2.4 Backstory and the Narrative Map
A The six-part specification that makes a persona reproducible as a character across months. describes a person. It does not describe a life in progress. That distinction is the difference between an account that people return to and one that becomes wallpaper.
Wallpaper is the failure mode of synthetic accounts. The face is consistent, the aesthetic holds, and after three weeks a viewer has seen everything there is to see, because nothing has happened and nothing can happen. A persona without a timeline cannot accumulate anything.
Three acts
The backstory is short and it has a shape. Three sentences are enough.
Before. Where she came from and what it cost her. This supplies motivation and explains her current situation. Two or three specifics: a city, a decision, something she left behind.
Now. The present project. She is doing something with a goal, and the goal is not yet achieved. This is the act she is living in, and it is where the content comes from.
Toward. What she is working toward, and what stands in the way. This supplies the reason a viewer returns next week: they want to know how it goes.
The third act is the one operators omit, and its absence is exactly why their accounts feel like a catalogue. A persona with no unresolved goal has no narrative tension, and narrative tension is what makes the next post interesting rather than merely presentable.
Worked shape, using the archetypes from 2.2:
| Guide | Companion | Underdog | |
|---|---|---|---|
| Before | Trained in a field, left it | Moved to a new city, knows nobody | Failed at something publicly |
| Now | Teaching what she learned, building an audience | Building a life, one small ritual at a time | Trying again, visibly, with a deadline |
| Toward | A complete body of work | A settled, familiar world | A specific target, with a real chance of failing |
Notice the underdog has the strongest third act and the guide the weakest — which is the structural reason chapter 1.4 recommends the underdog archetype to anyone worried about the trust gap.
The twelve-month map
The backstory is fixed; the timeline moves. Sketch twelve months before you publish anything, because this map is what turns "what do I post today" into "what is happening this week."
| Month | Season | Event | Appearance | Content consequence |
|---|---|---|---|---|
| 1 | Launch of the present project | Base look | Introduction, positioning | |
| 2 | First milestone attempt | Progress, failure, adjustment | ||
| 3 | A setback that is planned in advance | Vulnerability, the underdog beat | ||
| 4 | Seasonal change | Hair or wardrobe shift | Visual refresh, new locations | |
| 5 | Travel or a location change | Environmental variety | ||
| 6 | Half-year review | Reflection format, audience Q&A | ||
| 7 | A collaboration or a new recurring character | Two-hander formats | ||
| 8 | A skill or habit acquired | Fitness or styling change | Progression content | |
| 9 | A recurring event (a competition, a season, a tradition) | Anniversary and callback formats | ||
| 10 | A small crisis | Where the canon from 2.3 is tested | ||
| 11 | Consolidation | Best-of, lessons-learned | ||
| 12 | Anniversary of month one | Second base look | Retrospective, next year's plan |
Three rules for using it.
Plan the low points in advance. A persona whose timeline only rises is not credible and is not interesting. Decide now when she struggles, so the struggle is designed rather than improvised.
Change the appearance on a schedule, not on impulse. Hair and styling changes are the cheapest visible evolution available (2.3), and they are also the fastest way to break identity consistency if they happen often and unpredictably. One or two per year, planned.
Leave at least four months empty. The map is a skeleton, not a script. Events that arrive from outside — a trend, a platform change, a brand approaching you — need room.
Continuity management: the register
The bible holds the rules. The register holds what has already happened.
Keep it as a simple table in the persona's folder, appended after each published post that establishes anything:
| Date | Fact established | Where | Constraint it creates |
|---|---|---|---|
| 2026-11-03 | Sister named, mentioned once | Post caption | Name fixed; frequency capped by canon |
| 2026-11-14 | Moved apartments | Photo carousel | Prior apartment is retired until a planned move |
| 2026-12-01 | Started a running routine | Story | Fitness arc opens; needs follow-up in weeks 2–6 |
The third column is the one people skip. A fact with no recorded constraint is a fact you will contradict, because the constraint is the only part that affects future decisions.
The rule for handling contradictions: extend, never contradict. If a post implies something that conflicts with the register, you have three options, in order of preference. Reframe the new post so it is compatible. Treat it as a change of circumstances and say so in character. Or, in the worst case, correct it in character — which is a strength if the The ten facts a persona cannot break, each specific enough to check in ten seconds. from 2.3 includes how she handles being wrong.
Reviewing the register
The register is only a register if it is read. Two cadences, both cheap.
After each publishing session, append anything the post established. This is part of the production slot from 2.6 and takes under a minute, because it happens while the details are in your head. Reconstructed later, it is guesswork.
Weekly, in the review slot, read the last two weeks of entries and ask one question: does the next month's map still fit what has actually happened? A timeline drifts the way an account does — not through one bad decision but through five small ones that were individually reasonable.
The register's second use is answering a question you will be asked constantly and cannot answer from memory: has she already said this? Repetition is the most common continuity failure in a persona account, it is invisible to you, and it is obvious to an audience that has read everything.
When the audience catches a contradiction
They will. A synthetic persona's audience is unusually attentive to inconsistency, because inconsistency is the evidence they are looking for about whether a person is behind the account. A commenter quoting a March caption against an August post is not being hostile; they are doing exactly the thing the medium invites.
Three responses, and only one of them is good.
Delete the post. Worst option. It confirms the inconsistency, removes the evidence, and reads as concealment.
Ignore it. Sometimes survivable. It stays in the comments as the top reply, which means every new viewer reads it.
Answer in character, within the canon. Best option, if the canon gives you a mechanism. A persona who is allowed to be wrong, and who has a stated habit of correcting herself within a week, can turn this into the most human moment in her timeline. That mechanism does not exist unless you wrote it in chapter 2.3's canon before you needed it.
This is the practical argument for the whole part. The bible and the timeline are not administrative overhead; they are the thing that lets you respond well at the moment you are least able to think.
Where this goes wrong
No third act. The account presents a person with nothing unresolved, and viewers have no reason to return.
A timeline that only rises. No setbacks means no stakes, which returns you to the trust problem from 1.4.
Improvised events with no register. Continuity is your memory's job until the library outgrows it.
Appearance changes on impulse. Every unplanned change costs you identity consistency work.
Filling all twelve months. You have removed the capacity to respond to anything.
Deleting a contradicted post. It converts a continuity slip into an admission.
The gate
- Write the three acts in three sentences.
- Is the third act unresolved, and can a viewer see what is at stake?
- Does the twelve-month map include at least two planned low points and four empty months?
- Is the continuity register created, with at least the constraint column defined?
- Does the canon contain a stated mechanism for being wrong?
What you leave with
A three-act backstory with a third act that supplies tension, a twelve-month map that turns content planning into retrieval, three rules for using it without removing flexibility, a continuity register with the constraint column that most operators omit, a three-option response to a public contradiction ranked by outcome, and the reason all of it exists: it is what lets you answer well on the day you cannot think.