Freelance, agency, or in-house: choosing a video editing setup

The model usually picks itself once you write down two numbers: how many finished pieces you need in a normal month, and how far that number swings between your busiest month and your quietest. A brand shipping four to six social cutdowns a month with a spike at each launch has a freelance problem. A sports team shipping thirty pieces a month on a fixed calendar, from footage that cannot leave the building, has an in-house problem. Most companies sit between those two and lose a quarter picking a side.

What decides whether the choice was right is almost never the rate. It is what happens in month seven, when the editor who knows your loudness target, your caption font, and how your founder likes to be cut goes quiet for three weeks or takes a staff job. Continuity is the expensive line item, and it never reaches the spreadsheet.

Treat the three options as risk structures, not quality tiers. There are exceptional freelancers and average in-house editors, and the reverse is just as common. What differs is how cost behaves in a slow month and who absorbs the problem when someone disappears.

The five variables that decide it

Score yourself on five things before you look at anyone's rate. All five describe your operation rather than any editor, which is why the answer rarely changes when a better candidate turns up.

• Monthly volume, counted in deliverables. Count finished, approved pieces, not shoot days. Fifteen 9:16 cutdowns pulled from one master is a very different week from three 90-second brand films carrying three revision rounds each.
• Predictability of that volume. A steady twelve pieces a month supports a retainer with a floor. Twelve one month and two the next does not, and pretending otherwise is how retainers get cancelled in month four.
• Required turnaround. Weekly delivery suits a good freelancer. Recaps due by 9 a.m. the next morning, or highlight cuts due two hours after a final whistle, need a calendar you control or a company that staffs a shift.
• Sensitivity of the material. Unreleased product, pre-earnings financials, patient footage, talent under NDA, and adult content narrow the field to people who will sign specific terms and work only on your storage. For adult projects, starting from the adult work listings beats trying to convert a corporate editor.
• Total budget, including what never reaches an invoice. Storage, software seats, review tooling, a hardware refresh cycle, and your own coordination hours belong in the same number. Compare an hourly rate to a salary alone and you will be confidently wrong.

The pattern is not subtle. High volume, high predictability, tight turnaround, and sensitive material point at in-house. Low or spiky volume with comfortable deadlines points at freelance. High volume with wide format variety and no internal appetite for managing creative work points at an agency.

How the three models compare

Here is the short version. Read the continuity row twice, because it gets waved away during the decision and then dominates the outcome.

Two rows need a caveat. The freelance ceiling is one person's ceiling, so a specialist documentary editor beats a general agency team on a documentary and loses badly on a six-format social package. Low agency continuity risk holds at company level only: your named editor can roll off to another account unless the contract names them.

What a freelancer costs versus a salary

A freelance day rate and a salary are not comparable numbers, and treating them as comparable is the most common mistake in this decision. Base pay is where employee cost starts, not where it lands.

Employer payroll taxes, benefits, and paid time off sit on top of base pay. Then add what a freelancer already owns and you would now buy outright: a workstation on a three or four year refresh cycle, a calibrated monitor, an editing seat in Premiere Pro or Resolve Studio, the plugin set the work needs, and working storage plus an archive that grows every month once you are shooting 4K. A freelancer folds that into a rate and replaces it on their own schedule.

Then subtract the time that is not editing. An editor's week absorbs ingest, transcode, file naming, asset hunting, meetings, and revision rounds, so plan on three to four productive editing days out of five, and quiet weeks between campaigns are paid weeks. Someone also writes the briefs, runs the reviews, keeps the person busy in slow periods, and holds a career conversation twice a year. That someone is you.

None of this argues against hiring. It argues against setting an hourly figure beside a salary figure and calling that a decision. For figures on both sides of the comparison, what editing work costs a buyer is the piece to read next.

What the agency markup buys

An agency's blended rate sits above the same editor's freelance rate. You are paying for four things a single freelancer structurally cannot sell you.

1. 1Account management. A producer builds the schedule, chases the review round that has sat untouched for four days, checks deliverables against platform specs, and warns you when your feedback will blow the date.
2. 2Redundancy. If the assigned editor gets sick on the Wednesday before a Friday launch, somebody else opens the project, which works only if the project was built to be opened by somebody else.
3. 3Range under one roof. Offline edit, motion design, color, sound mix, subtitling, and versioning across 16:9, 9:16, and 1:1 without brokering four relationships.
4. 4Turnover absorbed elsewhere. When an editor quits, the recruiting cost and the gap come out of their margin, not your launch date.

The markup does not buy the same hands every time. Ask who is actually cutting, ask for a named lead and a named backup in the statement of work, and ask what notice you get if the lead rolls off. A vague answer is information, so price it in.

The hybrid most growing teams land on

The setup that survives contact with reality is one anchor plus a bench. The anchor is a freelancer on a monthly retainer with a floor: days or deliverables you pay for whether or not you use them, in exchange for priority in their calendar and a committed turnaround. The bench is two or three editors you have already briefed and paid.

Write the retainer terms properly the first time: the floor in days or deliverables, the first-cut turnaround, revision rounds included before hourly billing starts, whether unused days roll over (cap that at one month or they pile into a debt nobody can schedule), the rush multiplier under 48 hours, and a 30 day notice period both ways. A reusable brief format does more for turnaround than any clause in that retainer, since most delay comes from ambiguous instructions rather than slow editing.

A bench only works if it stays warm. Give each bench editor one small paid job a quarter so their access still works and your folder structure is still in their head. Keep them on the anchor's project template, naming convention, and drive layout, share your calendar a month out, and pay invoices fast. An editor ignored for eight months, then messaged Friday afternoon about a Monday delivery, will say no.

Restocking the bench is ongoing work rather than an emergency task. Note the good editors you meet on projects you do not book, with their software and usual turnaround. Filtering a directory of editors by software and location builds a shortlist faster than writing a job ad.

Continuity and the handover pack

Whichever model you pick, the failure mode is identical: the person holding the project files stops answering, and six months of brand work turns out to live on a drive you have never seen. Require a handover pack from day one, in the contract, delivered on a schedule rather than at the exit.

• Native project files on a schedule: .prproj, a Resolve .drp or project archive, .fcpbundle, Media Composer bins, or .aep with linked assets.
• A consolidated archive with media attached, built through Project Manager in Premiere Pro or Media Management in Resolve, so relinking works on another machine.
• Original camera and audio media on your storage in the format it was shot, not re-encoded H.264 proxies.
• Editable graphics: After Effects compositions with fonts listed, .mogrt templates, titles, and lower thirds.
• LUTs, grade stills or PowerGrades, and export presets matching the specs you publish to.
• Music and stock licenses in your company's name with license IDs, never in an editor's personal account.
• A one-page production note: folder structure, naming convention, frame rate and codec per deliverable, and caption workflow.
• Shared drives, review tools, and asset libraries owned by your account, with the editor added as a member.

Collect it quarterly, not on the last day, and expect the weak point to differ by model. A departing freelancer is the worst person to rely on for a careful export. An in-house editor's local scratch drive is the biggest single point of failure on most small teams. An agency returns flattened exports and keeps the projects unless the master services agreement lists this pack beside the work-for-hire language.

Resistance here is diagnostic. Someone who explains their archive process in two calm sentences has done it before. Someone who treats the request as an insult belongs with the other warning signs worth screening for before you sign anything.

Signals it is time to switch

Models expire. The setup that fit at six pieces a month will not fit at twenty five, and an in-house hire that made sense during a campaign push stops making sense when the push ends.

Move up, toward a retainer, an agency, or a hire, when your anchor starts declining briefs on schedule grounds, when you spend more than a few hours a week coordinating editors, when three people touch the same brand and the output drifts apart (caption placement, loudness, safe margins), when you need same-day turnaround more than once a week, or when legal decides footage cannot leave your systems.

Move down, toward freelance or hybrid, when your editor is idle or doing non-editing work for much of the month, when volume drops after a campaign and does not recover, or when you keep booking outside specialists anyway because one generalist cannot cover motion design, color, and long-form. Two of those at once means you are paying fixed cost for variable work.

Going back to freelance after an in-house hire has not worked out is a correction, not a confession. Before reposting the same job, check whether the role was defined wrong: a team needing a motion designer three days a week and a documentary editor two days a week does not need one video editor; it needs two specialists on retainer.

Agency relationships expire more quietly. When the account team has churned twice, when the work has become repetitive template versioning a freelancer could handle for less, or when you are paying producer time to manage a workload you understand better than they do, move the recurring work to a retainer and keep the agency for pieces that need range. Nobody at the agency will raise this for you.

Start with three numbers, then a paid test

Open your last three months and write down three figures: finished pieces shipped, pieces you wanted to ship and did not, and hours you spent coordinating the work. Those three settle the argument faster than any comparison table, including the one above.

Then test the model rather than only the person. Run one genuine paid deliverable with a real deadline, a real brief, and two rounds of feedback, and watch where the friction lands. Friction on the editor's capacity is a volume problem and switching models fixes it. Friction on your own review turnaround is a process problem and hiring anyone will not fix it. The full hiring sequence covers the brief, the test, and the contract in order. Whichever way the test points, put the handover terms in that first contract: the setup you choose this quarter is not the one you will be running in two years.