DCHS Insights CX Human-Centered Design @ DCHS PROTOTYPE no real data Customer journey mapping
What it is
A visualization of the end-to-end process a client goes through to accomplish one goal — built with frontline staff and clients with recent experience of one specific program, grounded in real research. The April 2025 maps were made by leadership for a general journey; the power of the method comes from doing it narrow and evidenced.
"Where, step by step, does the experience break down — and how does it feel?"
What it tells you and what it cannot
| Tells you | Cannot tell you |
|---|---|
| Where the pain points and emotional lows sit, in sequence | How frequent each pain point is — that needs a quantitative pairing |
| A shared cross-role picture executives read in seconds (the emotion curve) | The backstage causes — that is a service blueprint |
| Which fixes matter most, tied to owners | Anything reliable, if built from assumptions instead of research |
How it runs
| Who does the work | Participants | Elapsed time | Cost |
|---|---|---|---|
| A facilitator/designer + a cross-functional group | 5–15 in the room; synthesizes 5–15 client interviews | 1–2 days with research in hand; 2–6 weeks full cycle | $$ — staff and client time, plus research and facilitation |
Strengths, watch-outs and quality signals
aligns the organization around the client's lived reality; makes problems legible instantly; converts pain points into owned, prioritized opportunities.
"fantasy maps" built from assumptions; over-generalizing one persona's journey; and failing to assign priority opportunities to owners.
grounded in real research with verbatim quotes; a genuine numeric emotion curve; every top pain point mapped to a fix with an owner and a metric.
Equity & consent
Whose journey gets mapped is the equity decision: map the hardest journeys (limited English proficiency, disability, first-time applicants), not the average. Mixing clients, frontline staff, and supervisors in one room can silence the clients — consider separate sessions and consent for quotes and photos.
When to choose it
to synthesize evidence from other methods into a shared view of a journey.
Nielsen Norman Group's journey-mapping structure (lens → experience → insights); Civilla's Project Re:Form, which used hundreds of hours of fieldwork to cut Michigan's 42-page benefits application to 18 pages.
Scenario: How this can help examine HR1 impacts
Journey mapping can be the continuity spine for evidence about the affected cohort. It places SNAP and Medicaid requirements on one lived timeline: notice received → applicability or exemption understood → proof gathered → submitted → accepted → benefit retained, lost, or restored → alternative offered → contacted → enrolled → service received. That sequence makes duplicated proof, emotional lows, missed compliance steps, and uncertain ownership visible across program boundaries.
Measure elapsed time, retellings, evidence requests, submission and acceptance separately, treatment gaps, and the difference between a referral being offered and help being received. The HR1 scenario tests a reusable screen, consented proof packet, shared deadline view, pre-closure outreach, and closed-loop alternative tracking within DCHS's service-coordination role.
Blind spot: a map is synthesis, not prevalence or causal proof. If it is built only from staff assumptions or one composite persona, it becomes a confident-looking fiction. Ground a real version in client evidence, case review, and operational logs. See the HR1 listening plan and HR1 Impact Report.
Customer journey map — Janice Brown, crisis entry to stability
Scope & method
One persona, one journey, one goal.
Persona tile for Janice Brown, a composite client who entered through the DV hotline and touches seven programs.
Janice Brown
Housing insecurity and fragmented services complicate one mother's efforts to find stability and reunite with her children.
Goal: Stability that unblocks reunification
Scenario. Janice left a violent home and is stabilizing several needs at once: immediate safety, housing, food and benefits, behavioral health support, and a CPS matter that cannot move toward reunification without stability. Goal: a stable footing that unblocks reunification. She is a protective parent who acted under pressure, continues seeking help through exhausting application and appointment processes, and is clear about what she can realistically manage — the map treats those strengths as data, not as garnish.
Evidence base (as-if). This mock is written as the output of a half-day mapping workshop with 6 frontline staff from the programs on her path plus synthesis of 9 client interviews with recent crisis-entry experience. A journey map without underlying research produces fiction; a real version of this artifact must be grounded in real interviews and field evidence before it is used to prioritize anything.
The journey at a glance
The map below is the workshop's central artifact: six stages, standardized department-wide, with what happens, where it happens, how it feels, what grinds, and which indignities attach at each stage — and the emotion curve running underneath on the same columns. Timeframes are relative to the night she left.
Six-stage customer journey map from crisis entry to an unresolved stability loop, with indignity tags on every stage and an emotion curve beneath that starts at -3, never rises above -2, and dips back to -3 at the retelling stage and the housing loop.
Night 0 — the night she left
Days 1-3 — safety first
Weeks 1-2 — the story again
Weeks 2-5 — separate clocks
Weeks 4-6 — one miss echoes
Day 47 and counting — the loop
Arrives via the DV hotline and enters a safe house during an immediate crisis.
Safety, a bed, food. Told about resources, handed numbers.
Separate approaches for housing, benefits, behavioral health, and the CPS matter. Each requires the story again.
Services start on separate clocks. Nobody holds the whole picture; appointments collide.
Progress in one area; a missed appointment elsewhere sets it back. No coordinated view.
Housing remains unresolved, so the reunification process cannot move forward.
DV hotline / safe house
Safe house and phone calls
Multiple programs
Program offices
Calls and offices
Housing queue / CPS
A night of disclosure and no whole-system plan
Multiple numbers to call while in crisis
9 retellings across 7 doors
Conflicting appointments and forms
A missed appointment becomes a closure risk
47 days to first stable service
3. Service Connection 9 retellings across 7 doors — this dip is administrative, not clinical
6. Stability & Transition The curve ends where it began: housing loops back to reunification
The curve never crosses −2, and the workshop's reading of it is the report's headline: the deepest lows after the entry night are system-made — the retelling stage and the unresolved housing loop — not features of the underlying crisis.
Her scoreboard across this journey, from the persona record (illustrative, not measured caseload data): 7 doors · 9 retellings · 6 forms · 47 days to first stable service · needs met 3 of 8.
Pain-point deep-dives
Four pain points carried the workshop's attention. Each traces back to a stage on the map above.
The story restarts at every door — Stage 3, the deepest system-made dip
Phone photo of a blank intake form on a clipboard at a reception counter.
Intake counter, mid-morning: a fresh clipboard with the same demographic page she has completed five times
Between the hotline, the safe house, housing intake, benefits, behavioral health, and the CPS matter, interview participants described giving the same account of the worst night of their lives eight to ten times; the persona record carries 9 retellings across 7 doors. At the third door in nine days, one participant put it in five words: "I already told somebody this."
Taken one at a time, every program's intake is locally reasonable — each needs its own record, its own consent, its own risk assessment. It is the sum that retraumatizes, and the workshop judged that sum the single strongest driver of the curve's second dip. Nobody designed the repetition; that is precisely why no single program can fix it.
Appointments collide — Stage 4
Client quote: I can do the appointments. I cannot do them all at once.
I can do the appointments. I cannot do them all at once.
In the mapped fortnight, three programs independently scheduled required contacts into the same two weekday windows, and two collided outright. No one holds the whole calendar, so compliance with one program manufactures absence from another. The point the workshop kept returning to: she can meet any one program's expectations. It is the set that fails — and the set is ours, not hers.
The dependency nobody can see — Stage 6
Phone photo of a folded letter and notebook on a made bed in a plain room.
A borrowed room at the safe house: a folded program letter on a made bed
Reunification requires stable housing, but housing is managed through a separate queue and no plan of record connects the two. Workshop participants identified this dependency when they drew the staff lanes side by side. It was not visible in any single program's view.
Client quote: If housing is the thing everything depends on, why is it the last thing moving?
If housing is the thing everything depends on, why is it the last thing moving?
A miss reads as refusal — Stage 5
A behavioral-health appointment missed because of a same-morning benefits appointment appears in the record as a no-show, and a second miss would trigger closure review. The schedule shows a collision, but the record classifies the missed appointment as noncompliance. That classification may affect whether her case remains open.
Prioritized opportunities
Zone C — every pain point maps to an owned, measurable change. Placement judgments are the workshop's.
Severity-by-effort matrix: making the housing-CPS dependency explicit and collision-aware miss codes are quick wins; the traveling intake and shared calendar view are major projects.
S = Severity / impact (1 low - 5 high) · E = Effort to fix (1 low - 5 high)
Findings table linking the four pain points to owned recommendations with metrics.
| Finding | Recommendation | Owner | Evidence · metric |
|---|---|---|---|
| The story restarts at every door (9 retellings, 7 doors) | Pilot one consented intake narrative that travels across programs | Intake redesign workgroup | Journey strip stage 3; workshop interview synthesis Retellings per new multi-program case |
| Appointments collide across programs | Stand up a cross-program appointment view and sequence new contacts against it | Program supervisors, jointly | Stage 4; two collisions in the mapped fortnight Collisions per client per month |
| Housing-reunification dependency is invisible | Name the dependency in both the housing and CPS plans and review it jointly | Housing and CPS program leads | Stage 6; surfaced only when staff lanes were drawn together Multi-program cases with documented dependencies |
| Missed appointments read as refusal | Add a collision-aware miss code and require it before any closure step | Records / QA lead | Stage 5; no-show coding precedes closure review Share of misses coded with a documented reason |
Limitations & quality self-check
Limits of this map. One composite persona's journey, reconstructed retrospectively — recency and memory bias apply, and one journey does not generalize to all clients. This mock's evidence base is synthetic by design; a real map must be built from real interviews, field observation, and frontline participation before its priorities are acted on.
Open research questions the map surfaced but cannot answer (deliberately left unresolved):
- "Which retellings are legally required and which are habit"
- "What information can travel with consent today"
- "Which missed appointments could be prevented by sequencing"
Self-check. Emotion curve numeric on the −3..+3 scale: yes. Every pain point mapped to an owned fix with a metric: yes. Quotes verbatim-style and tagged: yes. All fabricated figures marked: yes. Banner present: yes.
Janice's HR1 proof-and-transition journey
Scope & method
This map is the continuity spine for the four-report HR1 package. It follows one narrow journey: from the first benefit notice through an affected 2027 renewal, any benefit lapse or restoration, and the completion—or failure—of an alternative-service connection. It combines a synthetic Janice interview, her synthetic 28-day diary, a synthetic 12-case QSR round, and synthetic two-week staff friction logs. A real map would require real, consented evidence from the affected cohort.
Janice Brown post-HR1 persona tile describing her benefit, safe-house, behavioral-health, CPS, and housing context.
Janice Brown (post-HR1)
SNAP and Virginia Medicaid Expansion recipient in a safe house; four children with kinship caregivers during a CPS matter, so no child exemption reaches her; her SNAP work-requirement status governs both programs.
Goal: Understand what applies, prove it once where possible, preserve treatment, and reach usable help if coverage changes
Workshop participants used one common outcome chain: notice received → affected status or exemption identified → proof available → submitted → accepted → benefit retained, lost, or restored → alternative offered → contacted → enrolled → service received. The map deliberately splits steps that administrative systems often collapse.
The journey at a glance
The current-state path spans September 11, 2026 through February 8, 2027. Dates and outcomes match the Janice record in the other three HR1 reports.
Six-stage synthetic HR1 journey for Janice from two notices through understanding, proof gathering, submission, short benefit lapses, restoration, and completed alternative connections, with emotional lows at proof gathering and benefit lapse.
Sep 11–15, 2026
Jan 8, 2027
Jan 12–14
Jan 15–28
Jan 22–Feb 8
Jan 22–Feb 8
SNAP and Medicaid notices reach the safe house on different days.
An assistance call establishes that she is a non-exempt SNAP ABAWD. Nobody asks whether she is unfit for work, and nobody notes that her SNAP status is the only reason the Medicaid work requirement does not apply to her.
Janice requests provider documentation and gathers safe-house, household, and activity records.
A portal attempt fails; an assisted resubmission is received; the Medicaid renewal packet follows.
SNAP closes for unverified information, not for the time limit, then restores; while it is closed the Medicaid work requirement starts applying to her; Medicaid displays a short gap; a treatment visit is rescheduled.
Review restores both benefits; food and medical-navigation pathways reach service received.
Safe-house mail
Phone and benefits navigation
Safe house and behavioral-health clinic
Portal, Customer Relations, benefits
Eligibility systems, clinic, CPS plan
Review channels and referral network
Two clocks start before their combined meaning is clear
The determination that governs both programs is made without being named as such
Evidence depends on other people and requires another account of trauma
Receipt, match, and acceptance are not one status
One unresolved proof state ends food and opens a second work rule
A referral needs several owners before it becomes help
4. Proof submitted Continuity point: submitted, received, matched, and accepted are distinct.
5. Decision or lapse Cascade boundary: sequence and risk are visible, but causation is not assumed.
6. Restoration or alternative Connection rule: offered, contacted, enrolled, and service received are separate.
The curve has two equal lows. Stage 3 is the human burden: treatment and trauma must be turned into evidence. Stage 5 is the consequence risk: a proof item that has been transmitted but not accepted can affect food, coverage, and scheduling before anyone owns the mismatch. Stage 6 rises only to +1 because restoration does not remove the next six-month renewal or resolve housing.
Pain-point deep-dives
The notice arrives where stability is temporary — Stage 1
Generic benefits envelope and unreadable notice on a bedside table in an anonymous temporary room.
An unreadable benefit notice and sealed envelope on a small safe-house bedside table
The first notice arrived September 11 and the second September 15. Safe-house staff could document receipt without interpreting eligibility. The program clocks existed, but no one owned a cross-program explanation. A reusable affected-status screen would not change the rules; it would make the two sets of questions, evidence, deadlines, and owners visible together.
Proof gathering turns care into paperwork — Stage 3
Janice requested behavioral-health documentation January 12. A provider had to decide what could be disclosed with consent, prepare a letter, and route it to a system that would make a separate sufficiency determination.
The state was required to look first. The rule says it must use information it already holds — payroll data, adjudicated claims, encounter data — before asking someone for documents, and until December 31, 2027 a signed statement is enough for medical frailty. Janice is Medicaid-enrolled and in treatment, so the claims data exists. Nobody looked, and nobody wrote down that nobody looked. Janice spent four days carrying paper between her clinic and the eligibility system to produce a document the process may not have needed.
Janice asks who needs the paper, what it must say, and whether it was received.
I already told somebody this. I need to know who needs the paper, what they need it to say, and whether they got it.
Stage 3 evidence-gathering reconstruction
The proof packet has no single status — Stage 4
Organized generic proof packet with unreadable documents, consent card, deadline checklist, and phone receipt.
A staged consented proof packet: covered documents, owner checklist, deadline card, and generic receipt screen
The January 15 portal failure, January 16 assisted resubmission, January 18 receipt, and January 22 SNAP closure show why “uploaded” is not an outcome. The packet needs a four-state audit trail—submitted, received, matched, accepted—plus the responsible role and next date. The client should be able to see the same plain-language status.
A missed compliance step can become a service cascade — Stage 5
In the synthetic sequence, SNAP lapsed January 22 and restored February 5; Medicaid displayed a February 1–7 gap and restored February 8. A behavioral-health visit was rescheduled during coverage uncertainty, and that change was documented for the CPS service plan.
The closure reason matters, and it is easy to get wrong. Janice's SNAP case closed for unverified information, not for exhausting the three-month time limit. The distinction decides what restoration takes: a verification closure is cured by producing the evidence, which is why a February 5 restoration is possible at all, whereas a time-limit closure is only cured by 80 hours over a 30-day period, an exemption, or the end of the 36-month window. Both routes are live for her, and a record that does not say which one fired will mislead everyone who reads it later.
The closure did something less visible too. While SNAP was closed she was no longer a SNAP household member, so the exclusion at 42 CFR 435.554(c)(7) ended and the Medicaid work requirement started applying to her for the first time. This case illustrates a plausible SNAP closure → Medicaid requirement → coverage → treatment → service-plan → reunification risk. It does not establish that the first event caused the others or that every client will experience the cascade.
Pre-closure outreach checks for unmatched evidence, warns that closing SNAP would start the Medicaid requirement, and coordinates a continuity plan while the determination is made. Where a closure does go ahead, she has ten days to request a hearing and keep benefits running. Mail reaches a safe house late, so that date needs saying out loud, not only printing.
Offered is only the first referral state — Stage 6
The food pathway moved from offered January 22 to contact January 23, intake and groceries received January 24. Medical-navigation help moved from offered February 1 to contact and case opening February 2 and appointment help received February 3. A workforce option was offered but Janice did not pursue it during the review window. All three facts belong in the report; only two reached service received.
Prioritized opportunities
The workshop prioritized changes DCHS can make without implying control over federal or state eligibility standards.
Synthetic severity-effort matrix prioritizing a reusable HR1 screen, proof-state vocabulary, and pre-closure outreach as quick wins, with proof packets, deadline views, and closed-loop referrals as larger projects.
S = Avoidable burden reduced (1 low – 5 high) · E = Effort to implement (1 low – 5 high)
Five synthetic journey findings with owners and metrics for screening, proof packets, pre-closure review, deadlines, and referral completion.
| Finding | Recommendation | Owner | Evidence · metric |
|---|---|---|---|
| Program notices became one unowned client task | Use one screen to distinguish program, requirement, exemption question, proof, deadline, and owner | Benefits policy and Customer Relations | Separate September arrival and January combined explanation Affected records with all six fields completed |
| Proof gathering depended on provider workflow | Build a consented proof packet that records source, minimum necessary disclosure, and reuse permission | Behavioral health with privacy lead | Provider request January 12; assisted submission January 16 Duplicate evidence requests per affected case |
| Receipt did not prevent action | Trigger named pre-closure review when submitted proof remains unmatched | Benefits operations | Receipt January 18; SNAP lapse January 22 Unmatched submissions resolved before closure date |
| No one held every deadline | Pilot a consented shared deadline view with contact preferences | Cross-program operations group | Benefit, provider, treatment, and CPS dates stored separately At-risk cases with a current deadline view and named next-step owner |
| Two referrals became services; one remained an offer | Track offered, contacted, enrolled, and received and honor a client decision not to pursue | Referral network lead | Food and medical navigation received; workforce option not pursued Stage conversion and client-declined status by pathway |
The sequence for implementation is: agree on definitions first; pilot the screen and proof states; add pre-closure outreach; then test the more complex shared packet, deadline, and closed-loop infrastructure. Client testing must include people who cannot or do not use a portal.
Limitations & quality self-check
- One prospective composite journey. This map is not a prevalence estimate or a finding about actual Janice-like clients. A real version needs recent client interviews, record review, and frontline evidence.
- Synthesis can conceal disagreement. Program rules and staff interpretations should remain in separate lanes until an authorized source resolves them.
- Sequence is not causation. The map shows when a coverage gap, treatment change, CPS documentation, and housing barrier coexist. It does not prove one caused the next.
- Integrated steps are hypotheses. A shared screen, proof packet, deadline view, and outreach can reduce avoidable burden; none guarantees eligibility, an exemption, or benefit retention.
- Unknowns preserved, unresolved. “Whether Janice is physically or mentally unfit for employment under SNAP rules.” “Whether proof is matched ex parte, as the rule requires, before staff ask her to submit it again.” The map keeps these as research and eligibility questions rather than filling them with a preferred answer.
- Where the rule, not the map, supplies the answer. Two questions this map previously left open are in fact settled by the regulatory text and are now stated rather than hedged: no child exemption reaches her, because none of her children lives with her; and while SNAP continues she is excluded from the Medicaid work requirement. Both are determinations DCHS records, not opinions the map holds.
- Quality checks. Numeric −3 to +3 emotion path present; six stages match the persona; every priority has an owner and metric; receipt and acceptance are distinct; referral rungs are distinct; all fabricated evidence is marked.