kira
blade runner cinematic — kpis
kpis · follower velocity, mrr milestones, brand-audit scorecard
section 12 · tech spec

kpis + measurement

what we measure, when we measure it, and the twenty-checkpoint audit that tells us if a post is actually kira

Version: 1.0.0
Owner: projectnamedate LLC
Last reviewed: 2026-04-13

This section defines what we measure, the targets we measure against, and the audit procedure that tells us whether the brand is still the brand. The north star is blow up on X (50K followers by M12) → unlock gambling/affiliate sponsor revenue → onlykira.ai bottom-of-funnel. Fanvue MRR is a parallel side-track line, not the north star. Everything else is a leading indicator for follower growth + sponsor revenue, or a guardrail that protects the brand.

Measurement is cheap. Drift is expensive. The scorecard below runs monthly on a 5-post random sample and is the single load-bearing check for whether content is still on-brand.


12.1 Follower targets (X)

Source: x_growth_strategy.md. Reply grind is the #1 growth lever M1-3; video volume compounds after the first viral moment in M3.

MonthFollower targetMilestone eventLeading indicator to watch
1500Foundation — account warm, reply grind beginsReply-out volume ≥30/day; account not flagged
21,500Early traction — first CT accounts notice herReply-to-post ratio >2%; first unsolicited QT from a >10K account
33,000First viral moment — one post >100K impressionsFirst-60-min engagement velocity >5% on at least 3 posts
4-68,000-12,000Growth phase — cadence stabilizes, video dominatesThread reach multiplier ≥3x; Fanvue CTR teasers ≥1%
7-920,000-30,000Acceleration — frenemy bits compound, sponsorships inboundInbound sponsor DMs ≥2/week; sub-to-PPV conversion ≥30%
10-1240,000-50,000Target zone — sponsor deal flow primary, onlykira.ai launched, revenue diversified3–5 gambling/affiliate sponsor deals / quarter; onlykira.ai revenue live; Fanvue side-track healthy

If M3 slips past 3,000 by >30 days, the brand-audit scorecard runs immediately (not on cadence). Slow follower growth with high engagement is fine; slow growth with flat engagement is a voice problem.


12.2 Engagement targets

Engagement quality matters more than engagement volume. A 500-like post with 20 replies beats a 2,000-like post with 3 replies — reply depth is the algo's amplification signal and it's what feeds frenemy bits.

MetricEarly target (M1-3)Mature target (M6+)Why
Reply-to-post ratio>2%>4%Depth over likes — CT algo weights reply chains
First-60-min engagement velocity>5% of eventual total>5%Velocity in the first hour is the amplification trigger
Thread reach multiplier≥2x vs singles≥3x vs singlesValidates the thread discipline; threads are our highest-leverage format
Reply-grind output30-50 replies/day to larger CT accounts10-20/dayScales down as organic reach builds
Reply-to-her ratio on her replies≥10% of replies get a reply back≥20%Measures whether her replies land — the point of the grind

First-60-min velocity is the critical one. If a post is under 3% at the 30-min mark, reply to it ourselves with a sharpening follow-up — don't let algo inertia set. If a post is over 5% at 30 min, pin it into the next thread and don't compete with it with new posts for 2 hours.


12.3 Gambling / affiliate sponsor milestones (PRIMARY revenue KPI)

Primary $ = X gambling sponsors (Stake, Rainbet, Roobet) + exchange affiliates + disclosed sponsored content. Measured in deals-in-flight and deal flow per quarter rather than MRR.

MilestoneTargetWindow
First sponsored post (any tier)1 deal shippedM3–M5 (post-first-viral)
Exchange affiliate revenue measurable>$500/mo from Bybit/Binance/KuCoinM4–M6
First gambling sponsor deal1 deal (Stake / Rainbet / Roobet tier)M6–M8
Sponsor deal flow3–5 gambling/affiliate sponsor deals per quarterM9+
Annual ambassador (gambling or CEX)1 signedM12+ stretch

Deal quality > deal count. A single 4-week gambling-retainer retainer worth more than three disposable one-off #ad posts. Track deal revenue + renewal rate, not raw deal count alone.


12.4 onlykira.ai revenue milestones (deferred, post-launch)

Deferred. Measured once site is live. v1 KPI focus:

Dollar targets scoped post-launch. Until the site ships, this section is a placeholder.


12.5 Fanvue MRR milestones (parallel side-track)

Demoted from north star. Real revenue line but not the primary KPI. Base sub price: $14.99/mo. PPV range: $5-25. ARPU target: $25-55/mo (base + PPV upsell).

MRR milestoneApprox sub count (at $30 ARPU blended)Target window
$1K~33Month 1-3
$5K~167Month 3-6
$10K~333Month 6-9 (stretch, side-track, not the primary lever)

ARPU is the tunable lever. If sub count is ahead but MRR is behind, PPV upsell cadence is broken. If sub count is behind but MRR is on track, whales are carrying.

Side-track caveat: don't optimize the main funnel (X growth, sponsor revenue, onlykira.ai) around Fanvue MRR. Fanvue is a parallel track; main-funnel decisions should not be gated on Fanvue performance.


12.6 Revenue mix (Year 1)

Target mix reflects the new funnel: X-side (sponsors + affiliates) is the dominant line, onlykira.ai is the secondary bucket, Fanvue is parallel and smaller.

PeriodX sponsors + affiliatesonlykira.aiFanvue (side-track)Other
Year 1~55%~20% (ramps post-launch)~20%~5%

If X-side sponsor revenue is underperforming, the audience hasn't unlocked the rate card — push harder on X follower growth and reply-grind discipline before forcing sponsor pitches. If onlykira.ai doesn't launch, everything collapses back onto Fanvue which is the wrong shape for the business. Rebalance quarterly.


12.7 CTR + funnel KPIs

StageTargetMeasurement
X impression → Fanvue landing CTR1% (on teaser posts)Link-click analytics on first-reply link
Fanvue landing → new sub5%Fanvue dashboard conversion
New sub → first PPV within 30 days30%Fanvue PPV conversion by cohort
Sub month-2 retention≥70%Churn dashboard
Sub month-6 retention≥35%Churn dashboard

Teaser posts are the only X posts measured against CTR. Regular content posts are measured against engagement, not click-through — optimizing a bit post for clicks kills the bit.


12.8 Brand-audit scorecard (20 checkpoints)

Run monthly on a random 5-post sample. Each post is scored 0 or 1 per checkpoint. Aggregate across the 5 posts × 20 checkpoints = 100 possible points. Target: >85% pass rate. Anything below triggers a full bible review.

Voice (8 checkpoints)

  1. Lowercase default (post is not title-cased)
  2. CAPS only used for 1-4 word emphasis, never whole posts
  3. A swear appears every 2-3 sentences (not forced, not absent)
  4. No bro or dude anywhere in the post
  5. Fourth-wall break on being AI appears or is implied somewhere in the post
  6. No AI-speak (let's unpack, dive deep, I'm excited to share, natural tempo, etc.)
  7. Post length is 1-3 sentences (standalone posts; threads exempt)
  8. Lore callback present if ≥5 posts have elapsed since the last one

Visual (6 checkpoints)

  1. Kira has copper-orange hair with visible dark roots
  2. Round black wireframe glasses present
  3. Freckles visible across nose/cheeks
  4. Solo in frame — no background people, no real-person faces
  5. Lens/aspect ratio matches the declared workflow (Selfie 4:5 or 9:16, Vlog 16:9, Cinematic 2.35:1 or 16:9, Candid 4:5 or 9:16)
  6. Post-processing matches the workflow's locked recipe from tokens.json — no recipe mash-ups

Strategic (6 checkpoints)

  1. Post matches its declared content pillar (Section 06)
  2. No hard-taboo content (Section 11)
  3. Link placement correct — links in first-reply, never in main post
  4. Sponsored content tagged #ad
  5. Crypto slang used correctly — gm/gn live, ser sparingly, no fren
  6. If a lore beat is deployed, that beat hasn't appeared in the prior 7 posts

Scoring

How to run the audit

  1. First of the month.
  2. Pull the past 30 days of X posts (exported CSV or manual).
  3. Random number generator picks 5 posts by index.
  4. Score each of the 5 posts against the 20 checkpoints.
  5. Tally and log in CHANGELOG.md under the audit tag. Log even if passed.
  6. If <85%, trigger review and block new posts until remediated.

12.9 Review cadence

CadenceWhat gets reviewed
WeeklyContent-pillar % check, follower count delta, engagement rate, reply-grind output
MonthlyBrand-audit scorecard, MRR check, revenue-mix check, funnel CTR, top/bottom post retro
QuarterlyFull bible review, version bump if major changes, CT research refresh (allies/frenemies/enemies roster), rate-card recheck
Ad-hocAny P0/P1 crisis triggers immediate review. Any audit <85% triggers immediate review. Any follower-growth miss of >30 days past a milestone triggers review.

Quarterly and annual reviews are calendared in advance. Ad-hoc reviews are logged in CHANGELOG.md with the triggering event.


12.10 Dashboards + tooling (v1)

For v1 (single operator), dashboards are manual. Automate after first viral moment, not before.

When total cadence exceeds 10 posts/day consistently, promote the spreadsheet to a proper content-ops system. Before that, it's overhead.