Self-Initiated · UX Design

First
Service

A power-user production tool, run by a volunteer covering Sunday at the last minute. This project redesigns the first hour at the controls so a first-timer can run a live service without fear of breaking it.

Projected task success rate
Before
~55%
Target
90%+
Project Snapshot

Everything a reviewer needs, at a glance

ORGANIZATION

Renewed Vision

Role

UX Researcher + Product Designer

Target Users

Last-minute or unexperienced users of Pro Presenter

Duration

Three weeks

Team

Self

Tools

Figma · VS Code · Claude

The Problem

Low confidence with first-time users

ProPresenter is professional-grade live production software trusted in major venues, yet most of the people running it on a given Sunday are volunteers with minutes of preparation. The recurring complaint in public reviews is not that the tool lacks power. It is that a first-timer cannot reach that power safely while a room watches.

Research

Understanding where the learning gap is.

The same software is praised and resented for the same reason. Across public reviews one contradiction repeats: the depth people love is the depth that strands a first-timer. Capability is not the problem. The climb to use it under pressure is.

Research goals, assumptions & questions

Goals

  • Locate the specific moments where a first-time operator loses confidence.
  • Separate genuine learning gaps from interface defects that can be designed out.
  • Identify the single highest-leverage change rather than a longer list of fixes.

Assumptions to test

  • Volunteers are motivated to serve; motivation is not the bottleneck.
  • Most distress clusters around going live and recovering from a mistake.
  • The full power-user surface is overwhelming as a first surface, not as a tool.
Method: two tracks
✓ Secondary research, completed

A corpus of verified public reviews on Capterra and GetApp, plus Renewed Vision's own Volunteer Operator's Guide. Reviews were coded for the moment of failure and the emotional language used. This grounds the project in real operator voices rather than assumption.

◆ Primary research, protocol designed

A planned interview study across four groups: volunteer operators, worship leaders, church media directors, and experienced operators. Five to eight participants per round, semi-structured, focused on the first-service narrative. Designed and ready to run; results would replace the projected baselines below.

Voice of the operator · verbatim, from verified reviews

“it can be very daunting to know where to start”

Verified review · Capterra

“give a crash course before service to fill in”

Church media volunteer · GetApp

“WAY to complicated for the average person”

Ron B., worship leader · Capterra
Competitive analysis · onboarding & error behavior across the category

Five tools operators encounter in live-production and worship settings, rated qualitatively against the dimensions that matter for a nervous first-timer. Ratings are a heuristic assessment, not benchmarked scores.

ToolLearnabilityDiscoverabilityOnboardingError preventionRecovery
ProPresenterLowMediumVideo onlyWeakUnclear
OBS StudioLowLowCommunityWeakScene reset
vMixLowMediumDocsMediumMedium
EasyWorshipHighHighGuidedMediumMedium
Stream Deck workflowsHighHighPhysicalStrongStrong

Read across: EasyWorship trades power for approachability, and Stream Deck succeeds by reducing a service to a few large, labeled, unmistakable buttons. The opportunity is to keep ProPresenter's depth while borrowing that approachable, recoverable first surface.

Affinity mapping · from coded reviews to opportunities
Overwhelm at launch11 notes
“daunting to know where to start”
“too many buttons”
“froze the first time”
OpportunityReduce the first surface, not the product.
Fear of a live mistake9 notes
“terrified I'd put up the wrong thing”
“blacked the screen by accident”
“no idea how to undo it”
OpportunityMake the safe path the obvious path, live.
No safe place to learn7 notes
“crash course right before service”
“first run was the live run”
“only learned by watching someone”
OpportunityBuild confidence before Sunday, not during it.
02b Stakeholders & ecosystem

The person at the controls is not the person who bought the software.

ProPresenter is purchased by a media director or pastor, configured by a technical lead, and then operated week to week by whoever is available. That gap between buyer, configurer, and operator is where onboarding either holds or fails. Mapping it first keeps the rest of the work honest about who actually feels the problem.

Stakeholder map · influence and exposure to the onboarding problem
Exposure to the first-service problem → Control over the tool → sets it up, rarely operates decides AND feels it peripheral lives the problem, low control Volunteer operator Media director Technical lead Worship leader Congregation RV support team configures for → recruits absorbs the support calls

Red = highest exposure to the problem. The volunteer carries the most exposure and the least control, and sits furthest from the people who could fix the setup. Onboarding is the only lever that reaches them directly.

User ecosystem · one service, many hands

A single Sunday touches every role on the map. The operator’s confidence is shaped upstream by how the technical lead configured the file, and it is felt downstream by everyone watching the main screen. Onboarding is not a private tutorial. It is the seam that holds a volunteer-run production together.

Business impact · why a confident first hour pays back
Adoption
Churches keep the tool they can staff

A product a volunteer cannot run on week one is a renewal risk for the whole organization, not one user.

Retention
Volunteers return when they did not fail in public

The cost of a lost volunteer is recruiting and training the next one, every few weeks.

Support burden
Fewer panicked Saturday-night tickets

First-run confusion and live-mistake recovery are predictable, repeatable support themes that onboarding can absorb.

Brand & trust
Reliability is visible to the whole room

A mistake on the main screen is seen by hundreds. Confidence at the controls is a brand surface, not a private setting.

These links are a working model of how onboarding connects to outcomes, drawn from the support and review themes below, not measured revenue figures.

03 Problem statement

One sentence the rest of the project answers.

A volunteer operator with minutes of preparation must run a live service inside a power-user interface, where the controls that advance the service sit beside the controls that can break it on the main screen, and where no obvious path returns the output to a safe state.

User pain

Overwhelm at launch, fear of a visible mistake, no rehearsal, no calm way back when something goes wrong.

Environment

Dark booth, live audience, last-minute hand-off, no expert present to ask.

Product limits

Dense default surface, destructive controls near navigation, system vocabulary, no recovery affordance.

Business impact

Lost volunteers, higher support load, and reliability concerns visible to the whole room.

04b Research toolkit

The instruments, so the findings can be trusted or challenged.

Secondary research across public reviews and Renewed Vision’s own Volunteer Operator’s Guide is complete and quoted earlier. The primary study below is a designed protocol, ready to run with a church partner. Showing the instrument matters as much as showing a result: a reviewer can judge whether the questions would surface real behavior rather than opinions.

Interview guide · 30 minutes, contextual, behavior-first
Warm-up & context
5 min · lower the stakes, establish the real setting
  1. Walk me through the last service you ran. Where were you standing, who handed it to you, how much notice did you get?
  2. How did you learn the software? Listen for: nobody taught me, I watched once, I just clicked.
The first hour
8 min · locate the confidence drop
  1. What was the first thing you weren’t sure about?
  2. Was there a moment you were afraid to click something? What was it, and what did you think would happen?
  3. Show me what you do to get a service ready. Observe rather than ask. Note hesitation, backtracking, mouse hovering.
Going live & recovery
10 min · the two peak-anxiety events
  1. What goes through your mind right before the first slide goes on the main screen?
  2. Tell me about a time something went wrong live. What did you do? How did you get back? Listen for: I froze, I asked someone, I restarted.
  3. If you could undo one thing instantly during a service, what would it be?
Practice & close
7 min · rehearsal behavior and wrap
  1. Do you ever practice before a service? What stops you, or what would make it easy?
  2. If you were training the next volunteer, what is the one thing you’d warn them about?
  3. Anything I should have asked but didn’t?

Open, non-leading, weighted toward observed behavior over stated preference. Questions map directly to the two anxiety peaks the peak-end model predicts: going live and recovering.

Recruitment matrix · who to talk to and why
SegmentTargetRecruit criteriaWhat they tell us that others can’t
Volunteer operators
primary
6–8Ran < 10 services, no AV background, recruited by their churchThe first-hour confidence drop while it is still fresh, before it normalizes.
Worship leaders2–3Direct a live service that depends on the operatorThe cost of an operator mistake to the room and the service flow.
Media directors2–3Own the ProPresenter setup, recruit and schedule operatorsWhy volunteers drop, and what they configure to protect the screen.
Technical directors2Configure files and displays, present pre-serviceBackstage constraints and what a safe default would have to respect.
Experienced operators
contrast
2–350+ services, fluent in the full surfaceWhich early fears were real defects versus a learning curve that fades.

Recruitment plan, not a roster of completed sessions. The experienced-operator contrast group exists to separate genuine interface defects from a curve that simply takes time.

Mental model · operator expectation vs. product behavior
What a first-timer expects

There is a clear “start the service” path; the rest can wait.

Clicking a slide previews it before anyone sees it.

A visible way exists to take the screen back to safe.

The thing I touch most is the biggest thing on screen.

What the default surface does

Every capability is present at once, with no first path marked.

Clicking a slide can send it live immediately.

Recovery is possible but unlabeled and discovered under stress.

Navigation sits beside destructive controls at equal weight.

Every gap between the columns is a confidence cost. The redesign closes the four that cluster around going live and recovery, the highest-anxiety moments.

Confidence by stage · where the first hour breaks (projected)
Open the file62
Find the start40
Go live (1st slide)24
Mid-service jump34
Recover a mistake18
End on the logo48
holding wobbling breaking

Projected confidence index, modeled from review themes and the peak-end framework, to be calibrated by the study above. The two troughs, going live and recovering, are exactly where the redesign concentrates.

Competitive scorecards · the three jobs that matter for a first-timer
Discoverability can a new user find it?
First Service4.4
ProPresenter2.3
EasyWorship3.2
OBS1.5
vMix1.3
Learnability confident in one sitting?
First Service4.2
ProPresenter2.1
EasyWorship3.0
OBS1.2
vMix1.1
Recovery graceful when it breaks?
First Service4.3
ProPresenter1.9
EasyWorship2.5
OBS2.2
Stream Deck2.9

Heuristic scores (0–5) from a structured expert evaluation of each tool’s first-run experience, not user-test data. “First Service” is the proposed design scored against the same rubric to show the target, not a shipped result.

04 Personas

Who is at the controls.

Three evidence-based personas drawn from the review corpus. They are provisional until validated by the planned interviews, but each is anchored to language operators actually used. The design is built for the least confident of them.

M
Primary

Maria, the Sunday fill-in

“I just don't want to be the reason something goes wrong up there.”
Goals

Get through the service accurately and hand it back without incident.

Motivations

Genuinely wants to serve and be trusted to do it again.

Behaviors

Arrives early, asks for a quick walkthrough, holds the run of show in her head.

Frustrations

Too many controls, unclear what is live, no rehearsal, no expert on the day.

Technical proficiencyLow
Key workflow

Open this week's service, advance cues in order, return to the welcome loop at the end.

D
Secondary

Doug, the greeter who said yes

“They told me it was easy. It is not easy.”
Goals

Help out for one week without embarrassing himself or the team.

Motivations

Community and a sense of contribution, not an interest in the software.

Behaviors

Cautious, reads every label, stops the moment something looks unfamiliar.

Frustrations

System vocabulary, dense layout, the feeling that the tool was not built for him.

Technical proficiencyVery low
Key workflow

Advance to the next slide on cue and clear to logo between segments. Nothing more.

Experienced

The regular operator

Fluent, fast, and protective of the keyboard accelerators and deep show controls. Any simplification must not slow this person down or hide what they rely on.

Stakeholder

The media director

Owns the system, trains the team, fields the support calls. Wants volunteers who can run a service unsupervised and a clear path from beginner to confident.

Edge

The bilingual operator

Runs a Spanish-language service. The interface and the flow cannot assume English. A first-class bilingual path, not an afterthought.

05 Journey map

The first service, drawn as an anxiety curve.

Mapping the experience stage by stage exposes where confidence is lost. By the Peak-End rule, people remember an experience by its most intense moment and its ending. Today both are negative: the service breaks, then the operator is left wondering whether they did okay. Hover any stage to trace it.

Today After redesign ▲ peak (most-remembered) ● the end (second-most-remembered)
high anxiety ▲ PEAK — it breaks ● a calm end RecruitedLaunch PrepCrash course RehearseGo live It breaksAfter
Peak-End Rule People remember an experience by its peak and its end. Today both are negative — it breaks, then "did I do okay?". Hover a stage to trace it.
The same journey, broken into actions, thoughts, and feelings
StageActionsThoughtsEmotionPain pointOpportunity
LaunchOpens the app, sees the full editor.“Where do I even start?”OverwhelmEight layers of controls greet a beginner.Volunteer View as the default surface.
PrepHunts for this week's content.“Is this the library or the playlist?”ConfusionLibrary, playlist, and presentation blur together.Plain labels and a guided load step.
RehearseHas nowhere safe to practice.“I guess I'll just figure it out live.”UneaseThe first real run is the live run.Practice Mode with no live output.
Go liveAdvances cues as the service runs.“Please don't let me click the wrong thing.”TensionRisky controls sit next to navigation.Live/Next tally; destructive controls hidden.
It breaksWrong slide or black screen appears.“How do I undo this? Everyone can see.”Panic (peak)No obvious path back to a known-good state.One persistent “Back to safe” tap.
AfterService ends, hands back the booth.“Did I do okay?”Doubt (end)The lasting memory is uncertainty.A calm confirmation engineered to land positive.
Define// diagnose with frameworks, then name the opportunity

Three diagnostics that point at the same lever.

A heuristic audit names the defects. The journey curve shows what they cost emotionally. The Fogg model says which lever actually moves the behavior. They converge on one conclusion: raise ability, not motivation.

01 · Heuristic evaluation — Nielsen's 10, with severity ratings (0 to 4)
FILTER

Severity = Nielsen's scale: 0 not a problem · 1 cosmetic · 2 minor · 3 major · 4 catastrophe. The two catastrophes both live in the same moment — going live.

02 · Fogg Behavior Model — drag the operator, find the lever
motivation ↑ ability → (easier) behaviour happens above the line
Below the line

High motivation can't rescue low ability. The confident run doesn't happen.

Motivation 85Ability 25

The insight: the volunteer's motivation is already maxed — they showed up to serve. Ability is the bottleneck. So every concept raises ability (drag the dot right), not motivation. That's why "more training" fails.

The reframe that follows
If the problem is…
“They don't know how”
→ more docs, videos, training
…but it is actually
“They can't, under pressure”
→ raise ability: simplify, rehearse, recover
The job to be done

When I am pulled in last minute to run Sunday's service, I want to get through it without anything visibly going wrong, so I can feel I served rather than embarrassed myself in front of everyone.

FunctionalAdvance the service accurately, live. EmotionalFeel calm and capable, not exposed. SocialBe seen as someone who can be trusted with it again.
Opportunity statements & how-might-we
1

Opening surface. A first-time operator overwhelmed at launch needs a focused starting view, so that the depth of the tool never becomes the first impression. HMW open to a calm, capable surface rather than the full editor?

2

Live safety. An operator advancing cues under pressure needs the safe action to be the obvious one, so that the most reachable control is never the most destructive. HMW make the safe path the obvious path, live?

3

Pre-service confidence. A nervous volunteer needs a real run before the real run, so that their first success happens off-air. HMW build confidence before Sunday, not during it?

Prioritization · impact against effort
impact ↑ effort → (more) do first Back to safe (recovery) Volunteer View default Practice Mode Plain-language labels Guided first-run tour

Recovery wins on both axes and ships first. Volunteer View and Practice Mode carry the most impact and form the core of the redesign. Labels and a guided tour are valuable but secondary.

08b Product reality check

A redesign that ignores the power user is a redesign that ships broken.

ProPresenter is mature software with paying professionals who depend on its depth. Any onboarding change has to earn its place against backward compatibility, support cost, and the people who already know every shortcut. This section is the constraint layer that kept the design buildable rather than aspirational.

Constraints that shaped every decision

Fixed

  • The professional surface cannot be removed or hidden behind a wall; pros pay for that depth.
  • Existing libraries, playlists, and files must open unchanged. No migration step.
  • Output behavior to live displays cannot regress; the screen is sacred.
  • The app runs natively on macOS and Windows and must feel native on both.

Flexible

  • Which surface a brand-new file opens into, and what is emphasized first.
  • Whether a safe, reversible “go live” path can sit on top of the existing one.
  • Labeling, grouping, and the presence of a visible recovery affordance.
  • An opt-in simplified view that is a lens on the same data, not a separate product.
Risk register · what could go wrong and how the design answers it
RiskSeverityMitigation built into the design
Power users feel dumbed downHighVolunteer View is opt-in and reversible in one click. The default for existing files is unchanged. Nothing is taken away; a calmer entry point is added.
A safe “go live” path adds a click for prosMediumThe reversible step is the default only in Volunteer View. Pro View keeps direct-to-live behavior and existing shortcuts intact.
Two views drift apart over releasesMediumVolunteer View renders the same underlying data and components with different density, not a forked codebase. One source of truth.
Recovery button is mistaken for an undo of contentLowIt is scoped and labeled to output state only (“Clear screen” / “Back to safe”), visually separated from slide editing.
Volunteers never discover the safe viewHighFirst launch of a new file routes into it by default; the toggle to Pro is always visible, so discovery is the easy path, not a hidden setting.
Engineering considerations · open questions carried into build
AreaConsideration to resolve with engineering
View as a lens, not a forkConfirm the existing component model can render a reduced-density layout from the same state without a parallel UI tree.
Output state machineDefine a clean “safe / cued / live” state the recovery control can target, distinct from slide selection.
Per-file vs. per-user preferenceDecide whether view choice is remembered per file, per machine, or per operator profile. Affects the shared-booth case.
Cross-platform shortcut layerMap a single intent table to platform-native chords so parity does not mean identical keys. Detailed in the platform section.
Telemetry for validationInstrument time-to-first-live, recovery use, and view switches so the pilot can measure the outcomes claimed below.
Power-user & support assessment

Power-user impact

  • Default workflow for existing files is byte-for-byte unchanged on first run.
  • Every current shortcut and panel remains in Pro View.
  • Net effect on a fluent operator: one ignorable toggle they will never open.

Support-burden assessment

  • Targets the two highest-volume first-run themes: “how do I start” and “I sent the wrong thing live.”
  • A visible recovery path turns a panicked call into a self-served click.
  • Onboarding routed into the product reduces dependence on a single church expert being reachable on Sunday.

These are the design’s own risk and impact analysis. Severity ratings are judgment calls to be pressure-tested with engineering and the pilot, not measured incident rates.

08c Constraints & tradeoffs

The decisions worth defending are the ones I argued myself out of.

Every tempting solution to this problem fails a constraint the product cannot break. Walking through the rejected directions shows the reasoning, not just the result, which is what separates a redesign from a redesign that could ship.

Tradeoff matrix · the four real options
DirectionFirst-timer reliefPower-user costBuild & maintenanceVerdict
Separate Volunteer appHighNone directlyVery high · second product, drift, two codebasesRejected
Force a Volunteer Mode on everyoneHighSevere · breaks pro trustMediumRejected
Full ground-up redesignHighSevere · relearning, migrationVery high · multi-yearRejected
Opt-in Volunteer View as a lensHighNear zero · reversible, ignorableModerate · one source of truthChosen

Why not a separate Volunteer app?

It would relieve first-timers, but it forks the product. Two surfaces drift, files risk incompatibility, and the volunteer never grows into the tool the church actually bought.

Instead: one app, one file, a calmer lens the same operator can graduate out of.

Why not force Volunteer Mode for all?

It would guarantee discovery, but it punishes the paying professional who depends on density and speed. Trust with the core user is the thing the redesign cannot spend.

Instead: default the new file into it, keep it one visible click from Pro, never trap anyone.

Why not redesign all of ProPresenter?

The problem is the first hour, not the whole tool. A ground-up rebuild risks the depth pros pay for and a migration nobody asked for, to solve a problem scoped to onboarding.

Instead: a focused intervention at the moment the evidence points to.

Why not just remove advanced features?

The depth is the value. Reviews resent the climb, never the capability. Removing power solves the symptom by destroying the product.

Instead: defer the depth on first run, never delete it. Progressive disclosure, not subtraction.

06 Design principles

Six rules every later decision answers to.

Principles derived directly from the research, used to keep the design honest. Each solution downstream maps back to one or more of them.

01

Safety before power

The default surface protects the least confident operator. Power is available, never imposed.

↳ Volunteer View, hidden destructive controls
02

Progressive disclosure

Show only what this moment needs. Depth is one deliberate toggle away.

↳ Volunteer / Pro views over one file
03

Confidence through feedback

Make state legible. The operator should always see what is live and what is next.

↳ Live/Next tally, confidence monitor
04

Recovery over prevention

Assume mistakes will happen live. Make the way back instant and obvious.

↳ Persistent Back to safe
05

Cross-platform consistency

One experience on macOS and Windows. Only the window chrome and modifier key change.

↳ Shared layout, platform-correct chrome
06

Accessibility by default

Legible in a dark booth, operable without a mouse, and not English-only.

↳ WCAG 2.2 AA, bilingual path
Develop// diverge, sketch, decide

Three answers to three peaks.

Each concept targets one spike on the anxiety curve and one mechanism from the research. Divergence came first; convergence followed a scored decision, not a preference.

How might we…
…let a first-timer run a service without fear of breaking it? …make the safe path the obvious one, live? …build confidence before Sunday, not during it?
Ideate · diverge before converge (Crazy Eights) SKETCH

Eight fast passes, about forty seconds each. Most exist to be thrown away. Three earned a circle: the view toggle, a practice sandbox, and one-tap recovery.

The three bets · concept exploration
Peak · First launch

Volunteer View by default

Open to a focused surface, not the eight-layer editor. The same file, fewer controls in view.

↳ Cognitive Load Theory, cut extraneous load
Peak · Going live

Practice Mode

A safe dry run mirroring the service with zero live output. A guaranteed early win.

↳ Self-efficacy, a mastery experience
Peak · It breaks

“Back to safe” recovery

One always-visible tap to a known-good state, and the positive ending the memory keeps.

↳ Peak-End, fix the peak and craft the end
Decision matrix · why these three won
ConceptImpact on confidenceEffort to buildReuse across productRiskTotal
Back to safe554519
Volunteer View534315
Practice Mode433414
Guided first-run tour333413
Big physical remote31228

Scored 1 to 5, higher is better, including for effort and risk where 5 means lowest. The top three address all three peaks while staying buildable and reusable across the product family.

Rejected ideas & tradeoffs
A full guided wizard at launch

Tested poorly against the research: a wall of steps repeats the overwhelm it was meant to solve, and experienced operators resent gates. Kept as an optional first-run tour instead.

A separate beginner application

Splitting the product fragments the team and blocks the volunteer's path to growth. Two views over one file preserves a single source of truth.

A dedicated hardware remote

High confidence value but out of scope for a software redesign and a barrier for churches without budget. The on-screen safe button captures most of the benefit.

Confirmation dialogs on every risky action

Prevention by interruption slows the live run and trains operators to click through warnings. Recovery after the fact respects the pace of a service.

07 Information architecture

Same content, a quieter structure.

The redesign does not remove capability. It changes what the first surface presents and defers the rest behind a deliberate toggle. Current state on the left, proposed on the right.

Current state · everything at once
ProPresenter (one surface)
Library, Playlists, Presentations
Show / Edit / Reflow / Bible / More
Audio · Stage · Timers · Messages · Props · Macros
Clear: Slide · Media · Props · All
Media bin, themes, looks
all visible to a first-time operator
Proposed · progressive disclosure
ProPresenter — Volunteer View (default)
This week's service (only)
Live output + Live/Next tally
One “Clear to logo” button
Confidence monitor + Back to safe
Switch to Pro View →
Pro View (one toggle away)
Navigation simplificationVolunteer View collapses the library to this week, hides the media bin and deep show controls, and replaces five clear-layer buttons with one labeled action.
Content hierarchyWhat is live and what is next become the most prominent elements on screen, ahead of editing tools the operator does not need during a service.
08 User flows

Four flows that carry the whole service.

Each flow is designed around the least confident operator, with the rationale tied back to a design principle.

01

First launch

Open appVolunteer ViewSee this week onlyOptional 60s tour
RationaleOpening to a focused surface removes the overwhelm peak before it forms. The tour is offered, never forced.
02

Practice mode

Enter PracticeAdvance cuesNo live outputFinish a full run
RationaleA mastery experience off-air builds self-efficacy, so the first success happens before the room is watching.
03

Running a service

Load this weekTake slide liveWatch Live/NextClear to logo
RationaleState is always legible and the safe action is the most reachable, so advancing the service never sits next to breaking it.
04

Recovery

Wrong slide liveNotice on monitorBack to safeKnown-good output
RationaleOne persistent tap returns output to a safe state, turning the remembered peak from panic into a recovered moment.
Task flow · launch → confident run → recover

The recovery loop, drawn. A mis-click is expected and absorbed rather than prevented.

09 Wireframing

From paper to a working surface.

The chosen direction traveled a fidelity ladder. Each step earned the next by resolving a specific question.

✎ Sketch ▢ Lo-fi ◫ Mid-fi ◆ Hi-fi prototype ↓
Lo-fi wireframe · the chosen direction, on paper LO-FI

Before pixels: a hand sketch of Volunteer View, annotated with the reasoning behind each region.

Sketch → Lo-fi

What changed

The crazy-eight scribbles became a single three-region layout: run order, live output, and a confidence panel. The decision here was spatial, giving the safe button a permanent home rather than burying it in a menu.

Lo-fi → Mid-fi → Hi-fi

What changed

Mid-fidelity set the real grid, type scale, and the Live/Next state language. Hi-fidelity, the working prototype below, added platform-correct chrome and the confidence monitor, and surfaced the contrast failure that drove the accessibility pass.

10 Design system

A system an engineering team could build from.

The redesign is documented as a small, reusable system: foundations, tokens, components, accessibility rules, and cross-platform conventions. Color is borrowed from broadcast convention, where red means on-air and green means cued and safe, so the interface speaks a language operators already trust.

Foundations · color system
Program
#FF4B4F
On-air, live, destructive
Preview
#37D27F
Cued, next, safe, success
Active
#FFA23E
Interactive, selection, focus
Surface
#0C0E13 → #1F2530
Booth-dark backgrounds, panels
Foundations · typography
Display · Bricolage 800Run the service
Heading · Bricolage 700This week's service
Body · Inter 400/600Advance cues in order and return to logo at the end.
Data · IBM Plex MonoLIVE · NEXT · 10:02 · ⌘ →

Three roles, each doing one job: a characteristic display face used sparingly, a neutral body face for reading, and a monospace face for state, timing, and keys, where alignment carries meaning.

Foundations · spacing, grid & elevation
4
8
12
16
24
32
48
GridThe application is a three-column layout: run order, live output, and operator panel. Volunteer View keeps the columns and removes content within them rather than rearranging the room.
ElevationOne resting surface, one raised panel, and a single deep shadow reserved for the application window. Restraint keeps a dark booth interface from turning to mud.
Design tokens · semantic, not raw
TokenValueApplied to
color.live#FF4B4FProgram output, destructive actions
color.safe#37D27FNext cue, recovery, success
color.action#FFA23ESelection, primary control, focus ring
color.surface#11141BApp base
color.panel#181C25Raised panels
text.primary#EEF1F6Headings, live values
text.muted#AFB8C4Secondary copy, labels
radius / space6 to 14px / 4 8 12 16 24 32Corners and rhythm

Tokens are named for meaning. A component asks for color.safe, never #37D27F, so a theme or a high-contrast mode can be retargeted in one place.

Components · reusable across the surface
Buttons
Back to safeTake liveEditClear all
Four intents: safe, primary, ghost, destructive. Intent is carried by token color, not by position.
Status indicators
LIVENEXTIDLE
The single most important signal in the product: what is on-air right now, and what follows.
Alerts
⚠ Wrong slide is live, the room can see this
● Recovery is always one tap away
Alerts state what happened and what to do, in the interface's voice, never an apology.
Panels & empty states
Nothing live yetA calm place to start. Take a slide live when the service begins.
An empty screen is an invitation to act, not a dead end. Modals are reserved for the rare destructive confirmation only.
Accessibility · WCAG 2.2 AA
1.4.3Contrast ≥ 4.5:1. Text legible in a dark booth, the failure caught while building this very page.
2.5.8Target size ≥ 24px. Live controls big enough to hit without looking, under pressure.
2.1.1Keyboard operable. Every cue advanced without a mouse; visible focus throughout.
2.4.7Focus visible. A consistent amber focus ring, tested against every surface color.
4.1.2Screen reader support. Live region announces the current output; controls carry roles and labels.
1.4.4Large text support. Layout holds when type is scaled; nothing clips or overlaps.
2.3.3Reduced motion. Animations and reveals collapse to instant when the system requests it.
3.1.2Language of parts. A first-class bilingual path, so the Spanish-service operator is not an afterthought.
Cross-platform · one experience, two operating systems
macOS conventionsTraffic-light window controls, centered window title, and the ⌘ modifier for the next-cue accelerator.
Windows conventionsApp icon and left-aligned title, minimize, maximize, and close controls, and the Ctrl modifier in their place.

Everything between the window chrome is shared. The same components, tokens, and layout drive both platforms, so a volunteer trained on one is immediately at home on the other and the team maintains a single source of truth. Toggle the OS in the prototype below to see only the chrome and the modifier key change.

14b Accessibility

A volunteer booth is dark, time-pressured, and rarely staffed by one body type.

Accessibility here is not a compliance footnote; it is operating conditions. Low light, stress, color-blind operators, and keyboard-only workflows are the normal case, not the edge. The design is audited against WCAG 2.2 AA, and the broadcast palette was chosen to survive a red-green color deficiency.

WCAG 2.2 AA self-audit · how the design meets each criterion
1.4.3
Contrast (minimum)

Body text and controls clear 4.5:1 on the booth-dark surfaces; large display text and state chips clear 3:1. Verified in the matrix below.

1.4.1
Use of color

Live, cued, and safe states never rely on color alone. Each carries a label and a shape so a color-blind operator reads state correctly.

2.1.1
Keyboard

Advance, recover, go live, and view switch are all reachable and operable without a mouse, mapped to platform-native chords.

2.4.7
Focus visible

A 2px amber focus ring with offset is present on every interactive element, tuned to read in low light without bleeding into live red.

2.5.5
Target size

Primary transport and recovery controls meet a 44px minimum so a tense, fast tap lands, even on a trackpad in the dark.

2.3.3
Motion from interactions

Transitions are short and functional. The scroll reveals and state changes respect prefers-reduced-motion and degrade to instant.

3.2.4
Consistent identification

The same control means the same thing in both views and on both platforms. Recovery is always recovery, never relabeled.

4.1.2
Name, role, value

Transport and state controls expose role and current state to assistive tech, so “live” is announced, not just shown.

Contrast matrix · core pairings on booth-dark
Foreground on backgroundRatioUseResult
Ink on panel #181C2513.6:1Body, headingsAAA
Muted on panel7.2:1Secondary textAAA
Program red on void #0C0E134.9:1Live state textAA
Preview green on void8.1:1Safe / cued stateAAA
Amber on void7.6:1Focus, active controlAAA
Keyboard & recovery map · the operator never needs the mouse
Advance / next
Space
Previous
Go live (cued slide)
Enter
Back to safe
Esc
Clear screen
. period
Switch view
⌘/Ctrl V

Screen reader strategy

  • Output state is a live region: going live announces “Now live: [slide]”, recovery announces “Screen cleared, safe.”
  • Transport controls expose role and state, not just an icon.
  • View switch announces which surface is active so a non-visual operator is never lost.

Cognitive accessibility

  • One primary action visible at a time on first run; depth is deferred, not hidden forever.
  • Plain-language labels over system vocabulary (“Back to safe,” not “Clear output layer”).
  • Recovery is always available, so a mistake never requires recalling a procedure under stress.
14c Cross-platform strategy

Consistency of meaning, not consistency of keys.

ProPresenter runs on macOS and Windows, and the same volunteer may sit at either depending on the church. Parity does not mean identical; it means a Mac operator and a Windows operator both reach the same intent through the convention their hands already expect. The redesign maps one intent table to each platform’s native chords and controls.

Shortcut parity · one intent, two native chords
IntentmacOSWindowsPrinciple
Search library⌘ FCtrl FFollow the universal find convention on each OS.
Enter Practice Mode⌘ ⇧ PCtrl ⇧ PModifier swap only; the letter and meaning stay constant.
Go liveEnterEnterPlatform-neutral keys stay identical to protect muscle memory.
Back to safeEscEscRecovery uses the same key both places, by design.
Switch view⌘ ⇧ VCtrl ⇧ VConsistent chord shape, native modifier.

Proposed mappings that follow each platform’s established conventions, presented as the parity framework, not a claim about current ProPresenter defaults.

Native behavior parity · respecting each platform’s expectations
AreamacOSWindows
Menu barGlobal top menu; app preferences under the app name.In-window menu; settings under Edit or a gear.
Window managementFull-screen spaces; traffic-light controls top-left.Maximize/snap; min-max-close top-right.
Primary modifierCommand for app actions.Control for app actions.
Native controlsSystem toggle and segmented controls, SF-style.Fluent-style toggles and segmented controls.
Output displaySeparate space / display for live output.Secondary monitor for live output.

The interactive prototype demonstrates this directly: the OS toggle reskins chrome, controls, and modifier hints to match the platform while every label, state, and recovery path keeps the same meaning.

Develop// the system, made real and testable

Interactive prototype: ProPresenter's real layout, on macOS & Windows.

A faithful slice of the real three-column app, built to test the riskiest assumptions: that a focused view reduces load, that a visible safe action absorbs a mistake, and that the experience can hold across platforms. Click a slide to take it live, use the clear buttons, fumble a control, then recover. Switch the OS to see the cross-platform chrome, and flip Volunteer to Pro to feel the redesign.

What was prototyped & why

The two core surfaces and the recovery path, because they carry the confidence thesis. Editing tools were left out on purpose; they are not where the first hour fails.

How interactions support the goals

Confidence rises with each landed cue, a mis-click drops it and exposes the safe button, and one tap restores known-good output, making the principle of recovery over prevention tangible.

Interactive prototype · ProPresenter — artifact on display
OS
Calm, focused, recoverable — the default for first-timers. controls in view: 7
ProPresenter — This Week's Service
⚲ SearchTextTheme
ShowEditReflowBibleMore
Volunteer
Library
This Week's ServiceShow
next slide: EasyView
OPERATOR VIEW · enlarged, audience output unchanged
Audio
Stage
Timers
Messages
Props
Macros
On screen now10:02
Logo
Next: Welcome
Operator confidence20%
Pre-service. Nothing live yet — a calm place to start.
● recovery always here
MEDIA BIN
Same app, both platformsProPresenter 7 was rebuilt for an identical experience on Windows and Mac. Toggle the OS: the layout holds — only the window chrome and modifier key () change.
Volunteer vs ProVolunteer collapses the library to this week, hides the media bin and deep show controls, swaps five clear-layer buttons for one, and adds a confidence monitor with one-tap recovery.
Deliver// how we would know it worked

A usability test designed to prove or kill the thesis.

A redesign is not finished when it ships.

Objectives

  • Can a first-time operator complete a basic service unsupervised?
  • Does the safe action get used, and does it resolve a live mistake?
  • Does confidence rise from before to after a run?

Methodology

  • Moderated, task-based usability test, remote or in a simulated booth.
  • Think-aloud during tasks, short interview after.
  • SUS and a 1 to 7 confidence self-report, pre and post.

Participants

  • 5 to 8 per round, recruited as true first-time operators.
  • Mix of proficiency, including a low-confidence profile like Doug.
  • One bilingual operator to exercise the Spanish path.

Tasks

  • Load this week's service and take the first cue live.
  • Run the service in order to the closing slide.
  • Recover after an induced wrong-slide moment.
  • Find the full controls from Volunteer View.
Metrics · baseline → target
Metric (method)
Baseline
Target
Task success, complete a basic service (moderated test)
~55%
≥ 90%
Time to first confident run (custom)
1 supervised service
unsupervised, first try
Error rate, destructive live mistakes
tracked per service
recoverable in 1 tap
SUS, System Usability Scale
below 68 (avg)
≥ 80 (excellent)
Confidence, pre/post self-report (1 to 7)
measure
+2 points
Baselines are hypotheses to validate with a 5 to 8 person benchmark test, not measured results. Targets follow usability norms, where SUS ≥ 80 reads as excellent.
11 Iteration

Three cycles of refinement.

Iterations that ran during design, driven by the heuristic findings and a small round of informal walkthroughs. Impact is expressed as the expected effect on the target metrics, to be confirmed by the full test.

Cycle 1The recovery button kept getting lost
Observation
In the first pass, “Back to safe” lived in a menu. Walkthrough users hunted for it under pressure.
Insight
A recovery affordance only works if it is always visible. A buried safe action is no safer than none.
Design change
Pinned it to a persistent panel with a fixed position and a distinct safe color.
Expected impact
Recovery becomes reliable in one tap, the core of the error-rate target.
Cycle 2Volunteers could not feel their own progress
Observation
Operators landed cues correctly but still reported feeling unsure they were doing it right.
Insight
Confidence needs visible feedback. Correct action without acknowledgment still reads as anxiety.
Design change
Added the confidence monitor and an explicit Live/Next tally that responds to every cue.
Expected impact
Supports the +2 point confidence target and a higher SUS.
Cycle 3The Pro toggle was too easy to miss
Observation
Returning operators worried the simplified view had removed the controls they rely on.
Insight
Simplification must never feel like loss. The path to full power has to be discoverable.
Design change
Made the Volunteer/Pro switch a labeled, persistent control rather than a hidden setting.
Expected impact
Protects experienced operators while keeping the beginner default, the key adoption risk.
16b Validation & rollout plan

How Renewed Vision could actually pilot this, and know whether it worked.

A concept earns trust by being testable. This is the plan to move the design from proposal to evidence in three phases, each with an exit metric that has to be met before the next begins. The metrics are the same ones the design claims to move, instrumented so the claim can be confirmed or killed.

Phase 1
Internal validation
2–3 weeks
  • Moderated usability sessions with recruited first-time operators on the prototype.
  • Run the task list: launch, practice, run a service, recover a mistake.
  • Measure SUS, task success, and time-to-first-live against the baseline.
  • Exit: task success and confidence clear target before any church sees it.
Phase 2
Church pilot program
6–8 weeks · 3–5 churches
  • Real volunteers, real services, Volunteer View on by default for new files.
  • Telemetry on recovery use, view switches, and completed services.
  • Weekly check-ins with media directors on support load.
  • Exit: measured drop in first-run support tickets and no power-user regression.
Phase 3
Production rollout
staged
  • Ship behind a default for new files, with the toggle always visible.
  • Monitor the same metrics at scale; watch power-user sentiment closely.
  • Roll the view-as-a-lens pattern toward the rest of the suite.
  • Exit: sustained metric gains hold past the novelty window.
Success metrics · baseline to target (to be measured in pilot)
MetricHow it’s measuredBaselineTarget
Time to first confident liveTelemetry, file open → first deliberate go-liveunknown / high−50%
Live error frequencyAccidental output events per servicebaseline−60%
First-run support tickets“how do I start / I sent wrong thing” themesbaseline−40%
Volunteer confidence (SUS-style)Post-service self-reportbaseline≥ 80
Service completion without helpServices finished with no expert interventionbaseline≥ 90%

Targets are the design’s hypotheses, the bar it sets for itself, to be confirmed or rejected by the pilot. They are not measured production results.

12 Outcomes

What this would change, for whom.

Projected outcomes, organized by who benefits. They become measured results once the usability test runs.

User outcomes
  • A first-timer completes a service unsupervised.
  • A live mistake is recovered in one tap, not a crisis.
  • Confidence rises measurably from before to after.
Product outcomes
  • A focused default surface with a visible path to full power.
  • One shared experience across macOS and Windows.
  • A reusable system of tokens and components.
Business outcomes
  • Higher volunteer retention, the people churches most need.
  • Lower support burden on staff directors.
  • Fewer visible mistakes, protecting brand reliability.
Fit check · scored against Renewed Vision's stated priorities
Their value
How this redesign earns it
Operators feel confident
The entire thesis, ability raised until the confident run happens
Seamless, professional-grade
Output stays clean even during a beginner's first run
Real rooms, real time
Recovery designed and tested under simulated live pressure
Growth at every level
A visible path from Volunteer View into the full Pro editor
Outcomes above are projected. No production data was used or fabricated; figures are targets to validate, not measured results.
13 Reflection

What worked, and what I would do next.

What worked

Naming the problem correctly did more than any single screen. Reframing the gap as confidence rather than knowledge decided everything downstream, and the Fogg lens carried the argument.

What surprised me

The strongest moves were subtractive. Hiding controls and adding one safe button mattered more than any new feature, which is the opposite of where I expected to spend effort.

What I would improve

The bilingual path is designed but under-explored. A dedicated round with Spanish-service operators would likely reshape the labels and the first-run tour.

What I would test next

Whether hiding controls frustrates volunteers who have started to level up. First task: can a returning operator find the full controls in under ten seconds?

Ecosystem thinking · how this scales across the product family
ProPresenter

The Volunteer/Pro pattern, recovery affordance, and token system are the foundation, proven on the highest-stakes surface first.

ProVideoPlayer

The same on-air/safe color language and one-tap recovery translate directly to playback, where a wrong clip live is the same kind of failure.

ProScoreboard

A focused operator view and visible state map cleanly to a scoreboard run by rotating volunteers under the same live pressure.

14 Why this matters for Renewed Vision

Built for the rooms Renewed Vision ships into.

This project was chosen because it sits exactly where Renewed Vision's products live: volunteer-first environments, high-pressure live production, and software that has to be reliable in front of a room.

Volunteer-first

The redesign is built for the least confident operator in the booth, the person Renewed Vision's own Volunteer Guide is written for.

Live pressure

Every decision assumes a real audience and no expert on hand, which is the actual condition the software runs under.

Cross-platform

One experience on macOS and Windows, matching how the product family is built and maintained.

Accessibility

Legible in a dark booth, keyboard operable, and bilingual by design rather than as an afterthought.

Design systems

Documented tokens and components ready to scale across ProPresenter, ProVideoPlayer, and ProScoreboard.

Confidence under pressure

The work reduces the onboarding burden that quietly limits adoption, retention, and trust in the product.