized
Intended
One advertiser outcome: a campaign delivered well
fragments into a relay of specialists
Step 01
Account Manager
CRM
Sets the brief and the advertiser's intent.
→
context by hand
Step 02
Campaign Manager
DSP console
Builds and traffics the campaign.
→
context by hand
Step 03
Trader
Bidding tools
Optimizes pacing and performance.
→
context by hand
Step 04
Analyst
Reporting / BI
Reports results and builds wrap decks.
→
late catches
Step 05
QA / Ops
Ticketing
Catches by hand what slipped through.
Reality
Owned by no one end to end. No shared source of truth.
Roles and tools are generalized for confidentiality. The structural truth is the point: the same campaign object is re-entered and re-interpreted across several roles and systems, with no single owner and no shared source of truth. Each handoff is a place context is lost and rework begins.
02 · The reframeNot a UX problem. A margin problem
A research readout describing disconnected workflows would have earned a polite nod and no budget. The work that mattered was translating the finding into the language leadership actually funds against. Fragmentation isn't an inconvenience; it's a tax. Every manual handoff is operator time spent re-entering context instead of serving advertisers. Every dropped detail is rework, or an error caught late. Every silo is a ceiling on how many accounts one operator can carry, which is the ceiling on how the business scales.
Reframed this way, the question stopped being "how do we make the workflow nicer" and became "how much growth, capacity, and retention is the current operating model leaving on the table."
Original ask
Smooth out the disconnected workflows so teams can move campaigns through more easily.
Research reframe
Treat fragmentation as a measurable tax on operator capacity, advertiser retention, and growth. It is an operating-model problem, not a UX one.
Strategic shift
The conversation moved from improving a workflow to deciding who owns the end-to-end outcome and what system holds the truth.
03 · The strategyFive patterns, five levers, mapped one for one
The diagnosis had to resolve into something a leadership team could act on without trying to fix everything at once. I distilled the fragmentation into five systemic patterns, and answered each with a single strategic lever, deliberately mapped one for one, so the strategy clearly answered the diagnosis rather than floating beside it.
The pattern-to-lever mapping
Strategy answers diagnosis, one for one
Pattern
Work is a relay of disconnected tools
The same campaign object is rebuilt and re-entered across several systems.
→
Lever
Unified workflow
One connected flow with a shared source of truth for the campaign object.
Pattern
Manual effort fills every gap
Operators hand-stitch handoffs, status, and reporting between roles.
→
Lever
Intelligent automation
Automate the connective tissue so operator time goes to judgment, not re-entry.
Pattern
The system only looks backward
Teams react to what already happened instead of anticipating what's next.
→
Lever
Predictive & proactive
Surface risks and opportunities early, before they become rework.
Pattern
Errors are caught late
Quality depends on people noticing problems downstream, after they cost something.
→
Lever
Error prevention over detection
Build guardrails that stop errors at the source instead of finding them later.
Pattern
Client-facing work is rebuilt each time
Reporting and deliverables are reassembled by hand for every campaign.
→
Lever
Client-ready by default
Make the system produce advertiser-ready outputs as a byproduct of the work.
The most persuasive single artifact in the strategy. Each lever sits directly across from the pattern it answers. The mapping is what turned a diagnosis into a strategy leadership could believe. The fix clearly resolves the problem rather than sitting beside it.
04 · The roadmapSequenced so it could actually be funded
Five levers all at once is not a plan; it's a wish. The strategy was sequenced into horizons so leadership could fund a first step, see it land, and decide on the next. That lowers the cost of saying yes.
Now
Stop the bleeding
Low-cost error-prevention wins and the instrumentation needed to baseline what the current model actually costs.
Next
Connect the flow
Unify the core workflow around a shared source of truth and automate the highest-friction handoffs.
Later
Get ahead of the work
Predictive capability and client-ready-by-default outputs, building on the connected foundation.
05 · The askThe one decision only leadership could make
The deck didn't end on a metric. It ended on a decision. The single thing that makes everything else stick is not buildable by a product team. It has to be authorized. Naming an end-to-end owner is the structural change that turns a relay back into a system.
Decision 01
Name an end-to-end owner
One accountable owner for the campaign outcome across roles. This is the change only leadership can authorize, and the one that makes the rest hold.
Decision 02
Designate the system of record
Choose the single source of truth for the campaign object so the relay stops re-entering and re-interpreting it.
Decision 03
Fund the first horizon
Resource the Now wins and the instrumentation to baseline success, with a checkpoint before committing to Next.
06 · What changedFrom a research finding to a funded direction
The work moved the organization from a stack of disconnected observations to a shared understanding of the system and a decision it could act on. Since then, early wins from the roadmap have shipped, connecting parts of the workflow that used to depend on manual handoffs and automating work that operators previously did by hand.
Decision impact
Fragmentation became fundable. Reframing the problem as an operating-model and margin issue gave leadership a reason to act and a single decision to make, rather than a list of workflow complaints to triage.
Operational change
Setup measured in weeks now takes days. Early automations and workflow connections have collapsed campaign time-to-live and removed redundant re-entry of the same object across tools.
Quality shift
From catching errors to preventing them. Guardrail work moved quality upstream, stopping errors at the source rather than finding them late, where they cost the most.
Strategic alignment
One outcome to optimize for. Product, engineering, and operations gained a shared target, the connected advertiser outcome, instead of each improving their own silo in isolation.
07 · ReflectionThe contribution wasn't the finding. It was the frame
The foundational research mattered, but it wasn't the hard part. Plenty of organizations can see that their workflows are fragmented. What they usually can't do is turn that observation into a decision: to translate a pile of friction into a single, fundable choice that someone with authority can actually make.
The value here wasn't a prettier journey map or a longer list of pain points. It was reframing a UX finding as an operating-model problem, mapping the diagnosis to a strategy one for one, and sequencing it so leadership could say yes to a first step. The research told the organization what was wrong. The reframe told it what to do.
A request to fix disconnected workflows became a decision about how the business operates: who owns the outcome, what holds the truth, and where to invest first.