Dashboard Operating Manual
How to capture work, run the business, and keep Faktory Ops trustworthy without turning dashboard maintenance into another job.
1. The operating model
Faktory Ops is the operating record for what is happening across Faktory Studios: current work, backlog, experiments, customer evidence, metrics, finances, commitments, decisions, and project health.
| Cadence | Purpose | Typical actions |
|---|---|---|
| During the day | Do not lose information | Use Universal Capture for ideas, activity, expenses, feedback, metrics, and tasks. |
| Daily | Reconcile reality | Clear captures, scan Attention, confirm non-GitHub activity, and let GitHub/ChatGPT reconciliation update evidenced development work. |
| Weekly | Choose and learn | Set one primary outcome, supporting tasks, review experiments, refresh key metrics, clean backlog. |
| Monthly / Quarterly | Run the business | Reconcile actual revenue/expenses monthly; review portfolio, strategy, positioning, and competitors quarterly. |
Trust hierarchy
- GitHub proves code activity, not human validation or customer publication.
- A missing metric means unknown, not zero.
- Recorded financial actuals remain separate from estimates and subscription projections.
- Human validation and publication are explicit states.
2. Home: the daily command center
Open Home first. It should answer: What deserves attention? and What should I do next?
- Universal Capture - fastest place to record something before it is forgotten.
- Review cadence & data freshness - what information is becoming stale.
- This week's outcome - the result the week is organized around.
- Next actions - selected work, not the whole backlog.
- Attention queue - exceptions such as blockers, overdue tasks, stale data, feedback, and experiment reviews.
- Upcoming commitments - dates that can surprise you if they are not visible.
3. Universal Capture: your default input
Use Universal Capture whenever something matters but you do not want to interrupt your current work to fill out a full form.
| Prefix | Use for | Example |
|---|---|---|
activity: | Something that happened | Posted the ORDO article to X and Bluesky |
task: | A concrete action | Follow up with the ORDO voucher recipient |
idea: | A possibility worth preserving | Add a living project map to ORDO |
experiment: | A test designed to learn | Test whether an ORDO Short drives store visits |
feedback: | Customer evidence | First scan feels overwhelming |
metric: | A measured value | ORDO pageviews were 420 this week |
expense: | Money spent | Paid $39 for a Unity plugin |
revenue: | Money received | ORDO sale posted this month |
decision: | A choice worth remembering | Pause Sideline Atlas until ORDO 1.3 ships |
Prefixes are optional. The raw capture remains intact until you convert or archive it.
Conversion
- Capture the rough sentence.
- Confirm or change the suggested record type.
- Select a project if applicable.
- Fill only the required facts.
- Convert. The structured record is created while the original note is preserved.
4. Weekly planning: focus without losing the backlog
- Choose one primary outcome.
- Select a small number of supporting tasks.
- Estimate hours when it will expose capacity problems.
- Defer work explicitly instead of pretending everything is active.
- At week end, record completed, carried forward, or dropped outcomes and lessons.
4A. CEO reviews: preserve the executive view
Use the CEO Reviews tab after the ordinary weekly closeout and for a consolidated monthly rollup. It keeps dated highlights navigable without replacing the underlying tasks, metrics, finances, evidence, or canonical strategy documents.
- Record the review period and cadence (weekly or monthly).
- Attach a source period and basis to every key metric.
- Separate wins, blockers, and opportunities.
- Close with exactly three priorities.
- List unavailable or partial data explicitly; missing never means zero.
- Link the canonical Markdown review under
business/ceo-reviews/.
Adding a review
- Close the Weekly plan or reconcile the monthly Money scorecard.
- Verify source values and preserve mixed periods or gross/net bases separately.
- Create the canonical weekly
YYYY-MM-DD.mdor monthlyYYYY-MM.mdfile from the repository template. - Add the structured mirror through
/api/insights?resource=ceo-reviews. Corrections become explicit amendments rather than silent rewrites.
5. Backlog, ideas, and outcomes
Tasks
Use tasks for concrete executable work. Use due dates only for real deadlines; use revisit dates for “think about this again later.”
Ideas
Use ideas for possibilities that have not earned execution yet. Move them through exploring, experimenting, adopted, parked, or rejected.
Outcomes
Use outcomes to group tasks around a business result.
6. Experiments: turn uncertainty into learning
Use an experiment when you are deliberately testing a belief rather than doing routine work.
| Field | Question |
|---|---|
| Hypothesis | What do I believe will happen? |
| Method | What exactly will I do? |
| Success criteria | What would count as evidence in favor? |
| Stop criteria | When should I stop? |
| Effort/spend limit | What is my maximum exposure? |
| Review date | When will I evaluate the result? |
Example: If an ORDO YouTube Short demonstrates a painful Unity problem and links directly to the Asset Store, it will generate measurable product-page traffic within seven days.
7. Evidence and metrics
Prefer a small number of metrics that change decisions over a large collection of numbers.
- Create stable metric definitions per project.
- Log dated observations so trends become visible.
- Use CSV import when a service exports structured data.
- Use manual metric capture for one-off measurements.
- Record source, period/cohort, and caveats when relevant.
- Never enter zero because data is unavailable.
Data freshness
The freshness panel records whether a source was reviewed. It does not create metric or financial values.
| Source | Cadence |
|---|---|
| Non-GitHub activity check-in | Daily |
| Unity Asset Store analytics | Weekly |
| Kajabi analytics | Weekly |
| YouTube analytics | Weekly |
| Social channel metrics | Weekly |
| Revenue and sales actuals | Monthly |
| Expenses and subscriptions | Monthly |
| Portfolio and strategy review | Quarterly |
8. Money: actuals vs estimates
- Revenue/expense entries - dated business actuals.
- Bank imports - reviewed transactions rolled into actuals.
- Project revenue - current monthly estimate or live synced value.
- Subscriptions - recurring projected costs.
- Recorded Actuals - historical actual entries only.
- Income vs Expenses estimate - revenue estimates vs normalized subscription costs.
9. Activity, project pages, commitments, and decisions
Activity
Log feature-level shipped / next / blocked / note updates. Avoid commit-level noise.
Project pages
Use the project page to understand current status, notes, release/validation facts, GitHub evidence, updates, links, and project metrics.
Commitments
Use commitments for dated obligations not already represented by a task due date, release date, experiment review, or subscription renewal.
Decisions
Record choices likely to be re-litigated later. Capture rationale, tradeoffs, evidence to revisit, and review date when useful.
10. Working with ChatGPT and other agents
- ChatGPT can reconcile GitHub activity with dashboard state and write high-confidence updates with
source=chatgpt. - You can say “Capture this for Faktory: ...” instead of opening the dashboard.
- Development evidence can often be reconciled automatically.
- Metrics, money, human validation, publication, experiment decisions, and strategy should only change when evidence is direct.
- If sources conflict, surface and reconcile the conflict instead of silently choosing one.
11. Recommended routine
Daily - 3 to 5 minutes
- Open Home.
- Capture anything GitHub cannot know.
- Triage the capture inbox.
- Scan Attention.
- Confirm the non-GitHub activity check-in.
- Let daily ChatGPT reconciliation handle evidenced development changes.
Weekly - 15 to 30 minutes
- Review last week's outcome and lessons.
- Set one primary outcome and supporting tasks.
- Review experiments due soon.
- Refresh weekly metric sources.
- Clean stale backlog items.
- Review the next two weeks of commitments.
- Publish the verified weekly CEO highlights and three priorities.
Monthly - 30 to 60 minutes
- Import/review bank data.
- Confirm revenue and sales actuals.
- Review scorecard vs targets.
- Review subscriptions.
- Inspect metric trends.
- Publish the monthly CEO rollup only after reconciliation; list any missing sources.
Quarterly - 60 to 90 minutes
- Review the portfolio.
- Review product positioning and competitors.
- Evaluate experiments and customer evidence.
- Decide what to continue, pause, or invest in.
- Update strategic decisions and next-quarter outcomes.
12. Quick reference: where does this go?
| I have... | Put it in... |
|---|---|
| A rough thought I may forget | Universal Capture |
| A concrete action | Backlog Task |
| A possible direction | Idea |
| A testable belief | Experiment |
| Something shipped, blocked, or learned | Activity Update |
| Customer evidence | Feedback |
| A number measured on a date | Metric Observation |
| Money received or spent | Revenue / Expense |
| A hard date or obligation | Commitment or Task Due Date |
| A choice worth remembering | Decision |
| Work selected for this week | Weekly Plan |
| Verified executive highlights and next priorities | Canonical CEO review + CEO Reviews tab |
| Something requiring intervention | Attention Queue (generated from the underlying record) |