Pitch Deck
A sequence of presentation slides that tells a persuasive story to win a decision, one idea per slide.
Pitch Deck
Section titled “Pitch Deck”A pitch deck is a sequence of presentation slides that tells a persuasive story to win a decision - investment, a deal, a go-ahead - one idea per slide, built to be presented aloud or skimmed quickly. The format is defined by its presentation context: the speaker carries the thread, each slide carries one idea, and the sequence builds conviction before asking for anything. A slide that stands alone without a speaker has likely been written as a document, not a deck.
The canonical pitch deck moves through a deliberate narrative arc: establish the problem and who it hurts, reveal the solution and why this approach works, show the market and the moment, demonstrate traction or evidence, introduce the team’s capacity to execute, and close with a clear ask. Not every deck uses this exact sequence, but every effective deck follows the same logic - earn the ask before making it. Slides that skip the problem or bury the ask undermine the format’s persuasive structure.
Canonical template
Section titled “Canonical template”## Slide 1: Problem[Who is the customer? What is the pain? How widespread and severe is it?]
## Slide 2: Solution[What do you do? Why is this the right answer to the problem on Slide 1?]
## Slide 3: Why Now / Market[What has changed? How large is the opportunity? Why is this the moment?]
## Slide 4: Product[How does it work? One or two key screens or mechanics, not a full feature tour.]
## Slide 5: Traction / Evidence[What have you proven? Revenue, users, retention, partnerships, or pilot results.]
## Slide 6: Business Model[How does value capture work? Who pays, how much, and with what unit economics?]
## Slide 7: Team[Who is building this and why are you the right people to win?]
## Slide 8: Ask[What are you requesting? Amount or approval, use of resources, and milestone by when.]When to use
Section titled “When to use”Pitching to investors, partners, or executives to win funding, a deal, or a strategic go-ahead; presenting at a demo day, accelerator, or board meeting where narrative speed and visual clarity matter; aligning a stakeholder group on a direction when the story arc matters as much as the details; sending a deck ahead of a live meeting so the audience can skim the narrative before the conversation.
When not to use
Section titled “When not to use”When the decision-maker needs to reason through the full argument independently without a presenter; when the content is technically dense enough that slide-level text cannot carry the necessary precision; when the audience expects a written deliverable they can annotate, share, or return to as a reference.
Pairs well with
Section titled “Pairs well with”product-thinker, executive, confident, candid, problem-solution
Often confused with
Section titled “Often confused with”one-pager: A one-pager is a single-page prose document that makes one argument or presents one situation, compressed to fit on a single page and designed to be read independently by a decision-maker without a presenter. A pitch deck unfolds across many slides in a deliberate narrative sequence, relies on minimal text per slide and visual momentum, and is built for a live presentation moment where the speaker carries the story thread. The one-pager can circulate and be absorbed on its own; the pitch deck typically needs a presenter to deliver its persuasive force.
- A sequence of discrete slides, each carrying exactly one idea or claim
- Slide titles that advance a narrative arc - problem, solution, market, traction, team, ask
- Minimal text per slide: a headline, a few bullets, or a single stat - not paragraphs
- Visuals and data that carry the persuasive argument rather than decorate
- A deliberate sequence that earns the ask before placing it at the end
- An explicit Ask slide naming what is requested, the amount or approval, and the milestone
- Designed for live presentation or rapid skim, not solo deep reading
Anti-patterns
Section titled “Anti-patterns”- Loading each slide with paragraph-length prose so the deck can be read without a presenter - A pitch deck is built for a presentation moment where the speaker carries the thread; text-heavy slides break the spoken rhythm and produce a poor substitute for a proper written document.
- Treating it as a one-pager spread across many slides - writing prose sections designed for an asynchronous solo read and distributing them across slide boundaries - A one-pager is a single-page prose document designed to make one argument a reader can absorb independently; a pitch deck builds persuasion through narrative sequence and visual momentum with a live presenter. Conflating the two produces a deck that neither presents well nor reads well on its own.
- Leading with the solution or product features before establishing the problem and who has it - The persuasive arc of a pitch deck depends on making the audience feel the problem before the solution is introduced; a solution without a felt problem has no story and no urgency.
- Placing the ask before traction and team credibility have been established - The ask earns its authority from everything that precedes it in the narrative; an early or unsupported ask reads as a negotiation move rather than a conclusion the evidence supports.
Failure modes
Section titled “Failure modes”- Over-persuades into brochure mode - the deck becomes pure selling, omitting risks and weaknesses, until no sophisticated audience trusts it - Surface at least one genuine risk or constraint; acknowledging hard truths reads as credibility, not vulnerability, because it signals the presenter knows the landscape.
- Narrative becomes theater - so much visual staging and story drama that the deck feels produced but empty; the audience senses it is being sold at rather than shown evidence - Anchor each slide in a concrete claim or number; story beats without substance dissolve the moment the audience asks follow-up questions.
Instruction
Section titled “Instruction”Write as a pitch deck. Organize the content as a sequence of discrete slides, each carryingone idea. Follow a persuasive narrative arc: open by establishing the problem and whoexperiences it, then introduce the solution and why it works, show the market and the moment,demonstrate traction or evidence, introduce the team, and close with a clear and specific ask.Keep each slide to a headline plus a few bullets or a single supporting visual - notparagraphs. Write slide titles that advance the story, not just label the category. The askmust be earned: do not place it before the problem, evidence, and team have been established.If a slide can stand alone without a presenter, it likely has too much text.Template
Section titled “Template”See the Pitch Deck template.
Related
Section titled “Related”Pairs well with
Section titled “Pairs well with”Product Thinker, Executive, Confident, Candid, Problem-Solution
Avoid with
Section titled “Avoid with”Confessional, Reverent, Pastoral
Often confused with
Section titled “Often confused with”Examples
Section titled “Examples”- Whether the team should move to async-first standups
- Designing a sustainable morning routine
- Choosing Postgres vs DynamoDB for a new service
- Telling stakeholders a committed feature is being cut this quarter
- Getting a new engineer productive in their first two weeks
- Writing to thank a mentor who shaped your career
- Reflecting on keeping a discipline of rest
- Marking a long-serving colleague's departure
- Marking the team shipping a hard, long project
- Arguing a public position on return-to-office
- Announcing a new product to an outside audience
- A personal year-end reckoning with a difficult year
Slide 1: The standup is failing half the team
Section titled “Slide 1: The standup is failing half the team”- Team: 11 engineers across 4 timezones - US Pacific, US Eastern, UK, India
- 9am Pacific is 9:30pm IST
- India engineers averaged 3.2 of 5 standups in Q1; US engineers averaged 4.6
- This is not a discipline problem. It is a scheduling problem.
Slide 2: Swap the meeting for a three-field Slack post
Section titled “Slide 2: Swap the meeting for a three-field Slack post”- Engineers post to #team-standup by 10am local time
- Three fields: Shipped, In progress, Blocked or at risk
- Blockers @mention the person who can resolve them - no waiting for 9am Pacific
- Status persists in Slack; searchable; no information lost between sessions
Slide 3: The team crossed the timezone threshold
Section titled “Slide 3: The team crossed the timezone threshold”- At 6 engineers: a 9am Pacific slot fits the whole team
- At 11 engineers across 4 timezones: it does not
- India headcount grew across the last two hiring rounds
- We are already operating async in practice - only the standup pretends otherwise
Slide 4: One post per engineer, by 10am local time
Section titled “Slide 4: One post per engineer, by 10am local time”- Post: Shipped / In progress / Blocked or at risk in #team-standup
- On-call engineer: reads the channel by 9am Pacific, responds to @mentions within 30 min
- Thursday: 60-minute working session at 8am Pacific replaces the weekly sync slots - reserved for discussion that requires real-time exchange
- No new tools. Slack templates only.
Slide 5: 14-minute meeting, 4 minutes of signal
Section titled “Slide 5: 14-minute meeting, 4 minutes of signal”- The standup averaged 14 minutes per session
- Past month analysis: 4.2 minutes per session changed someone’s behavior
- The remaining 10 minutes was status that required no response from anyone
- Last quarter: 3 incidents where an engineer spent 1+ hour on a problem already solved in a prior standup - no persistent record existed
Slide 6: Recover 70 minutes per engineer per week, add nothing new
Section titled “Slide 6: Recover 70 minutes per engineer per week, add nothing new”- Recovered: 70 person-minutes per engineer per week (the 14-minute daily slot disappears)
- Thursday session: 60 minutes replaces the five daily sync slots for the week
- New overhead: On-call engineer adds a daily channel-read to existing scope
- Tooling cost: Zero - Slack templates, no new subscriptions
Slide 7: The infrastructure is already in place
Section titled “Slide 7: The infrastructure is already in place”- Engineering manager: owns the template, the process doc, and the 30-day retro
- On-call rotation (already running): adds daily channel triage to existing scope
- No new roles. No new tools. No new budget.
Slide 8: Thirty days, then a decision
Section titled “Slide 8: Thirty days, then a decision”Ask: Approval to run a 30-day trial starting next Monday
At Day 30, the director of engineering reviews:
- Participation rate by timezone
- Blocker resolution time vs sync baseline
- Qualitative signal from 1:1 check-ins with all 11 engineers
Decision: Adopt, modify, or revert - with evidence.
Slide 1: Problem
Section titled “Slide 1: Problem”The first hour of every weekday belongs to someone else’s inbox
- Wake at 6:30, phone in hand before feet touch the floor
- First 45 minutes: Slack, email, news - reactive before the day begins
- 7:30am: family starts (breakfast, school dropoff, lunches). 9am: work starts
- Arrive at the desk already behind, already drained, already a step behind my own intentions
Three attempts in the past year to “just stop using my phone in the morning.” Each lasted four to six days. The pattern: removed a behavior, replaced it with nothing. The vacuum collapsed back.
Slide 2: Solution
Section titled “Slide 2: Solution”A four-step protocol that owns the 6:30-7:30 window before the day can claim it
- Water - 500ml within 5 minutes of waking (glass on the nightstand the night before)
- Light - 10 minutes outside or by an open window
- Movement - 15 minutes: walk, stretch, or bodyweight
- Planning - 10 minutes on paper, top three priorities for the day
Phone stays in the kitchen, plugged in, face down, until step 4 is done.
One structural rule: replace the behavior, do not just remove it. That is what the three previous failures missed.
Slide 3: Why Now
Section titled “Slide 3: Why Now”The 30-day test is complete. This is no longer a hypothesis.
- The v3.0 protocol launched 2026-05-14 and ran through day 30
- Data is in hand - not projected, not planned, not aspirational
- ADR-001 (2026-05-14) committed to a review at day 30; today is that review
- The question is no longer “will this work?” The question is “do we extend, and on what terms?”
The decision point is now, not next month.
Slide 4: How It Works
Section titled “Slide 4: How It Works”One decision removed: the phone is not within reach, so there is nothing to resist
The phone-in-the-kitchen rule is the structural load-bearing piece. Willpower is not the mechanism. Distance is.
Each morning logs one row in log/days.csv: date, wake time, steps completed, mood, notes. Retros happen weekly in log/retros/. The protocol is designed to be skipped and resumed, not abandoned on the first missed morning.
Travel breaks the protocol. That is a known gap. A travel variant is Month 2 work.
Slide 5: Traction
Section titled “Slide 5: Traction”23 of 30 mornings completed. The phone deferral held at 93 percent.
| Metric | Month 1 Result |
|---|---|
| Full protocol completion | 23 of 30 (76.7%) |
| 6:15 wake held | 19 of 30 (63.3%) |
| Phone deferred until step 4 | 28 of 30 (93.3%) |
| Most common mood | ”Steady” (14 days) vs. pre-protocol “tired” and “rushed” |
Misses are explained: 3 travel days, 4 sick days, 4 Tuesdays. No silent-failure cliff - every miss has a traceable cause.
Slide 6: What It Costs
Section titled “Slide 6: What It Costs”Sleep time, one unasked question, and a travel plan that does not exist yet
- Bedtime shifted from 11:30pm to 10:30pm - partially achieved, not fully locked in
- Spouse’s 6:45-7:15am coffee window runs inside the silent hour; conversation deferred to 7:15 by design
- The silent-hour arrangement has been tolerated, not explicitly agreed to - that conversation has not happened
- Travel: two of three trips attempted the protocol, both poorly; a travel variant must be written before the next trip
- Weekends: currently running the same protocol, verdict genuinely unclear
The accountability partner check-in threshold is 70 percent completion. Month 1 cleared it at 76.7.
Slide 7: Team
Section titled “Slide 7: Team”Three people the protocol depends on
- Me - protocol owner; designs, logs, adjusts, runs the weekly retros
- Spouse - absorbs the early-morning silence and the earlier bedtime; has not raised friction, but has not been directly asked
- Accountability partner - committed to a check-in if monthly completion drops below 70 percent; condition was written into ADR-001 and has not triggered
The accountability partner condition is the external forcing function. It has held without needing to activate.
Slide 8: Ask
Section titled “Slide 8: Ask”Two commitments for Month 2
-
Explicit family agreement by end of this week - a 10-minute conversation with my spouse to confirm the silent hour and earlier bedtime are genuinely sustainable, not just tolerated. Do not assume. Ask.
-
30-day extension with two additions - continue through day 60 with a travel variant documented before the next trip and a written weekend rule (even if the rule is “same as weekdays”).
Review at day 60. If completion drops below 70 percent before then, the accountability partner check-in triggers before any design change.
Slide 1: The Decision We Cannot Delay
Section titled “Slide 1: The Decision We Cannot Delay”Lattice Notify launches in six weeks. Choosing the wrong datastore now means 3-6 weeks of rework later - at the worst possible moment.
- 500K notification events per day at launch
- Two viable options: Postgres (known) and DynamoDB (better access-pattern fit, zero team experience)
- Decision gate: Friday 11am sync with Priya
- Sprint planning Friday 2pm requires a confirmed datastore to proceed
Slide 2: Recommendation - Postgres
Section titled “Slide 2: Recommendation - Postgres”Build the notification service on the existing Postgres cluster with a new notifications schema and a pg_notify-backed job queue.
- One operational surface for the 4-person on-call rotation
- Cross-database joins (users, accounts, workspaces) stay simple SQL
- Team has operated Postgres at this scale; DynamoDB production experience is zero
- Revisit threshold documented at 5M events/day sustained - migration path exists if growth demands it
Slide 3: Why This Decision Must Be Made Now
Section titled “Slide 3: Why This Decision Must Be Made Now”The spike is done. The architecture meeting reached consensus. Every day without a locked decision is a day the sprint team cannot commit work.
- DynamoDB spike (
experiments/notify-ddb/) complete: access-pattern fit confirmed, operational cost quantified - Wednesday architecture meeting (Ana, Marcus, Priya, Jordan, Sam) landed on Postgres
- The Slack-partnership deal that triggers the 10x growth scenario has not closed; designing for it now is premature
- Six weeks to first production traffic; the decision window closes Friday
Slide 4: How the Postgres Path Works
Section titled “Slide 4: How the Postgres Path Works”A single cluster carries the full notification lifecycle at launch scale.
- Write path: events land in
notificationsschema, picked up bynotification_jobsworker pool within ~50ms - Read path: read replicas absorb fanout reads; sub-second p95 delivery to end users
- Queue:
pg_notify+notification_jobstable with dead-letter, retry, and queue-depth on familiar tooling - Threshold: at 5M events/day sustained the on-call rotation owns the dashboard metric that triggers revisit
Slide 5: What the Evidence Shows
Section titled “Slide 5: What the Evidence Shows”We did not guess. We ran the spike and put both options through the architecture review.
- DynamoDB spike: access-pattern fit is real; operational overhead is also real (second runbook, second monitoring surface, second debugging skillset on call)
- Postgres at 500K events/day: no new tooling, no new runbooks, on-call rotation load unchanged
- Architecture meeting outcome: after operational cost was quantified, the room aligned on Postgres
- Marcus (DynamoDB advocate) is aligned on the recommendation and the revisit-threshold language
Slide 6: The Tradeoff We Are Accepting
Section titled “Slide 6: The Tradeoff We Are Accepting”We are buying operational stability now and accepting access-pattern impurity. This is the right trade at this team size.
- What we gain: ~3 weeks faster to first production traffic; single on-call surface; no new debugging skillset required
- What we give up: DynamoDB’s cleaner fit for write-heavy, point-lookup-by-user access patterns
- Rework cost if Postgres is outgrown: 3-6 weeks to migrate, triggered by a clear data signal (5M events/day)
- Rework cost if DynamoDB and cross-database joins are both needed: equivalent rework plus a team that learned the wrong tool for this situation
Slide 7: The Team That Will Execute
Section titled “Slide 7: The Team That Will Execute”We have experienced operators and a clear ownership structure for every milestone.
- 8 backend engineers with production Postgres experience
- 4-person on-call rotation covering Mon/Wed/Fri/weekend
- Ana Rivera (tech lead): owns schema design, Postgres partitioning roadmap, ADR-0023 authorship
- Sam:
notificationsschema andnotification_jobstable spec delivered by May 20 - Jordan: queue-depth and write-rate added to on-call dashboard by May 22
- First end-to-end internal traffic on the Postgres path targeted for May 29
Slide 8: What We Are Asking For
Section titled “Slide 8: What We Are Asking For”Two approvals this week unblock six weeks of execution.
- Priya - Friday 11am: confirm ADR-0023 Accepted and lock the datastore decision
- Engineering leadership - by Monday: review ADR-0023 so it can be published to the wider eng team and close the decision loop
- Sprint planning - Friday 2pm: first two weeks of build work committed, assuming decision is locked
- Success signal: first end-to-end internal traffic on the Postgres path by May 29 - on the six-week ship schedule
Slide 1: Title
Section titled “Slide 1: Title”Insights Dashboard: Q3 Decision and Path Forward
Product Update | September 2026 | Maya Chen, Product Lead
Slide 2: The Commitment We Made
Section titled “Slide 2: The Commitment We Made”We promised Insights to four enterprise accounts by end of Q3.
- Sales positioned Insights as a closing point on several deals
- Customers were given a firm Q3 delivery date
- Their expectation: saved views and scheduled reports, fully working
Slide 3: What Changed
Section titled “Slide 3: What Changed”A mandatory billing-system migration grew past its estimate and consumed all available engineering capacity.
- The migration is non-optional - regulatory and contract requirements require it before year-end
- The two workstreams cannot share capacity without risking both
- Insights and the migration cannot finish in Q3 at the same time
Slide 4: Shipping Now Would Still Break the Promise
Section titled “Slide 4: Shipping Now Would Still Break the Promise”The current Insights build is missing the two capabilities customers specifically asked for.
- Missing: saved-view persistence layer
- Missing: scheduled-report delivery
- Releasing without these delivers exactly the wrong experience at the highest-stakes moment
Slide 5: The Recommendation
Section titled “Slide 5: The Recommendation”Cut Insights from Q3. Ship a data stopgap. Deliver the full product in Q1 2027.
- Insights dashboard: deferred to Q1 2027, target March 13
- CSV export of the same underlying data: ships September 26
- Customers access their data now; the dashboard follows in Q1
Slide 6: The Stopgap Is Real and Ready
Section titled “Slide 6: The Stopgap Is Real and Ready”The underlying data is already queryable. The export is a two-week build.
- Backend endpoint complete: September 19
- Frontend integration and QA complete: September 24
- Release: September 26
- Customers open the file in any spreadsheet or BI tool - no setup required
Slide 7: Why a Clean Deferral Is the Lower-Cost Path
Section titled “Slide 7: Why a Clean Deferral Is the Lower-Cost Path”Rework after a disappointing release costs more than an honest delay.
- Billing migration completes without competing pressure - lower defect risk from context switching
- Q1 Insights ships to the full committed spec with no scope shortcuts
- Trust is repaired through direct outreach, not through a half-built product
Slide 8: The Team Has a Plan
Section titled “Slide 8: The Team Has a Plan”Owners are assigned. Timelines are set. Outreach starts this week.
- CSV export: Dario Reyes - build underway, September 26 release
- Customer outreach: Jordan Park - written notices to four accounts this week; calls scheduled for accounts that flagged strong Q3 dependency
- Insights Q1 kickoff: Maya Chen - design document starts October 6, after billing release stabilizes
Slide 9: What We Need from You
Section titled “Slide 9: What We Need from You”Two sign-offs and one action this week.
- Leadership: confirm March 13, 2027 as the Insights date we can put in customer-facing communications
- Sales: flag any of the four accounts that need a direct call before written notices go out - send to Jordan Park by Thursday
- No objections by end of day Friday means we proceed on this plan
Slide 1: The Same Two Failures, Every Time
Section titled “Slide 1: The Same Two Failures, Every Time”New engineer joins. Without structure, the same two things happen.
- Week 1: blocked on access and tooling, reluctant to interrupt busy teammates, nothing ships
- Week 2 or 3: first change finally lands - too late for the team’s first impression to be reversed, too late for the new hire to feel she belongs
We ship daily. We run a shared on-call rotation. Sustained hand-holding degrades team capacity. Unstructured onboarding does not disappear - it transfers the cost to the new hire.
Slide 2: The Fix - A Named Protocol, Not a Vague Expectation
Section titled “Slide 2: The Fix - A Named Protocol, Not a Vague Expectation”A two-week guided pairing protocol: one named buddy, one pre-scoped first change, explicit daily check-ins.
Three constraints that make it work:
- The team picks the first-change task before day one - scope locked to one service, no on-call blast radius
- Access and tooling have a named owner (the buddy) and are not done until the new hire verifies each item
- Notes belong to the new hire; the buddy does not maintain them
Slide 3: Why Standardize Now
Section titled “Slide 3: Why Standardize Now”The team is adding headcount. On-call load scales with headcount.
- Every unstructured onboarding produces a slower, more expensive ramp than the last as the codebase grows
- Belonging gaps do not heal on their own - each week without structure widens the cost
- We have a pilot running now with Priya (joined Jun 22) that lets us validate the protocol before committing to it as a standard
This is the lowest-friction moment to adopt: one pilot, one DRI, one set of results to review before the next hire.
Slide 4: How the Protocol Works
Section titled “Slide 4: How the Protocol Works”Three domains. Fifteen working days. One clear owner at each stage.
- Days 1-2 - Access and tooling: Buddy-owned checklist; each item marked done only after Priya verifies it herself
- Week 1 - Codebase orientation: Two 90-minute guided walkthroughs - service topology first, then deployment and on-call tooling; new hire keeps the notes
- Week 2 - Paired first change: Pre-scoped task, new hire drives, buddy reviews and pairs on blockers; new hire does the deploy
Goal: Priya’s name in the deployment log before the end of week two.
Slide 5: What Priya’s Week One Showed (Jun 22-26)
Section titled “Slide 5: What Priya’s Week One Showed (Jun 22-26)”The pilot ran the first week of the protocol. Results:
- Full access live by Monday afternoon; one setup doc gap found and patched same day
- Priya navigated three services independently by day 3, without prompting
- Week 2 change fully designed by Thursday; Priya caught an edge case the team had missed
- Priya attended a live incident response and knows the escalation path
- Three teammates have async threads going with her outside the formal onboarding plan
One open risk: staging credentials still pending in the infra queue (submitted Jun 23). Shared credential covers her for now; her own access is needed before the Jul 2 on-call alert drill.
Slide 6: The Real Cost
Section titled “Slide 6: The Real Cost”The protocol is not free. Budget it into sprint planning or it fails silently.
- Week 1: buddy loses roughly 30-40% of capacity
- Week 2: roughly 15-20% as the new hire drives independently
What you buy: an engineer who ships in week two, navigates the codebase on her own, and has real human threads with the team. The unstructured alternative costs the same capacity with no designed outcome and a week-four first ship.
Slide 7: Three Roles, No Dedicated Function
Section titled “Slide 7: Three Roles, No Dedicated Function”- Mei (DRI): Runs the protocol, owns the check-in schedule, tracks the two-week close
- Arjun (buddy): Paired engineer for week two; reviews Priya’s first PR; carries the access checklist in week one
- Team: Owns the pre-scoped first-change decision before each new hire’s start date - the one coordination step that cannot be delegated
No dedicated onboarding function required. The protocol runs on existing team structure.
Slide 8: The Ask
Section titled “Slide 8: The Ask”Adopt the two-week guided pairing protocol as the team standard for all future engineer onboarding.
Three approvals needed:
- Engineering manager: Confirm buddy capacity cost goes into sprint planning at each new hire’s start - not treated as an optional buffer
- Tech lead: Own the pre-scoped first-change task decision before each new hire’s day one
- Next DRI: Name the onboarding DRI for the next hire before Priya’s close-out retro on Jul 3
Two-week retro with Priya on Jul 3 closes the pilot and captures what to carry forward.
The Return on the Alderton Investment
Section titled “The Return on the Alderton Investment”Presented to Dana Forsythe by Sable Marchetti / June 2026
Slide 1: The Problem
Section titled “Slide 1: The Problem”Mentorship investments go unaccounted for.
- Senior contributors spend months of attention on junior colleagues; the returns arrive years later, quietly, in someone else’s work
- The people shaped by that attention rarely report back
- Dana put six months into Sable Marchetti in 2016 and received no accounting since
Slide 2: The Solution
Section titled “Slide 2: The Solution”This deck is the full accounting.
- A formal record of what the Alderton investment produced
- Evidence spanning ten years and one confirmed generation of transmission
- Delivered because the pattern just repeated in a way that made the source unmistakable
Slide 3: Why Now
Section titled “Slide 3: Why Now”The Cassava rebuild closed the loop.
- February 2026: Sable nominated Priya Osei to lead the Cassava data-pipeline rebuild before Priya felt ready
- Week three of March: Priya got stuck; Sable sat on her hands and did not take back the work
- June 2026: Priya shipped four weeks ahead of schedule
- The arc was identical to Sable’s in 2016
That recognition made a ten-year silence impossible to continue.
Slide 4: The Method
Section titled “Slide 4: The Method”What Dana did in 2016, reconstructed.
- Nominated before the person felt ready
- Stayed close without substituting judgment for Sable’s own
- Corrected privately, never during a meeting
- Did not accept the work back when it was offered in a moment of genuine panic
- Stepped back when footing was found
Applied to: Alderton platform migration, March 2016.
Slide 5: Evidence
Section titled “Slide 5: Evidence”What the Alderton investment produced.
- March 2016: Dana nominated Sable for Alderton; Sable said she was not ready
- Fall 2016: Alderton shipped; Sable led the post-mortem without Dana in the room
- 2018 onward: Sable managing and developing teams; the method in active use
- February 2026: method applied to Priya Osei; one generation of transmission confirmed
Slide 6: Return
Section titled “Slide 6: Return”One quarter of mentorship; ten years of compounding.
| What Dana invested | What it produced |
|---|---|
| Six months of attention, 2016 | Enterprise migration on Sable’s record |
| Reputational exposure on Alderton | Durable model for developing people under real stakes |
| Patience while Sable found her footing | Priya Osei’s trajectory in 2026 |
The returns are still running.
Slide 7: Presenter
Section titled “Slide 7: Presenter”Sable Marchetti - the original investment.
- Junior analyst Dana nominated for Alderton, March 2016
- Shipped the migration; led the post-mortem; moved into management in 2018
- Applied Dana’s method to Priya Osei, February 2026
- Writing this deck because the debt became visible and silence felt like a misrepresentation
Slide 8: The Ask
Section titled “Slide 8: The Ask”One thing: receive this accounting.
- No reply required
- No action on Dana’s end
- No debt outstanding
The only ask: know that the six months in 2016 produced something that is still running, and that the person who carries it forward has not mistaken it for luck.
The Case for One Day in Seven
Section titled “The Case for One Day in Seven”Committing to a weekly rest practice - Week 14 of the restart, June 21, 2026
Slide 1: The Problem
Section titled “Slide 1: The Problem”Six unbroken days accumulate a tax that compounds without showing up on any calendar
- Output never fully resets between working days
- Micro-recovery (capped evenings, short walks) was already in place and did not change the underlying pattern
- Decisions late in a long work stretch are measurably slower than at the start of one
- The pull to check one more thing wins every informal attempt at rest
Slide 2: The Practice
Section titled “Slide 2: The Practice”One full day off per week - completely, not “mostly”
- No work output
- No task completion
- No checking of messages or notifications
- The day is not a lighter workday; it is a structurally different kind of day
- A rule about what will not happen, not a schedule of what must
Slide 3: Why Now
Section titled “Slide 3: Why Now”The restart is at week fourteen and holding
- First attempt: ran six weeks before a deadline collapsed it; eleven months passed before a restart
- Restart began early 2026; three consecutive successful rest days as of June 21, 2026
- Phone-away window extended from four hours (week 13) to ten hours (week 14) in a single step
- This is not a new beginning; it is a continuation of a practice that has already proved it can hold
Slide 4: How It Works
Section titled “Slide 4: How It Works”Saturday at 8 p.m. - phone in a drawer until Sunday at 6 a.m.
- No messages checked; no replies drafted in the mind
- The Sunday productivity review (mentally scoring what the week produced) did not run
- One work message sat unanswered from Sunday afternoon until Monday morning - first time this has happened without a mental reply forming
- No clean starting point; the practice resists automation - you cannot schedule the willingness to stop
Slide 5: What Rest Is Not
Section titled “Slide 5: What Rest Is Not”Rest is not a lighter workday; the distinction is structural
Not rest:
- Low-intensity email scanning
- Passive content consumption with inbox awareness running in the background
- A reward earned after enough work is completed
Rest:
- A full stop from output, monitoring, and task-orientation
- Not a rule about what to do on the day - a rule about what will not be done
Slide 6: Evidence
Section titled “Slide 6: Evidence”One clear signal across three weeks
- Week 13: four-hour phone-away window
- Week 14: ten-hour window; first time a work message sat unanswered until Monday without a mental draft forming
- Monday after rest: two problems that had been circling for days resolved more cleanly than they had in the prior week
- The mechanism is not traceable; rest appears to have done something to the thinking
- This is one data point; more weeks required before the pattern is reliable
Slide 7: The Economics
Section titled “Slide 7: The Economics”What the practice costs and what it returns
Costs:
- One day of monitoring and availability per week, every week
- The discomfort of knowing a message went unread for a full day
- Anxiety at the start of the rest window when something unresolved is waiting
Returns:
- The remaining six days carry a different quality of attention
- Decision quality improves when it follows rest rather than arriving at the tail of a long stretch
- What looked like lost time returns as steadiness
Slide 8: The Team
Section titled “Slide 8: The Team”One person committed; fourteen weeks of evidence; one live risk on the table
- Running this: me, now in week fourteen of the restart, on a three-week consecutive streak
- The live risk: a background habit of productivity accounting that tallies whether the rest was “worth it” even during the rest itself
- I have tried to stop it through direct effort; that approach does not work
- Working hypothesis: the habit resolves as evidence accumulates over time, not through direct attack
Slide 9: Open Risks
Section titled “Slide 9: Open Risks”Two risks visible; one has no mitigation yet
Active risk - productivity accounting:
- Runs during rest, not only before and after it
- Named here because unnamed risks erode the practice quietly
- No mitigation identified; direct attack makes the problem worse
Lower risk - Sunday evening re-entry:
- The return to work happens too fast after the rest window closes
- Steadiness built during the day compresses in the first fifteen minutes after the phone returns
- Mitigation under test: cap Sunday evening review at thirty minutes, no replies, triage only
Slide 10: The Ask
Section titled “Slide 10: The Ask”Three commitments by June 28
- Extend the phone-away window to begin Friday at sundown and hold through Sunday evening
- Write a pre-rest log before each rest day - what I expect to feel and what I am afraid of missing
- Name the one check most likely to be rationalized as necessary - write the rule before the moment where I need it
The evidence is early. The commitment is the product. Hold the practice through the next phase of the test.
Slide 1: The Institutional Memory Is Walking Out the Door
Section titled “Slide 1: The Institutional Memory Is Walking Out the Door”- Howard Thayer, Operations Coordinator - 26 years at Crestfield Group
- Joined June 2000. Final day: June 27, 2026.
- Vendor escalation paths, incident decision trees, informal workarounds: none in the official systems
- Four utility contacts maintained personally; not recorded in any official directory
- A backfill hire is approved - the coordination load is covered; the institutional memory is not
Slide 2: Two Problems Are Wearing the Same Job Title
Section titled “Slide 2: Two Problems Are Wearing the Same Job Title”A new hire closes one gap. A continuity program closes the other.
Track 1 - Operational backfill (hire complete):
- Task scheduling and coordination
- Documented runbook execution
- Official vendor contacts and channels
Track 2 - Continuity program (needs approval):
- Undocumented institutional knowledge capture
- Vendor relationship handoff and context transfer
- Informal mentoring re-designated to senior ICs
Slide 3: The Window Is Already Closing
Section titled “Slide 3: The Window Is Already Closing”- Howard’s final day is June 27, 2026. The gap is live.
- Q3 is the first real stress test: on-call incidents without Howard available
- Six to eight weeks post-departure is when absence becomes operationally felt
- Knowledge that was not captured before he left cannot be recovered after
- This program shapes the response now, not after the first Q3 incident
Slide 4: The Program
Section titled “Slide 4: The Program”Three components. Named owners. Delivery dates.
| Component | Owner | Status / Target |
|---|---|---|
| Incident-response runbook | Dana Reyes | Complete, June 2026 |
| Knowledge wiki gap analysis | Marcus Okonkwo | August 15, 2026 |
| Quarterly knowledge-capture practice | Carolyn Marsh | Q3 2026 launch |
| Six-month mentee check-in | Team leads | December 2026 |
The quarterly practice does not end with Howard. It becomes how Crestfield handles every senior departure going forward.
Slide 5: The Foundation Is Already in Place
Section titled “Slide 5: The Foundation Is Already in Place”- Howard spent two sessions in May and June with Dana Reyes and Marcus Okonkwo documenting vendor escalation paths and informal incident decision trees
- Runbook now covers the four utility contacts that had never been recorded anywhere in the official system
- Mentee archive compiled: Priya Sandhu, Ben Holter, and four colleagues contributed specific practices Howard used
- System access and vendor credentials transferred to three named successors by June 20
- All-hands send-off held June 25: 40 in-person, 18 remote
The hard part is done. What is needed now is formal commitment to sustain it.
Slide 6: Senior Time Is the Only Cost
Section titled “Slide 6: Senior Time Is the Only Cost”No budget line. No new hires. No external spend.
- Quarterly knowledge capture: 2 hours per quarter per participating senior IC
- Gap analysis review: one team lead session, August 2026
- Mentee check-in facilitation: one session, December 2026
The alternative: absorbing the gaps at incident time, when the cost is highest.
Slide 7: The Team Already Runs This
Section titled “Slide 7: The Team Already Runs This”- Carolyn Marsh, Operations Lead - continuity workstream owner, point of contact for incoming knowledge contributions
- Dana Reyes - runbook author, on-call incident lead through Q3
- Marcus Okonkwo - knowledge wiki gap analysis lead, Q3 incident review
No new roles. No new reporting lines. Formal accountability added to an existing team.
Slide 8: Three Approvals by July 11
Section titled “Slide 8: Three Approvals by July 11”- Adopt the quarterly knowledge-capture practice as standing Crestfield operations policy
- Protect 2 hours per quarter per designated senior IC, starting Q3 2026
- Send the sponsorship signal to the team by July 11 - before the first Q3 on-call rotation without Howard
Howard spent 26 years making Crestfield resilient to any single departure, with one exception. This program does not replace him. It is the lesson the organization takes from how it lost him.
Slide 1: The Problem We Were Carrying
Section titled “Slide 1: The Problem We Were Carrying”- Cart abandonment elevated for three years - engineering knew the source was checkout
- Codebase: five years of emergency patches, no meaningful test coverage, session coupling the team did not fully understand
- Two previous refactor attempts stalled and were abandoned
- The math became undeniable. We stopped deferring.
Slide 2: The Approach We Chose
Section titled “Slide 2: The Approach We Chose”We had three options. We took the expensive one.
- Option A: Keep patching - 18-month estimate, high regression risk, no guarantee of reaching targets
- Option B: Big-bang cutover - one launch window, no rollback if something failed under real load
- Option C: Parallel run - build the new system alongside the old, migrate by cohort, keep the fallback live until proven
The only way to validate checkout under checkout load is to route real checkout load through it.
Slide 3: Why the Expensive Option Was the Right Option
Section titled “Slide 3: Why the Expensive Option Was the Right Option”- Every previous patch added debt faster than it removed symptoms
- A big-bang cutover gave the team one shot with no recovery path
- The parallel run cost more to operate but gave the team the ability to be right about readiness - not just optimistic
- The cost of not building it right had grown larger than the cost of building it right
Slide 4: Five Phases, Fourteen Months, No User Saw the Migration
Section titled “Slide 4: Five Phases, Fourteen Months, No User Saw the Migration”| Phase | Timeframe | Traffic | What Happened |
|---|---|---|---|
| 1 - Shadow mode | Months 1-4 | 0% live | New pipeline built and instrumented |
| 2 - Canary | Months 5-9 | 5% of sessions | A/B comparison running |
| 3 - Ramp | Months 10-12 | Up to 80% | Near-miss #1 caught and resolved |
| 4 - Pause | Month 13 | Full ramp held | Near-miss #2 found; launch date slipped |
| 5 - Launch | Month 14 | 100% | Final cutover completed under peak load |
Slide 5: What the Data Says Now
Section titled “Slide 5: What the Data Says Now”- Shipped June 13, 2026 - full cutover, zero rollback
- Peak weekend June 13-14 - sustained load, no latency spikes, no rollback triggers
- Cart completion rate held at the target the team modeled in January
- Both near-misses resolved before any user was affected
What the parallel run actually bought us: the ability to find problems under real conditions and fix them before they counted.
Slide 6: What It Actually Cost
Section titled “Slide 6: What It Actually Cost”- 14 months of dual operations - two live systems, two runbooks, two on-call rotations
- Two planned launch windows slipped - once for the February cart-state mismatch, once for the April payment callback rewrite
- Engineering bandwidth that produced no visible features for over a year
- Every slip was the correct call. Each was cheaper than shipping a broken checkout.
Slide 7: The People Who Held This
Section titled “Slide 7: The People Who Held This”Dani Rowe - called the hold on the March launch when the February near-miss was not yet resolved. The call cost three weeks. It was right.
Marcus Teel - filed the cart-state mismatch when he could have marked it low severity and moved on. The fix was his initiative.
Jordan Osei - rewrote the payment callback handler over a weekend when a smaller patch was available and tempting. He chose the right fix.
Sam Wickfield - held the regression bar on June 9 when the schedule pressure was at its peak and every hour of delay felt enormous.
Priya Vasquez - held the line with stakeholders through two slipped dates and kept “we ship when it is ready” from collapsing into “we ship on a date.”
None of this appears in a commit count. It is the kind of work that determines whether a rebuilt system holds under load or fails quietly three months after launch.
Slide 8: Three Things We Need to Close This Right
Section titled “Slide 8: Three Things We Need to Close This Right”-
Recognize the team publicly. The work was invisible to the rest of the organization for fourteen months. The close is the moment to change that.
-
Protect the June 30 retrospective. Fourteen months is not a standard sprint. The retrospective needs space - not a calendar slot - to do justice to what the team learned.
-
Hold the post-launch window. Mia Chen’s analytics team delivers the cart-abandonment baseline on July 7. The six weeks after that date need to be clear for the team to read the results and act before the next planning cycle.
The team earned the ask. These three things are how we close the project the same way we ran it.
Slide 1: Our “Flexible” Policy Is Costing Us on Both Sides
Section titled “Slide 1: Our “Flexible” Policy Is Costing Us on Both Sides”- Senior leaders: collaboration has atrophied; new-hire trust formation is measurably slower
- Employees: have relocated, restructured childcare, and accepted roles here because flexibility was sustainable
- Talent pipeline: two recent hiring cycles confirm geography closes doors under a full in-office mandate
- Ambiguity is not neutral - both groups are losing ground at the same time
Slide 2: Both Camps Are Right About Something
Section titled “Slide 2: Both Camps Are Right About Something”The office-first case is real:
- Trust builds faster in the same room
- Small decisions that once resolved in corridor conversations now languish in threads
- New hires without in-person exposure show coordination drag on mixed-tenure teams
The remote-first case is also real:
- Talent pools shrink when geography is a filter
- Commute time is real time taken from employees
- Flexibility was a stated condition of employment; reversing it breaks faith
The policy cannot dismiss either side. It has to hold both.
Slide 3: The Anchor-Day Hybrid
Section titled “Slide 3: The Anchor-Day Hybrid”Two fixed in-office days. Three genuinely flexible days. No binary required.
- Anchor days: Tuesday and Thursday - all employees not designated remote-eligible at hire
- Flexible days: Monday, Wednesday, Friday - employee decides where to work the day before
- Exceptions to anchor days require documented approval from manager and the people team
- Remote-eligible roles (defined at hire) are explicitly excluded from any mandate
In-person time is highest value when it is rare, anticipated, and shared - not compulsory and continuous.
Slide 4: The Window Is Closing Before the All-Hands
Section titled “Slide 4: The Window Is Closing Before the All-Hands”Two internal tracks are headed toward a public contradiction.
- Facilities is already operating a room-booking policy premised on five-day attendance
- HR has not communicated a work-location policy
- An all-hands announcement is scheduled
- If the two tracks are not aligned first, the official policy will contradict room-booking rules already in effect
The decision must be made before the announcement - not after it.
Slide 5: How the Policy Works in Practice
Section titled “Slide 5: How the Policy Works in Practice”| Day | What Is Required |
|---|---|
| Tuesday | In-office (anchor day, load-bearing) |
| Thursday | In-office (anchor day, load-bearing) |
| Monday, Wednesday, Friday | Flexible - employee decides |
- Flexible days require no approval in either direction
- Anchor-day exceptions require documented approval from manager and people team
- Anchor days are not preference days - they are the enforcement structure of the policy
Slide 6: What Each Side Gives and Gets
Section titled “Slide 6: What Each Side Gives and Gets”| Stakeholder | Gives | Gets |
|---|---|---|
| Office-first leaders | Five-day presence; constant hallway access | Two guaranteed shared windows per week |
| Remote advocates | Full schedule autonomy | Three genuinely flexible days, no approval required |
| Organization | Some recruitment breadth | Retained access to talent markets with no local office |
Full in-office closes the non-local talent market entirely. This model does not.
Slide 7: The Working Group Has Built the Case
Section titled “Slide 7: The Working Group Has Built the Case”Position Brief v2 circulated to the leadership team week of June 20.
Major objections answered in the brief:
- “Trust needs daily proximity” - Anchor days create predictable overlap; constant co-location is not required for spontaneous contact
- “Any mandate narrows the talent pool” - Two fixed days stated upfront is a smaller and more honest constraint than an unwritten five-day expectation discovered after hire
Open item - Manager FAQ:
- Covers anchor-day operations, accommodation handling, and new-hire onboarding under hybrid
- Target delivery: June 27
- Blocked until decision authority is confirmed
Working Group Lead: Priya Ahluwalia
Slide 8: Two Asks. Both Needed by June 28.
Section titled “Slide 8: Two Asks. Both Needed by June 28.”Ask 1 - Confirm decision authority. Which team owns the final policy call - Facilities or HR? Without a named owner, the Manager FAQ cannot be finalized and the all-hands announcement will contradict room-booking rules already in effect.
Ask 2 - Read Position Brief v2 by June 27. Flag any objection not already addressed in the brief. Goal: arrive at the June 28 leadership briefing with all known objections resolved on paper, not surfaced in the room for the first time.
Leadership cohort endorsement is scheduled for July 3. Both asks must be resolved before that session.
Slide 1: Problem
Section titled “Slide 1: Problem”Small product teams collect customer feedback everywhere. No one knows what matters most.
- Feedback lives in support tickets, chat threads, survey exports, and call notes - spread across whatever tools the team already uses
- Turning that scatter into a ranked priority list means someone reads everything manually, copies items into a spreadsheet, and argues about priority in a meeting
- Teams of 2 to 20 people own the roadmap but have no dedicated research function - so this work either does not happen or happens too slowly to drive decisions
Slide 2: Solution
Section titled “Slide 2: Solution”Tidemark turns scattered customer feedback into a ranked, shareable roadmap.
- Connects to your existing feedback sources - no migration required
- Identifies recurring themes across all sources and scores them by frequency and recency
- Produces a single ranked view your team can share with stakeholders in minutes, not days
Slide 3: Why Now / Market
Section titled “Slide 3: Why Now / Market”The gap between collecting feedback and knowing what to build is wider than ever.
- Product teams have grown without growing research capacity
- Roadmap tools assume you already know what to build; research tools require dedicated analysts to do the synthesis
- No good solution exists today for the connective layer between scattered input and a decision-ready view - not for small teams
Slide 4: Product
Section titled “Slide 4: Product”Three commands from scattered feedback to a shareable roadmap.
tidemark init # connect your existing feedback sourcestidemark sync # pull, rank, and surface recurring themestidemark share # send a read-only roadmap link to any stakeholder- Works with the tools teams already use; no data migration
- The shared view is read-only by default - stakeholders can see and comment, not override the ranking
Slide 5: Traction / Evidence
Section titled “Slide 5: Traction / Evidence”22 early-access teams. Zero support tickets. One clear workflow.
- 22 small teams completed the full feedback-to-roadmap loop before public launch
- All 22 imported feedback from scattered sources, received a ranked thematic summary, and shared a roadmap link with their stakeholders
- Not one team filed a support request during the cohort
- Every team at intake named the same pain: maintaining a separate spreadsheet to consolidate what users had asked for
Slide 6: Business Model
Section titled “Slide 6: Business Model”Free to start. $29 per month when the team grows.
- Solo plan: Free for individuals; full access, one user
- Team plan: $29/month for up to 15 seats; no feature gating
- Custom plan: Organizations that need SSO and audit logs contact us directly
- Self-serve on all plans; no sales call required to get started
Slide 7: Team
Section titled “Slide 7: Team”Product-led. Built for self-serve from day one.
- Marisol Veen, Head of Product - managed roadmaps at small companies where no research function existed; the problem Tidemark solves is firsthand
- Core team covers product, engineering, and design
- Go-to-market is community and organic reach; no dedicated sales headcount, by design
Slide 8: Ask
Section titled “Slide 8: Ask”Tidemark is live at tidemark.io as of June 30, 2026.
- Press and writers: a demo video, product screenshots, and a one-page summary are available at launch@tidemark.io - no embargo, no exclusivity conditions
- Community builders: one post or mention in the first two weeks helps more than a general announcement; a concrete before-and-after from our 22-team cohort is available to share
- Potential integration partners: the source connection layer is where we are building first; reach out at launch@tidemark.io to talk about what your platform’s feedback looks like
Slide 1: Two Load-Bearing Things Failed at Once
Section titled “Slide 1: Two Load-Bearing Things Failed at Once”The year produced two simultaneous losses.
- March: Meridian dissolved when the primary funder withdrew after eighteen months of coalition work
- April through August: Celeste’s friendship drifted past the point of easy return
- Neither failure was caused by poor work - both still cost exactly what they cost
The downstream risk is not the losses themselves. It is the default response: quiet retreat, shallower investment, less on the line next time.
Slide 2: The Default Response Is Already Happening
Section titled “Slide 2: The Default Response Is Already Happening”Retreat is not a decision - it is what happens when no decision is made.
The three available postures:
- Reframe - find the hidden gifts, produce a tidy growth narrative
- Protect forward - narrower commitments, shallower investment, less to lose
- Reckon honestly and choose exposure anyway
Option 1 is not honest. Option 2 is already hardening by default. That is the problem. A default is not a posture; it is a drift.
The ask is to name Option 3 explicitly before Option 2 becomes the answer.
Slide 3: Year-End Is the Last Moment Before the Pattern Calcifies
Section titled “Slide 3: Year-End Is the Last Moment Before the Pattern Calcifies”The window for a deliberate choice closes when the next large project starts.
- The losses are recent enough to examine; distant enough to see
- Professional capacity held all year - client work shipped, no major failures
- The pattern is already identified: conflated momentum with progress, managed perception instead of naming hard truths, confused giving-space with avoiding
- Drift toward protection is already in motion; naming it now is the only way to interrupt it
The next large project cannot start yet. That is the condition that keeps this choice available.
Slide 4: Three Moves Before Anything Else
Section titled “Slide 4: Three Moves Before Anything Else”What this approach looks like in practice:
- Write the Meridian retrospective - a real account for the eleven people who gave time to this work, naming what the decision-making looked like from the inside, not a press release (target: February)
- Answer Theo’s message from April - the longer the wait, the harder it becomes; that is not a reason to keep waiting (target: January)
- Repair before evaluating - do not assess the next large effort until at least one under-maintained relationship is actively rebuilt
Sequence matters. These are not parallel. Retrospective and repair come before any new evaluation.
Slide 5: The Capacity Held All Year
Section titled “Slide 5: The Capacity Held All Year”Evidence that this is not asking for something that is gone:
- Deliverables shipped on schedule, full year, no professional failures
- September self-diagnosis: identified the core error clearly enough to name it - managed perception to protect the project rather than deliver the hard truth, then delayed a reckoning that arrived anyway, with less trust intact
- Did not construct an exculpatory story about Celeste; named what the personal failure was
- This reckoning is happening rather than not happening
The capacity ran parallel to the losses. It did not recover after the losses. It was there the whole time.
Slide 6: What This Posture Costs and Returns
Section titled “Slide 6: What This Posture Costs and Returns”| Costs | Returns |
|---|---|
| Future losses land at full weight - no protection from the same kind of year | Capacity for full investment in the next meaningful work, by choice rather than inertia |
| Unresolved questions about Celeste stay unresolved | Attention freed from reconstructing a narrative that does not exist |
| Choosing exposure does not reduce the probability of similar loss | The Meridian work stands as what it was, without depending on anyone else to confirm it |
The alternative posture costs less upfront. It costs more across the next ten years.
Slide 7: Who Is Making This Pitch
Section titled “Slide 7: Who Is Making This Pitch”Marcus Delgado - track record for context:
- Led the Meridian coalition for eighteen months: held eleven contributors aligned, produced a technically sound infrastructure proposal, managed the coalition through the funder withdrawal
- Six years of close friendship with Celeste before this year’s drift
- Identified the failure pattern within the same year it played out, not in retrospect years later
- Maintained professional function under conditions that made function feel pointless
Known failure pattern: Starting new things to avoid finishing difficult existing ones. Meridian’s retrospective does not yet exist for this reason. Theo’s message from April has not been answered for this reason. Naming it here is the first move against it.
Slide 8: Three Commitments, In This Order
Section titled “Slide 8: Three Commitments, In This Order”The ask is not a vision. It is a sequence.
- Answer Theo before the end of January
- Draft and deliver the Meridian retrospective before the end of February - the real account, not the short version
- Do not evaluate or start a new large effort until both of the above are done and at least one under-maintained relationship is in active repair
The condition: The next project does not start before this sequence completes.
This is not a roadmap for what comes next. It is the precondition that makes whatever comes next worth trusting.