Production system and quality bar

A production system is the set of steps, templates, checklists and tools that turns an idea in your backlog into a published piece of content, the same way every time. A quality bar is the minimum standard every piece must meet before it goes out. Together they let you publish consistently without relying on heroic effort, and they make it possible to hand parts of the work to someone else later, as CRE-401 described when it discussed standard operating procedures. This lesson also asks you to write an AI use policy, drawing on CRE-410. The deliverable is a production system document of two to three pages.

Professional creators are often less talented than people assume and more systematic. They reuse templates, follow checklists, batch work and review results against a standard. You will not build a perfect system in week 4; you will build a first version, run it for a few weeks and improve it at mid-term using your log.

Mapping the workflow

Start by writing down every step from idea to published piece and the first day after publishing. For a typical long video, the stages are idea selection, research, scripting or outlining, pre-production (shot list, props, location, permissions), filming, data management, editing, captions, thumbnail and metadata, review, scheduling, publishing and the first round of community replies. For each stage note the input, the output, the tool used, the typical time from your log and who does it.

Mapping the workflow often reveals obvious improvements. Perhaps you script by typing and then re-type the script into a teleprompter app; perhaps you rename files by hand every time; perhaps you design each thumbnail from scratch. Each of these is a candidate for a template or a small automation. Look first at the longest stage, because a 20 % saving there matters more than a 50 % saving on a ten-minute task.

Scripts and outlines that save editing time

Many editing hours come from footage that was never needed or from searching for the right take. A tighter script or outline reduces both. For educational content, an outline with a strong hook, a clear promise, numbered steps and a payoff usually works better than a word-for-word script, which can sound stiff. Write the opening twenty to thirty seconds word for word, because that is where viewers decide whether to stay, and outline the rest.

Film with editing in mind. Clap or snap at the start of each take to mark it, say the step number aloud before each section, record a few seconds of silence for noise reduction, and film close-ups of key actions twice from different angles. These habits feel unnecessary on set and save large amounts of time at the editing desk.

File and data management

Lost footage is a common and avoidable capstone disaster. Use a consistent folder structure for every project, for example a project folder named with the date and slug, containing folders for footage, audio, graphics, exports and documents. Follow the 3-2-1 backup rule from CINE-320: three copies of important files, on two kinds of storage, with one copy off-site or in the cloud. Name exported files consistently so you can find the final version, and delete or archive raw footage only after the video is published and backed up.

Checklists

A checklist turns your quality bar into steps that cannot be forgotten. Aviation and surgery use checklists not because experts are careless but because everyone forgets things under pressure. Write three short checklists: pre-filming, pre-publish and post-publish. Keep each to one screen and use it every time.

The pre-publish checklist is where compliance lives. Put disclosure, rights clearance, accessibility and the made for kids setting on it so that they are checked every time, not only when you remember. CRE-320, CRE-330 and CRE-450 provide the details; the checklist ensures they are applied.

Defining the quality bar

A quality bar describes 'good enough to publish' in concrete terms. Without one, perfectionists polish forever and others publish carelessly. Write five to eight criteria that every piece must meet, phrased so that you can check them quickly. For example: audio is clear with no background noise that masks speech; every step is visible on screen; information has been checked against at least one reliable source; captions are accurate; the thumbnail is readable at small size; and nothing in the video could reasonably cause harm if followed.

Separate must-have criteria from nice-to-have ones. A slightly imperfect color grade is acceptable; an unsafe repair instruction is not. Audio quality matters more than video quality for most viewers, so a clip-on microphone is usually a better first purchase than a new camera.

Review before publishing

Build a short review step into every piece. Watch the final export from start to finish at normal speed, ideally after a break, with the pre-publish checklist beside you. If possible, ask a friend from the target audience to watch it too and tell you where they got bored or confused. Fresh eyes catch problems you can no longer see. For anything involving safety, health, money or legal claims, check facts against reliable sources and note them in the description where helpful.

An AI use policy

AI tools can save a great deal of time in a capstone: transcription, rough captions, outline brainstorming, noise reduction, thumbnail ideas and translation drafts, as CRE-410 explained. They also create risks: factual errors stated confidently, generic content, copyright and likeness concerns, and disclosure obligations for realistic synthetic media. Write a short AI use policy for your channel that says which tools you use, for what tasks, how a human reviews the output, what you will never use AI for, and how you will disclose AI use.

• Allowed with review: transcription, caption drafts, noise reduction, outline brainstorming, translation drafts, alternative title ideas.
• Allowed with disclosure: realistic synthetic images, voices or scenes, following platform labeling rules.
• Not allowed: inventing facts or quotes, cloning another person's voice or likeness without consent, generating fake reviews or testimonials.
• Always: a human checks facts, tone and accuracy before publishing, and keeps a record of significant AI use.

At the time of writing (October 2026), major platforms require creators to label realistic altered or synthetic content, and Canadian and US advertising regulators expect AI-generated endorsements or virtual influencers to be disclosed, as CRE-410 explained. Check current platform rules before publishing any realistic synthetic content, because they change often.

Writing the system document

Bring everything together in your production system document: the workflow map with times, folder and backup structure, the three checklists, the quality bar, the review step, the AI use policy and a short section on what you plan to improve after four weeks of use. Keep it practical. If someone helped you next term, could they follow it? That is the test of a good system.