How to Manage Video Production Projects From Brief to Final Delivery

Video projects die in revision rounds. The shop with the contract that says "2 rounds included, $500 each after" is the shop that ships on budget. The shop that takes "a few small tweaks" until the colorist gives up is the shop watching margin evaporate on a 60-second spot.

After eight years producing branded video — corporate explainers, e-comm product shoots, founder interviews, event recap reels — the pattern is the same. Pre-production gets rushed because the client signed late. Production day a memory card slot fails and nobody notices until wrap. Post-production turns into seven revision rounds because round one had the wrong music and nobody asked the CMO until round four. Delivery means cutting fourteen aspect ratios for paid social and somehow nobody has the brand kit.

Video production project management is the discipline that turns those moving parts — crew scheduling, gear logistics, 4K footage storage, client review rounds, music licensing, deliverable variants for social and full-length — into a process that ships predictable work at a predictable margin. This is the five-step process that's worked across roughly 200 client projects, with the contractual and operational guardrails that keep revisions from eating you alive.

Nothing happens until three things exist in writing: a creative brief, a deliverables list, and a paid deposit.

The creative brief is one page. Audience, single message, tone references (link 2-3 videos), brand guidelines, success metric. If the client can't say "the goal of this video is X for audience Y," you're not ready to shoot. Send the brief, get it signed, and stop work until it comes back.

The deliverables list is the contract. Be specific to the point of pedantry. "One 60-second hero video, 1920x1080, H.264, color graded, with two music license options" is a deliverable. "A great brand video" is a lawsuit. List every cut: hero, three 15-second social cuts, vertical 9:16 for Reels and TikTok, square 1:1 for feed, with-captions versions, without-captions versions. Each variant is a line item with a price.

Deposit is 50% before pre-production starts. For new clients, 50/50. For repeat clients with NET-30 history, 30/70 with the balance due before final delivery — never on receipt of files, because once they have the files your leverage is gone. Project kickoff happens after the deposit clears. Not after the PO is issued. After the deposit clears. This single rule has saved more projects than any tool.

Pre-production is where 70% of a smooth shoot is built. Skipping it to "save time" costs you a full day of fixes in post.

Location scout happens at the same time of day as the shoot. The window light at 9 a.m. is not the window light at 3 p.m. Walk the location with the DP, check power, ceiling height, sound bleed, restroom access, and where the talent will hold between takes. Negotiate a location release on the spot, not the morning of.

Talent and crew booking. Lock the DP, gaffer, sound op, and any second camera at least two weeks out. Confirm in writing with rates, hours, overtime trigger, and kit fee. Talent gets a separate release covering usage scope (paid social, web, broadcast — be specific), term (perpetual or N-year), and territory.

Gear list. The DP owns this, but you sign off. Camera body and backup, lenses, sticks, slider or gimbal if needed, key/fill/back lights, sound (lav per speaker plus boom backup), and at minimum two memory cards per camera with a backup card reader on set. Rentals get reserved in writing with pickup and return windows.

Shot list. One row per shot with location, talent, lens, framing, audio source, and which deliverable it serves. The shot list is what keeps a 14-hour day from turning into a 17-hour day. Print it. Tape it to the slate.

Production day is execution, not creative discovery. The creative was decided in pre-pro.

Call sheet goes out the night before by 6 p.m. Crew names, call times, location address with parking notes, talent call times, schedule by hour, contact numbers, weather, and the producer's emergency line. Anyone not on the call sheet is not on set. Anyone who shows up at the wrong time gets paid for the actual call time, not when they got there.

Talent and location releases get signed before the first take. Not at lunch. Not at wrap. Before the slate. A signed release at the start of the day prevents the "actually I'm not comfortable with that usage" conversation at 6 p.m. when you've shot the day's coverage.

Footage backup is the single most important thing that happens on a production day, and it happens before wrap. The 3-2-1 rule for media: three copies, on two different media, with one off-site. On set that means card-to-laptop SSD, card-to-shuttle drive, and one of those drives goes home with a different person than the other. Cards do not get formatted until two verified copies exist. Producers who format cards on the way home from the shoot eventually format the wrong card.

Post-production is where contracts get tested. The deal you wrote in step one is what protects you here.

The standard is rough cut plus two revision rounds. Rough cut goes to the client at roughly 70% — picture locked, scratch music, no color, no final mix. The point of the rough cut is to confirm narrative, pacing, and music direction before you spend ten hours on a final color grade for a video the CMO is going to ask you to re-edit. Two revision rounds after the rough cut: round one for picture changes, round two for fine tuning. Anything beyond round two is billed at $500 per round, in writing, agreed at contract.

Scope creep guardrails are written into the SOW. "New shots not in the original shot list trigger a change order." "Reshoots are billed at day rate plus crew." "Adding deliverables not on the original list is billed per variant." When the client asks for a vertical cut on round three, you don't argue — you send a one-line change order with a price and a new delivery date. Most clients will say yes. Some will reconsider whether they really need it. Either outcome is fine.

Music licensing is a place producers get sloppy and sued. Premium Beat and Artlist are the two services most working video shops actually use — single-track or subscription licenses with clear commercial usage rights. Read the license. "Web only" means web only. Broadcast and paid social often require separate licenses. If the client wants a known song, they pay for sync rights — those negotiations take weeks and cost more than the rest of the project, so flag it in the brief.

Delivery is not "send a Dropbox link." Delivery is a structured handoff that the client's team can use six months from now without calling you.

Final files go to the client's cloud — typically a shared Google Drive or Dropbox folder organized by deliverable type. Master files in ProRes 422 HQ for archive, H.264 1080p and 4K for web, plus every social cut as a separate file with naming convention "client_project_aspect_duration_version" — for example, "acme_summer_9x16_15s_v1.mp4." Captions get delivered as both burned-in and SRT sidecars.

Social cuts are their own deliverable list. A 60-second hero typically generates a 30-second cut, three 15-second cuts, vertical 9:16 versions of each, square 1:1 versions, and captioned versions of all of the above. That's potentially 24 files from one hero. Build the variant list in pre-production so you shoot for it.

Archive policy gets agreed in the contract. Project files (Premiere or Resolve project, raw footage, audio stems) live on your shop's archive drive for 12 months, after which they're deleted unless the client pays an annual archive fee. RAW footage is heavy — a one-day shoot can be 800 GB — and clients who say "keep it forever" rarely want to pay for it forever. Get the policy in writing so the conversation in month 13 isn't a surprise.

The final piece of delivery is the project debrief. What went well, what cost more time than expected, what to budget differently next time. The shops that do this consistently bid more accurately on the next project. The shops that don't keep underestimating post.