Tutorial Ultra-Premium Cinematic System for Fintech Neobank | increase watch time 32% | PID 01009

EXECUTION BRIEF — TUTORIAL ROLE Act as a senior commercial video director, motion designer, editor, and prompt engineer. Your job is to produce an executable shot plan and generation brief, not a generic “premium cinematic” description. Optimize for a believable finished video that an editor can assemble and a video model can reproduce consistently. PROJECT Industry: Fintech Neobank Audience: a mobile-first customer who wants faster everyday banking without sacrificing clarity or trust Primary problem: routine money tasks feel fragmented, slow, or difficult to understand Creative goal: increase watch time 32% Goal interpretation: progressive reveal Business objective: Hold attention by revealing useful information in stages. Avoid arbitrary speed ramps, constant camera motion, or delayed relevance. Category: Tutorial Target duration: 45–75 seconds, designed as modular 8–15 second scenes Format: 16:9 primary; keep center-safe composition for later crops CREATIVE CONCEPT Center the video on one banking action such as moving money, freezing a card, or checking a transaction. The viewer should understand the situation before the first third of the runtime. show the task completing cleanly with believable states; use supplied UI for exact product screens The creative should feel architectural interior, using large diffused key with subtle rim. Use a 85mm portrait visual language and gentle lateral track. Keep the camera movement subordinate to the message: motion should reveal, compare, demonstrate, or create emphasis—not exist merely to look expensive. CATEGORY EXECUTION Hook the task; establish starting state; perform 3–5 numbered actions; show completed state; end with one verification cue. Treat each step as its own shot. Maintain the same person, device, environment, wardrobe, and interface state across adjacent shots. Keep continuity locked across shots: same subject identity, wardrobe/materials, environment, light direction, and product geometry. SHOT-BY-SHOT PLAN 1. OPENING — Establish the subject and problem immediately. Specify the starting frame, subject position, action, and what changes during the first 1–2 seconds. No logo-only opening. 2. DEVELOPMENT — Show the single core action. Give each shot one job. If the story needs multiple steps, cut only when the information changes. 3. PROOF — Show an observable product behavior, physical detail, customer moment, or supplied evidence that supports the claim. Never substitute abstract “success” imagery for actual proof. 4. PAYOFF — Resolve the action with a clear end state. Leave enough clean screen time for approved copy or CTA when the format requires it. 5. END/LOOP — For looping formats, design the last movement to return naturally to the opening state. For non-looping formats, end on the strongest useful frame rather than a decorative logo animation. CONTINUITY RULES Keep the main subject consistent in shape, color, material, identity, wardrobe, hand position, and environment. Keep lighting direction and focal length stable unless a deliberate shot change is specified. Do not morph between unrelated objects. Do not change the product design from shot to shot. For exact interface or packaging text, use supplied assets and add typography in post. Do not ask the generation model to invent legible small text. AUDIO / EDIT Clear voiceover or on-screen step labels; use quiet transitions rather than constant music. Edit on action or information change. Avoid excessive whip pans, speed ramps, glitch transitions, or artificial “epic” impacts. Preserve enough natural pauses for comprehension. Captions, legal lines, product claims, and CTA copy should be added from approved text during post-production. CLAIM AND ASSET DISCIPLINE Do not fabricate balances, transaction history, regulatory claims, or financial performance. Only use product facts, customer quotes, statistics, certifications, medical/financial claims, property details, or performance claims that are supplied and approved. If a required fact is missing, mark it as [APPROVED INPUT REQUIRED] rather than inventing it. NEGATIVE / FAILURE CONDITIONS No flicker, temporal instability, object morphing, identity drift, warped hands, duplicate objects, impossible reflections, fake UI text, accidental logos, watermarks, unreadable captions, random camera shake, excessive motion blur, plastic skin, generic stock-footage look, or visual elements unrelated to the core problem. Do not add a second product, second narrative, or second CTA. OUTPUT SPECIFICATION Return: • a concise concept statement; • the ordered shot list with approximate seconds, framing, camera action, subject action, environment, lighting, and transition; • the exact generation prompt for each shot; • the negative prompt shared by the shots; • post-production notes for supplied UI, captions, CTA, color, sound, and export; • one fallback shot for any scene that is likely to fail in generation. QUALITY GATE A viewer should be able to name the task, follow the steps in order, and identify the final successful state without guessing. Also verify: - The first frame makes the subject and context identifiable. - Every shot has one clear purpose. - The core product/subject remains consistent. - No unsupported claims are introduced. - The requested aspect ratio and duration are achievable without relying on invisible or contradictory instructions. - The result could be handed directly to a director/editor without asking what “premium” means. - The creative goal is treated as an optimization target, not a guaranteed outcome. SUCCESS TEST A professional reviewer should be able to answer “Who is this for?”, “What is happening?”, “Why should I care?”, and “What should I do next?” from the finished cut. Hold attention by revealing useful information in stages. Avoid arbitrary speed ramps, constant camera motion, or delayed relevance.