China Video Supplier Scorecard: Pilot or Full Project?

Hiring Video Production in China 10/10

Two video suppliers in China both show relevant work and both appear to understand the brief. One proposes a low-cost test shoot; the other recommends moving directly into full pre-production. An overseas buyer may frame this as caution versus commitment: a pilot feels safer, while a full award feels faster.

The better question is not how little to spend first. It is whether the most important unproven risk can be isolated and tested.

If the uncertainty is a small set of demanding shots—controlling reflections on a mirror-finish product, repeating a prototype function, or obtaining a real-time decision from a remote brand team—a paid pilot may produce useful evidence. If the uncertainty sits upstream—factory access, sample readiness, approved claims, or global usage rights—a polished test clip will not solve it. A paid recce or pre-production package is usually more valuable. When the critical risks are already evidenced and the essential gates have passed, another pilot may only duplicate cost and schedule.

A supplier scorecard should therefore do more than assign one total score. It should separate non-negotiable gates from project-specific evidence, then direct the buyer to the smallest verification step that can change the award decision.

Start with pass-fail gates, not weighted scores

Some weaknesses cannot be offset by a strong creative proposal or a lower price. New Zealand Government Procurement's current guidance distinguishes pass-fail preconditions from qualitative evaluation of capability, method, risk and value. It is not a mandatory framework for commercial video, but the separation is useful for cross-border production buying.

These gates do not ask the production supplier to make decisions for the buyer or the factory. A credible supplier explains which decisions require direct brand authority, which production can coordinate and which only the factory or rights holder can confirm. Guide 01 maps that division in the factory-shoot delegation framework. Before money moves, guide 09 adds the entity, beneficiary, scope and acceptance checks.

If a critical gate fails, “shoot something small first” is not the remedy. The missing authority, document or owner must be resolved. Otherwise, the pilot can prove that a crew creates a certain image without proving that the result can be used, released or delivered on schedule.

Score evidence against this project's failure modes

There is no universal weighting for every production. A connected-device demo, an industrial corporate film, an international-cast lifestyle shoot and a multi-market social asset package fail in different ways. The dimensions can remain consistent; their importance must follow the assignment.

Add an evidence-location field beside every rating: which document, case section, team member or test supports the judgment? That produces four more useful states than a bare number: evidenced; plausible but still unverified; promise only; or in conflict with the brief.

Cost comparison must also use the same input. The AICP Bidding Reference Guide, updated in February 2026, asks whether budgets include social, stills, behind-the-scenes, post, talent, VFX, music and finishing. It also asks about filming restrictions, deliverables, remote video streaming and the point at which the job is fully or initially awarded. The implication is straightforward: a lower total may reflect a different scope, not better efficiency. Use the quote-leveling method in guide 02 before scoring price.

Route the award: pilot, paid pre-production or full project

Once the remaining unknowns are visible, ask three questions about each one:

1. What is the consequence of being wrong? Would the loss be one optional shot, or could it invalidate a shoot day, worldwide use or the launch date?
2. Can the risk be isolated? Can one to three shots or one defined task test it without recreating the entire project?
3. Can the result be reused? Will the recce notes, technical settings, media or decision carry into the formal production?

The answers lead to three practical routes.

Route A: award the full project

Choose a full award when all critical gates pass and the highest risks already have credible evidence. Factory access and production dates are confirmed. Final samples are available. The supplier can identify its role in relevant work. The actual team, availability, cost boundaries and approval route are known. The buyer is not relying on an untested visual technology or a sample that behaves unpredictably.

A full award does not mean skipping pre-production. Script, storyboard, sample testing, recce, call sheet and staged approvals remain necessary. It simply avoids separating mandatory preparation into a “pilot” that creates no additional decision value.

Route B: buy a recce or pre-production package first

This route is appropriate when the important unknowns are upstream and cannot be answered by the beauty of a finished shot:

• the factory has not confirmed areas, processes or people that can be filmed;
• product appearance, firmware, app screens or packaging are still changing;
• the buyer has not frozen approved claims, target markets or version requirements;
• locations, cast, travel and shoot days cannot be estimated responsibly until site information exists;
• the main risk is cross-functional approval and scope, rather than camera technique.

The package can deliver recce images and routes, a red-amber-green shot map, sample test record, risk and gap log, preliminary resource plan, refined estimate and recommended shoot-day structure. Its completion standard is not “looks promising.” It is whether the buyer can now decide to shoot, change scope or move the date.

Route C: commission a bounded paid pilot

A pilot is useful when a high-consequence risk can be separated and the test conditions are representative. Examples include:

• whether a reflective, transparent or illuminated product can achieve the required material look in the target environment;
• whether an engineering sample can repeat one key function and let the camera prove it clearly;
• whether a specific live-action and generative or VFX combination preserves product structure and movement continuity;
• whether headquarters can approve a critical shot within the available bandwidth, proxy workflow and time-zone window;
• whether a proposed performer can complete a product operation with unusually high handling demands.

The pilot does not need to recreate the full production, but it must preserve the condition that drives the decision. A tabletop lighting test cannot prove factory-access coordination. A product-only close-up cannot prove international-cast rights, multilingual delivery and paid-media version control.

Write an eight-part pilot task before work begins

A paid pilot is not “shoot a little and see.” The buyer and supplier should define at least:

1. the single risk being tested and the questions explicitly outside the pilot;
2. the product, firmware, interface, location, performer and reference versions;
3. the one to three required shots and permitted substitutions;
4. the actual team, critical equipment, working time, location and travel assumptions;
5. when the buyer reviews, who can approve, and what constitutes pass, adjust or stop;
6. whether delivery includes test notes, camera originals, graded samples, usable finals, or a combination;
7. ownership and permitted use of concepts, footage, performers, music and pilot outputs;
8. whether pilot fees credit toward the main award and how long the result remains valid for a product or location version.

Acceptance must match the original risk. For a reflective enclosure, “premium look” is too vague. A useful criterion might state that the logo, ports and material boundaries remain visible; moving reflections do not hide the function; approved aspect ratios retain usable composition; and the lighting footprint, control surfaces and time required for the full shoot are recorded.

The 4As' public guidance on project reviews says short-term assignments should not request spec work. Its idea-ownership guidance adds that when a large or complex review genuinely requires speculative work, the client should consider a fee and establish the scope, deliverables, measurement and ownership in advance. These documents concern agency reviews, not Chinese video-production law. The transferable lesson is narrower: paying for the test compensates the work; it does not, by itself, settle usage or ownership.

When a cheap pilot gives the wrong answer

A pilot is a poor supplier test when:

• The real challenge is system coordination. One polished shot cannot show whether factory access, samples, performers, remote approval, language versions and asset delivery work together.
• The conditions are not representative. A different prototype, temporary location or substitute performer makes success difficult to generalise.
• Start-up cost already resembles full pre-production. International cast holds, intercity travel, special locations or large equipment packages make the “small” test materially large.
• There is no time to learn. If the schedule leaves no room to adjust the script, technology or plan after the test, the pilot only adds an approval step.
• The buyer is actually collecting free production-ready ideas. Asking several suppliers to make usable concepts, then handing the preferred idea to the lowest bidder, neither tests collaboration fairly nor creates a clean rights chain.

An expensive pilot is not automatically useful either. It must reduce one real uncertainty and leave a record that can be carried into production. Otherwise, it merely divides one project into two bills.

The scorecard should end in a procurement action

Do not finish the review with a ranking alone. Each candidate should lead to one executable outcome:

• Award the full project: gates pass, major risks are evidenced, and team and scope can be locked.
• Award a paid pre-production package: access, sample, location or scope remains unknown and must become shootable and estimable.
• Award a bounded paid pilot: one critical risk is isolatable, conditions are representative and the output is reusable.
• Do not proceed yet: a critical gate fails, or the supplier will not expose assumptions, actual team, rights or recovery paths.

The right production partner in China is not necessarily the lowest bidder, the largest company or the supplier with the most dazzling deck. It is the team that puts each uncertainty in the correct stage and turns each paid step into a clearer decision.

For corporate, product or factory content, ONCE's corporate and factory video production solution outlines the public production scope from pre-production through the shoot and post. To decide whether your project needs a pilot, a recce or a full award, send the product and factory location, target market, intended use, schedule, sample status, current brief, candidate estimates and available materials. ONCE can first map the production scope and verification route; this is not a promise of sales, certification or legal outcome.

Sources checked 7 October 2026. The three-layer scorecard and three-route decision are ONCE's production-management synthesis, not a universal certification, legal opinion or proven commercial-outcome model.