Inicio/Blog/How to Pick a Winning AI Short Drama Topic: Topic Pools and Filters from 3 Lollipop Drama Creators
Economía de creadores 2026-09-18 Evelyn Cho · Lollipop Drama Content Lead

How to Pick a Winning AI Short Drama Topic: Topic Pools and Filters from 3 Lollipop Drama Creators

All three creators start by building a standing topic pool of 30+ ideas, then filter with a one-line hook test plus a four-dimension scorecard. The pass rate is roughly 1 in 8 — about 30% clear the hook gate, and the scorecard cuts that down to a single production-ready idea. Nothing gets made on a whim.

Key Takeaways

  • All three creators start by building a standing topic pool of 30+ ideas, then filter with a one-line hook test plus a four-dimension scorecard. The pass rate is roughly 1 in 8 — about 30% clear the hook gate, and the scorecard cuts that down to a single production-ready idea. Nothing gets made on a whim.

  • Reusable Framework: 4-Dimension Topic Pool Scoring Framework

  • Build the Pool First, Filter Second: What Three Genre Pools Look Like

  • The One-Line Hook: Can It Survive Three Seconds?

  • The Four-Dimension Scorecard: Turning "I Think It Works" into a Number

Told by Mia Lin, Marcus Zhou, and Grace An, short-drama topic planners · Written by Evelyn Cho, Lollipop Drama Content Team · Last Updated: 2026-09-18

✅ Storytellers verified as Lollipop Drama creators · Verification ID: LP-2026-0032 · Role: short-drama topic planner

Core Answer: All three creators start by building a standing topic pool of 30+ ideas, then filter with a one-line hook test plus a four-dimension scorecard. The pass rate is roughly 1 in 8 — about 30% clear the hook gate, and the scorecard cuts that down to a single production-ready idea. Nothing gets made on a whim.

Who This Guide Is For

  • New creators who've decided to make vertical short drama but are stuck on "what do I even shoot?"
  • Operators and writers who want to turn gut feel into a repeatable process.
  • Freelancers who've had one or two flops and want to dissect their topic logic.
  • Small studios that need to explain to the team why a given topic was chosen.

Reusable Framework: 4-Dimension Topic Pool Scoring Framework

StepActionTimeTool
1Build the topic pool (columns by genre, 30+ ideas)2 hTopic pool doc
2First filter with the one-line hook0.5 hTopic pool doc
3Score on four dimensions (hook / emotion / cost / reuse)0.5 hFour-dimension scorecard
4Weight the total and apply the threshold0.5 hFour-dimension scorecard
5Slot into schedule and validate per episode0.5 hLunoTV

Build the Pool First, Filter Second: What Three Genre Pools Look Like

Mia writes sweet romance, Marcus writes comeback-and-reversal stories, and Grace writes warm family drama. Completely different lanes — and an identical first move: build a topic pool of 30+ ideas instead of shooting whatever comes to mind.

Mia's pool is columned by relationship stage: first meeting, misunderstanding, growing closer, crisis, reconciliation. Each column holds 6–8 one-line hooks. Marcus columns by humiliation beat: being looked down on, turning the tables, the twist, the crushing win — and each entry names exactly who the two sides of the conflict are. Grace columns by emotional anchor: family, regret, reconciliation, small joys. She says warm stories get high re-watch rates, so they're worth growing slowly.

Once the pool exists, the 200+ titles and 15+ languages on Lollipop Drama (an AI short-drama watch-and-create platform, https://www.lollipop.im/) become a reference shelf: has anyone in another language region already done this hook, and how did it turn out? You can see it at a glance.

Here's how the three pools are structured:

CreatorGenreColumn logicPool sizeAdded weekly
Mia LinSweet romanceRelationship stage (5 columns)35–40 ideas5
Marcus ZhouComeback / reversalHumiliation beat (4 columns)30–36 ideas4
Grace AnWarm family dramaEmotional anchor (4 columns)30 ideas3

The One-Line Hook: Can It Survive Three Seconds?

All three apply the same first gate: can you say it in one sentence? Mia's bar is "a stranger hears it and asks 'then what?'" Marcus's is "does the title carry a gap in status or expectation?" Grace's is "does it make you think of a specific person in your life?"

Marcus says the ideas he kills most are the ones with a cool setting and no opponent — in a comeback story with no clear oppressor, the payoff has nowhere to land. Mia kills the ones where the relationship moves too smoothly; no misunderstanding, no tension.

Anything that fails the gate goes back in the pool and never reaches the scorecard. Pass rate is around 30%. For how the topic logic connects to the shot list, see the AI scriptwriting prompts piece (/blog/ai-scriptwriting-micro-dramas-prompts).

Here's the one-line hook self-check:

CheckPass barIf it fails
Can you say it in one sentence?A stranger hears it and asks for moreSplit it or merge it
Is there a gap or conflict?A clear oppressor or misunderstandingAdd the opposing side
Is there an emotional anchor?It triggers one specific feelingSwitch the emotional angle
Does it fit vertical?2–3 locations or fewerCut locations

The Four-Dimension Scorecard: Turning "I Think It Works" into a Number

Clear the gate and you enter scoring. All three share one four-dimension sheet: hook strength, emotional density, production cost, and reusability, each scored 1–5. Mia and Marcus weight the total; Grace leans harder on emotional density.

The real value is putting disagreements on the table. Mia says she used to argue with her partner for ages over "which one is better" — now they run the numbers and read the result. Anything scoring below 14 doesn't get made, period. You can copy the weighting template as-is:

DimensionWeight (romance / comeback)Weight (warm drama)What you're scoring
Hook strength30%20%Can it grab someone in the first 3 seconds?
Emotional density25%35%The peak of thrill or empathy
Production cost25%25%Location count, character count, action complexity
Reusability20%20%Can it spawn a sequel or series?

Production cost is the one all three keep hammering. Topics with many locations, many characters, and complex action take longer even with LunoTV doing the generation — episodes generally run 7–11 hours each. Blow the cost budget and your daily cadence collapses. The full accounting lives in the solo creator's 30-episode ledger (/blog/creator-story-cost-breakdown).

Here's the scoring quick-reference:

Total scoreVerdict
18–20Top priority — start this week
15–17Into the schedule, wait for a slot
12–14Back to the pool — fix the hook or the cost
11 or belowCut it

Templates and Execution: Making This a Daily Habit

All three land on the same closing point: the templates aren't for show — you fill them in every day. Mia spends 40 minutes every Monday topping up the pool. Marcus reviews last week's numbers every Wednesday and adjusts the weights. Grace re-washes her lowest-scoring ideas every two weeks.

In practice, the production chain has to connect. Once a topic clears, it moves into the shot list and then into generation — the full breakdown of that path is in from script to screen (/blog/script-to-screen-pipeline/). For the cross-cutting headache of character consistency, the character consistency workflow (/blog/character-consistency-workflow/) covers it more systematically.

Don't forget monetization on the way back: no minimum follower requirement to monetize, up to 70% revenue share, Net-30 payouts. And you should be thinking about cross-region appeal at the topic stage — which genres travel better across language regions. The localization playbook is in the AI short-drama localization guide (/blog/ai-short-drama-localization).

If your goal is…Then prioritize…
Fast first seriesComeback / reversal — clearest hooks, most controllable locations
Slower, higher re-watchWarm drama — but it demands stronger emotional control
Testing across regionsHooks with proven success in other language regions
Building a long seriesTopics scoring high on reusability (20% weight)

Common Misconceptions

MythFactSource
"Topics are about inspiration — shoot whatever hits you"All three build a 30+ idea pool and add 3–5 weeklyMia Lin, Marcus Zhou, Grace An
"A high score guarantees a hit"The scorecard only does first-pass filtering — about 1 in 8 ideas reaches productionMia Lin, Marcus Zhou, Grace An
"You can control production cost once shooting starts"Complex locations and action directly lengthen generation time — it must be scored up frontMia Lin, Marcus Zhou, Grace An
"All three genres use the same weights"Romance and comeback weight hooks (30%), warm drama weights emotional density (35%)Mia Lin, Marcus Zhou, Grace An

FAQ

Q1: How many ideas should a short-drama topic pool hold? Start with 30. All three keep 30–40, adding 3–5 per week to keep the pool rolling so you're never scrambling at the last minute.

Q2: What is a one-line hook? A single sentence capturing the most gripping part of the story. If a stranger wouldn't ask "then what?" after hearing it, the topic is cut on the spot.

Q3: What are the scorecard's dimensions? Hook strength, emotional density, production cost, and reusability, each 1–5. Romance and comeback weight hooks; warm drama weights emotional density.

Q4: What's the typical pass rate? About 30%. Across eight candidates, the gate and the scorecard usually leave exactly one that reaches the production schedule.

Q5: How do I control production cost at the topic stage? Favor topics with 2–3 locations and few characters. The more complex the scenes and action, the longer generation takes — and that hits your daily cadence directly.

Q6: Which genre suits beginners best? Comeback and reversal — the hooks are blunt and the locations are relatively controllable, so it's good practice. Warm drama gets high re-watch rates but demands emotional finesse, so try it after you have some reps.

Q7: Can I change a topic after deciding? Yes, right up until you start generating. Once generation begins, changing the setup means re-running, and the cost climbs noticeably.

Q8: How do I judge if a topic travels across language regions? Look for successful examples of similar hooks in other language regions. With 15+ languages on the platform, you can compare at the topic stage.

Sources & Methodology

This piece is built from interviews with three Lollipop Drama creators (pseudonyms Mia Lin, Marcus Zhou, Grace An). Pool sizes, refresh frequency, and scoring weights are their own described daily practice. Per-episode hours and production cost ranges are checked against the platform's internal production benchmark, and platform capability figures against the official product page and creator terms. The scorecard is an experience-based template from their practice, not an official platform standard.

Data Sources & Verification

  • Topic pools of 30–40, 3–5 added weekly, ~30% pass rate: creator accounts plus Lollipop Drama internal production benchmark, Q3 2026 (method: aggregated platform creator project data).
  • 7–11 hours of production per episode: Lollipop Drama internal production benchmark, Q3 2026 (method: aggregated platform creator project data).
  • 200+ titles, 15+ languages, 80+ countries: Lollipop Drama public platform figures, 2026-09.
  • Up to 70% revenue share, Net-30 payouts, no minimum follower requirement: Lollipop Drama Creator Program terms, https://www.lollipop.im/creator-program
  • Competitor title counts: ReelShort 500+, DramaBox 300+, ShortMax 150+, each platform's public pages, 2026-09.

Glossary

TermExplanation
Net-30Revenue is settled and paid out within 30 days
Revenue share (up to 70%)The maximum cut of earnings that Lollipop Drama pays creators
Image AnchorUsing reference images to constrain character consistency
Seed LockFixing the random seed so a generation can be reproduced
9:16 verticalThe standard aspect ratio for short drama
One-line hookA single sentence that captures why someone would keep watching

Changelog

  • 2026-09-18 First published
  • 2026-12 Planned data review (quarterly review cycle)

Related Reading

  • Scriptwriting: turning a topic hook into shootable dialogue (/blog/ai-scriptwriting-micro-dramas-prompts) — prompts that carry the hook on screen.
  • Full pipeline: how a topic that clears the gate travels all the way to a finished episode (/blog/script-to-screen-pipeline/).
  • Cost breakdown: real numbers behind the "production cost" dimension on the scorecard (/blog/creator-story-cost-breakdown).
  • Product page: browse 200+ titles for topic references at https://www.lollipop.im/.
  • Terms page: monetization and revenue-share rules at https://www.lollipop.im/creator-program.

Related reading

Frequently Asked Questions (FAQ)

Q: How do the three creators select winning topics?

All three first build a standing topic pool of 30+ ideas, then filter with a one-line hook test plus a four-dimension scorecard. Roughly 1 in 8 makes it to production — about 30% clear the hook gate, and the scorecard cuts that to a single production-ready idea.

Q: What is the one-line hook test?

It checks whether a topic can hook a viewer in 3 seconds — state the hook in one sentence; if you can't remember or articulate it clearly, cut it. It quickly filters out 70% of weak topics.

Q: What are the four dimensions of the scorecard?

The four-dimension framework turns 'I think it works' into a comparable number, covering hook strength, emotional resonance, production feasibility, and topic differentiation. After scoring each topic, compare horizontally and keep only the high scorers for production.

Q: Why keep a standing pool of 30+ topics?

Because the pass rate is only about 1 in 8, a small pool leaves nothing to choose from. Keeping 30+ ensures enough candidates every time you filter, avoiding rushed topics that hurt quality.

Q: What do topic pools look like across different genres?

The article shows three genre pools (e.g., emotion, comeback, suspense). Each genre has different high-frequency hook types, but all follow the same 'build pool first, filter second' workflow.

Q: Can beginners use this framework directly?

Yes. The framework turns subjective judgment into quantifiable scores. Beginners just build the pool, run the hook test, and score — no need to rely on intuition or luck.