Deliverable

---
ddx:
  id: example.deliverable.depositmatch-pilot-deck
  type: deliverable
  activity: iterate
  kind: deck
  status: draft
  authoring:
    home: repo
    export:
      - docs/helix/06-iterate/deliverables/assets/DEL-001-depositmatch-pilot.pptx
      - docs/helix/06-iterate/deliverables/assets/DEL-001-depositmatch-pilot.pdf
  links:
    - id: example.business-case.depositmatch
      kind: informed_by
    - id: example.product-vision.depositmatch
      kind: informed_by
---

# A three-month pilot wins back reviewer hours

## Brief

- **Audience**: the firm-owner sponsor and the two partners who approve product spend
- **Occasion**: monthly investment review, 20-minute slot, deck presented live then sent as PDF
- **Decision or action sought**: approve the three-month pilot budget and name who recruits pilot firms
- **Time slot or page budget**: 20 minutes, at most 10 content slides
- **Kind**: deck
- **Constraints**: no client names; figures marked as assumptions stay marked; house theme
- **Look**: editorial
- **Scope**: docs/helix/00-discover/business-case.md; docs/helix/00-discover/product-vision.md
- **Breadth**: deep-dive
- **Angle**: the investment decision
- **Must cover**: pilot scope; investment and breakeven; risks
- **Must omit**: none
- **Max messages**: 4
- **Takeaway**: a bounded CSV-first pilot is the cheapest way to learn whether reviewers trust suggested matches and whether firms will pay for the time saved

## Story

- **Flow**: proposal
- **Takeaway**: a bounded CSV-first pilot is the cheapest way to learn whether reviewers trust suggested matches and whether firms will pay
- **Messages** (ranked by how much each moves the decision to fund the pilot):
  1. Firms lose reviewer capacity to manual deposit matching every week, and the market is large enough to matter (evidence: docs/helix/00-discover/business-case.md#executive-summary; docs/helix/00-discover/business-case.md#opportunity-sizing)
  2. A CSV-first pilot learns about reviewer trust faster and cheaper than integrating bank feeds (evidence: docs/helix/00-discover/business-case.md#alternatives-considered)
  3. Year one costs $262,000 and the plan breaks even in month 18 if pilot conversion and pricing hold (evidence: docs/helix/00-discover/business-case.md#investment-required; docs/helix/00-discover/business-case.md#expected-roi)
  4. Three risks decide the pilot, each with a mitigation, and two conditions gate spend beyond it (evidence: docs/helix/00-discover/business-case.md#risk-assessment; docs/helix/00-discover/business-case.md#recommendation)
- **Beats and titles** (beat ids from the proposal flow in deck-flows.yml; titles written before any body):

  | # | Beat | Title (claim) | Message | Exhibit |
  |---|---|---|---|---|
  | 1 | recommendation | A three-month pilot wins back reviewer hours | takeaway | title slide |
  | 2 | situation | Firms spend $1.2B a year on reviewers matching deposits | M1 | two stat callouts |
  | 3 | complication | Firms lose reviewer capacity to manual deposit matching every week | M1 | three panels where evidence goes missing |
  | 4 | options-considered | A CSV-first pilot tests reviewer trust faster than bank feeds | M2 | two-column comparison with the do-nothing row |
  | 5 | recommended-option | Approve a pilot that shows reviewers the evidence | M2 | three-step process strip |
  | 6 | return | The pilot breaks even in month 18 on planned pricing | M3 | timeline to breakeven |
  | 7 | cost | The pilot costs $262,000 in year one | M3 | cost table with the total highlighted |
  | 8 | risks | Three risks decide the pilot, each with a mitigation | M4 | risk grid |
  | 9 | plan | Two conditions gate any spend beyond the pilot | M4 | checkpoint strip |
  | 10 | ask | Approve the pilot budget and name a recruiter | takeaway | ask band and steps |
  | 11 | sources | Sources | all | source table |

- **Concept coverage** (every concept group the inventory surfaced, in authority order):

  | # | Concept group | Authority | Status | Carried by / reason |
  |---|---|---|---|---|
  | 1 | Pilot scope: import, evidence-backed review, exception ownership | business case, vision | covered | M2 |
  | 2 | Market and opportunity sizing | business case | covered | M1 |
  | 3 | Investment and breakeven | business case | covered | M3 |
  | 4 | Risks and gating conditions | business case | covered | M4 |
  | 5 | Alternatives considered | business case | covered | M2 |
  | 6 | Strategic alignment with the firm's goals | business case | omitted | angle: the committee decides on cost and risk, and the goals row restates the pilot scope |
  | 7 | Product vision and positioning | vision | omitted | audience: the sponsor authored the vision; slide 3 carries the pain it names |

- **Horizontal-logic test**: pass. The titles move from the answer (1) through what the audience already accepts (2), what it costs them (3), the choices (4), what the pilot is (5), what it returns and costs (6, 7), what could go wrong (8, 9), and back to the ask (10); each shares a word with the one before, and the ask repeats "pilot" from the takeaway.

## Content

### 1. A three-month pilot wins back reviewer hours

**Pattern**: title
**Body**:
- Investment review, prepared for the sponsor and approving partners
**Visual**: kind: none | icon: search. Full-bleed dark surface with the deck title lower-left, the hero marker on the right carrying the deck's one icon, and the review date beneath
**Notes**: Open with the takeaway sentence from the brief. Say up front that the ask comes on the last slide and takes two decisions.
**Sources**: S1

### 2. Firms spend $1.2B a year on reviewers matching deposits

**Pattern**: stat-callout
**Body**:
- $1.2B spent yearly on reconciliation labor and tools across small firms
- $180M of that sits in firms of the size we can serve
- Both figures are planning assumptions until the research plan validates them
**Visual**: kind: stat | stats: $1.2B / spent yearly on reconciliation labor and tools across 60,000 small firms; $180M / within reach at the 9,000 firms of the size we can serve | label: planning assumption. Two stat callouts from the opportunity sizing, $1.2B in the primary hue and $180M in the secondary hue, the firm counts as baseline captions
**Notes**: The total assumes 60,000 small bookkeeping firms spending about $20,000 a year each. The reachable slice is the 9,000 firms with five to twenty-five staff. Say plainly that confidence is low and that the pilot is partly how we raise it.
**Sources**: S3, S4

### 3. Firms lose reviewer capacity to manual deposit matching every week

**Pattern**: claim-evidence
**Body**:
- Deposit matching is a weekly chore that competes with client work
- A reviewer reconciles each deposit by hand against three separate sources
- The reasoning behind a match lives in one person's head once the sheet closes
- Saved hours count for an owner only when the result can be checked afterwards
**Visual**: kind: panels | items: Exports / matching happens across exports, bank statements, and email threads; Spreadsheet / closing it throws away the evidence behind each match; Audit / owners pay for saved time only when they can audit the result. Three panels standing in for the annotated matching-spreadsheet screenshot, each marking a place evidence goes missing
**Notes**: Keep this to the lived experience; the sponsor knows the pain. The point is that time saved without an audit trail is not something a firm will pay for.
**Sources**: S5

### 4. A CSV-first pilot tests reviewer trust faster than bank feeds

**Pattern**: two-column-comparison
**Body**:
- CSV-first: learns about reviewer trust in weeks; needs pilot recruiting and careful data handling
- Bank feeds first: stronger automation story; longer build, higher integration risk, slower learning
- Doing nothing for a quarter: rejected; every reviewer hour stays lost to manual matching
- Verdict: CSV-first, because the open question is reviewer trust
**Visual**: kind: two-column | left: CSV-first pilot | right: Bank feeds first | rows: Learns about reviewer trust in weeks / Stronger automation story; Needs pilot recruiting and careful data handling / Longer build and higher integration risk; Against doing nothing, saves the quarter of manual matching / Against doing nothing, saves nothing until the build lands | prefer: left | verdict: CSV-first, because the open question is reviewer trust. Two columns from the alternatives table in the business case with the check mark on the preferred column and the do-nothing baseline as the last row
**Notes**: The alternatives table in the business case also rejects spreadsheet templates, which keep no evidence; mention them only if asked. The comparison is about what we learn fastest, and doing nothing learns nothing.
**Sources**: S5, S6

### 5. Approve a pilot that shows reviewers the evidence

**Pattern**: claim-evidence
**Body**:
- The pilot builds three things and nothing else
- Reviewers see the evidence behind every suggested match before approving it
- Every exception has a named owner instead of a shared spreadsheet
- Spend beyond the pilot waits for two measured conditions
**Visual**: kind: process-flow | steps: Import; Review; Own exceptions | highlight: 2 | caption: reviewers see the evidence first. Three-step process strip drawn from the pilot scope in the business case, with the review step highlighted
**Notes**: This is what the committee is buying. The next two slides say what it returns and what it costs; the two after that say what could go wrong and when we would stop.
**Sources**: S1, S2

### 6. The pilot breaks even in month 18 on planned pricing

**Pattern**: claim-evidence
**Body**:
- Year one revenue of $120,000 leaves a $142,000 shortfall we fund deliberately
- Breakeven arrives in month 18 if pilot conversion and pricing hold
- The three-year return rests on pricing the pilot has not yet validated
**Visual**: kind: timeline | milestones: Pilot starts; Year one closes with $120,000 revenue; Breakeven in month 18; Three-year return | now: 1 | marker: today | caption: the shortfall is funded until month 18 | side: left. Net line over three years drawn from the expected ROI section, with the breakeven month marked in the secondary hue
**Notes**: The model is in the expected ROI section: revenue ramps with pilot conversion, and the three-year return depends on the price point the pilot tests. Lead with the breakeven month; the return figure waits for a partner to ask.
**Sources**: S8

### 7. The pilot costs $262,000 in year one

**Pattern**: table
**Body**:
- Year one: $180,000 development, $12,000 infrastructure, $40,000 go-to-market, $30,000 operations
- Development takes $180,000 of the $262,000 total
**Visual**: kind: table | columns: Year one line / Amount | rows: Development / $180,000; Infrastructure / $12,000; Go-to-market / $40,000; Operations / $30,000; Total / $262,000 | highlight: 5. Five-row cost table for year one from the investment section with the total row highlighted
**Notes**: Years two and three are in the appendix source; do not present them unless a partner asks. Lead with the year-one number the committee approves.
**Sources**: S7

### 8. Three risks decide the pilot, each with a mitigation

**Pattern**: risk-matrix
**Body**:
- Exports vary too much: recruit pilots across three accounting systems, map per client
- Reviewers distrust suggestions: show evidence before approval, require reviewer acceptance
- Firms will not pay enough: validate willingness to pay before widening scope
**Visual**: kind: risk-grid | risks: Exports vary too much / high / medium; Reviewers distrust suggestions / medium / high; Firms will not pay enough / medium / high. Likelihood-by-impact grid from the risk assessment with the three risks as numbered dots and the mitigations from the body keyed by number on the right
**Notes**: The export-variability risk is the most likely and the one the recruiter can reduce before build starts, which is why the ask names a recruiter. The business case names no risk owners; the ask has the committee name one.
**Sources**: S9

### 9. Two conditions gate any spend beyond the pilot

**Pattern**: process-flow
**Body**:
- Recruit at least five pilot firms before expanding beyond import and review
- Both measures are read at the two-month checkpoint, before any expansion decision
- A paid product waits until weekly reconciliation falls under 3 minutes per client
**Visual**: kind: process-flow | steps: Recruit five firms; Read two measures; Expand or stop | highlight: 2 | caption: the two-month checkpoint is where the committee can stop the pilot. Three-step checkpoint strip from the recommendation's conditions, with the 3-minute paid-product target in the body
**Notes**: These are the conditions in the business case recommendation. They are the sponsor's protection against scope creep and the team's protection against building on unvalidated demand.
**Sources**: S2, S10

### 10. Approve the pilot budget and name a recruiter

**Pattern**: ask-next-steps
**Body**:
- Approve the year-one pilot budget of $262,000 today
- Name the partner who recruits five pilot firms, by month end
- Product lead reports median reconciliation time at the two-month checkpoint
**Visual**: kind: none | ask: Fund the pilot and name the recruiter today | owners: investment committee; a named partner; product lead | dates: today; month end; two-month checkpoint | consequence: No decision today delays the pilot a full quarter and keeps firms in manual matching. Ask band in the secondary hue carrying the decision, then three numbered steps with owner and date columns filled from the body
**Notes**: Stop talking after the ask. If the committee wants the three-year numbers, they are in the appendix source and the business case.
**Sources**: S7, S10

### 11. Sources

**Pattern**: appendix-sources
**Body**:
- Every figure and claim above maps to a section of the business case or product vision listed below
**Visual**: kind: none. Two-column source list in caption size, claim on the left, artifact section on the right
**Notes**: Not presented.
**Sources**: S1

## Sources

| Id | Claim or figure | Governing artifact and section |
|---|---|---|
| S1 | Recommended investment is a three-month pilot build for CSV import, evidence-backed match review, and exception ownership | docs/helix/00-discover/business-case.md#executive-summary |
| S2 | Decision: Conditional Go; conditions: recruit at least five pilot firms; measure median reconciliation time and suggestion acceptance accuracy during the first two months | docs/helix/00-discover/business-case.md#recommendation |
| S3 | TAM $1.2B annual workflow spend: 60,000 small bookkeeping firms x $20,000; planning assumption, low confidence | docs/helix/00-discover/business-case.md#opportunity-sizing |
| S4 | SAM $180M: 9,000 firms with five to twenty-five employees x $20,000; medium confidence | docs/helix/00-discover/business-case.md#opportunity-sizing |
| S5 | Firms are losing reviewer capacity to manual deposit matching; firm owners will pay for auditability plus time saved | docs/helix/00-discover/business-case.md#executive-summary, docs/helix/00-discover/product-vision.md#positioning |
| S6 | Alternatives: CSV-first pilot (carry forward), bank feed and accounting sync first (reject for v1), spreadsheet templates (reject), do nothing for one quarter (reject) | docs/helix/00-discover/business-case.md#alternatives-considered |
| S7 | Year one: development $180,000, infrastructure $12,000, go-to-market $40,000, operations $30,000; total costs $262,000 | docs/helix/00-discover/business-case.md#investment-required, docs/helix/00-discover/business-case.md#expected-roi |
| S8 | Year one revenue $120,000, net $142,000 shortfall (shown as -$142,000); breakeven month 18; confidence low until the pilot validates conversion and pricing | docs/helix/00-discover/business-case.md#expected-roi |
| S9 | Risks: CSV exports vary (high, medium); reviewers distrust suggested matches (medium, high); firms will not pay enough (medium, high), each with the mitigation quoted, each graded for likelihood and impact | docs/helix/00-discover/business-case.md#risk-assessment |
| S10 | Paid product path if weekly reconciliation time falls below 3 minutes per client; recommended next steps with owners and dates: the investment committee approves the year-one budget today, a named partner recruits five pilot firms by month end, the product lead reports reconciliation time at the two-month checkpoint | docs/helix/00-discover/business-case.md#executive-summary, docs/helix/00-discover/business-case.md#recommendation |

## Assumptions and gaps

- **Assumption**: the committee approves year-one spend only; years two and three are informational. Inferred from the occasion (monthly investment review). The sponsor should confirm.
- **Gap**: the business case names no pilot recruiter; the ask has the committee name one instead of inventing an owner.
- **Gap**: the business case names no owner per risk, which the proposal flow's checklist asks for; the risk slide carries mitigations only, and the ask names the recruiter who reduces the export risk.
- **Gap**: no artifact holds the review date or presenter; both come from the meeting invite at render time.

## Render

- **Theme**: deliverables/theme.yml (no project design-system declared), editorial look
- **Targets**: docs/helix/06-iterate/deliverables/assets/DEL-001-depositmatch-pilot.pptx, docs/helix/06-iterate/deliverables/assets/DEL-001-depositmatch-pilot.pdf
- **Gate**: script checks pass (check-deliverable.py, 0 blocking); titles and shapes pass (title.slop, horizontal_logic, shape.slop, and restatement report nothing); horizontal logic pass; vertical logic pass; fidelity pass (every stat callout re-read against the business case); voice pass (prose lint with the Helix styles, 0 errors); visual: every slide rendered to an image and inspected on 2026-09-10, one overflow on slide 7 fixed by shortening the second bullet; file: presentation validator pass; flow checklist: pass except risk owners, which no source names (recorded under Assumptions and gaps); coverage: 7 groups, 5 covered, 2 omitted with reasons, every must-cover concept carried, no must-omit concepts