Video Editing Subscription vs Production Crew: Audit First

Executive Summary

A video editing subscription is a capacity purchase for post-production: the provider edits material the buyer supplies, or animates from approved assets. It does not create footage that was never recorded. This report, written for e-commerce and SaaS marketing leads, founders and agency producers, argues that the choice between a subscription and a production crew should be made scene by scene, before any plan is priced. It covers service subscriptions only, because "video editing subscription" also matches software plans in public search.

Provider documents show why the audit comes first. One comparison guide answers whether a buyer must supply footage this way: "For most editing subscriptions, yes: they edit what you send" ^[1]. Design Pickle's scope article states "All media files and assets must be provided by the customer." ^[2] Kimp's plan page lists editing of up to one hour and short-form video of up to three minutes ^[3], while its FAQ limits supplied video to a combined 15GB and the trial to one video of up to 15 seconds. Limits therefore depend on request type and on trial or subscription status. Some providers do sell shoots: Superside describes video as "Script, cast, shoot, edit, and animate." ^[4] Kimp lists Video Shoots among requests it does not fulfil, and beCreatives says: "No, we only provide awesome editing services! We can add Stock videos from our libraries and use the footage you provide." ^[5]

Production is a separate purchase with its own labor market. The U.S. Bureau of Labor Statistics (BLS) reports a median annual wage of $75,420 for film and video editors and $74,990 for camera operators in May 2025 ^[6]. Wistia's guide separates planning, shooting and post-production and says of the shoot: "Whatever you do now will impact your workload in post-production" ^[7]. Authenticity adds a regulatory layer: the U.S. Federal Trade Commission (FTC) rule on reviews and testimonials prohibits writing, creating or selling a review or testimonial that materially misrepresents that a testimonialist exists ^[8].

The recommended method has three steps. First, list every planned scene and mark the evidence it must carry: a real customer, the real product, or an idea that type and 2D motion can express. Second, record source, audio, permissions, screen-capture freshness, brand assets and the missing shot. Third, route the job to editing, motion design, new production, or a sequence of those. The correct answer is sometimes to delay the subscription and commission a shoot first. Budget with total committed spend, using the reader's own quotes; this report publishes no industry-average price. Keyword demand is specialist: the publisher's own Google Ads estimates (US, via DataForSEO; not independently verifiable) averaged 260 searches a month, distorted by a January 2026 spike of 2,400, and show no growth.

Introduction and Background

Marketing leads commissioning recurring product demos, customer testimonials and short-form video face a procurement question that price lists do not answer: which service should be bought first? Flat-rate subscriptions sell editing capacity. Production crews sell capture. Motion designers sell animation built from assets that already exist. Buying the wrong one first produces idle capacity, or a queue of requests that cannot start.

Search demand reflects a niche. Google Ads keyword estimates for the United States in English, obtained through DataForSEO on 27 October 2026 (history October 2025 to September 2026), show "video editing subscription" averaging 260 searches a month, driven by a January 2026 figure of 2,400, with July, August and September at 10, 20 and 20. These figures are the publisher's own tool output and are not linked to a public dataset; they are advertising estimates, not organic clicks, and they are not read as market growth here.

Industry surveys describe a mixed picture of in-house and outsourced work. Wistia's 2026 State of Video report says "we surveyed 900+ professionals across industries" ^[9], and reports that around a quarter of companies also outsource video creation ^[9], with outsourced production seeing the bigger jump, at 7% ^[9]. Wyzowl's 2026 survey reports that 91% of businesses use video as a marketing tool ^[10], from 266 respondents in late 2025 ^[10]. Both publishers sell video products or production, so their figures are best read as vendor-run surveys.

Existing competitor coverage already compares plans. EditClub's guide, dated 16 October 2026, notes that "Some cap the number of active requests, some sell editor hours, some sell credits." ^[1] and also asks whether the buyer must supply footage, but it is published by a service it ranks: "EditClub (our service, judged on the same criteria as everyone else below)" ^[1]. Mark Studios' price database reports 102 figures from 35 publishers ^[11], says none of them is a quote ^[11], and discloses that its own prices are excluded from the sample ^[11]. Both are treated here as competitor coverage, not independent benchmarks, and their price medians are not repeated.

The publisher of this report, Tapflare, sells a flat-rate monthly design and website development subscription, run through a dedicated project manager, that the publisher describes as covering motion graphics and animation requests; video editing scope should be confirmed in writing. It is a direct provider in the category discussed and appears in the comparison table on that basis. It is not a production crew, campaign strategy, copywriting or media-buying service, and it is not suited to a single one-off job. The rest of this report applies one test to every route, including that one: what must exist before the work can begin?

Route 1: Subscription Video Editing

Capabilities

Editing subscriptions work on supplied files. Kimp's FAQ limits the combined size of supplied video to 15GB ^[12] and states that editing and audiograms can run up to one hour, with an SRT (SubRip caption) file required for audiograms ^[12]. Kimp's trial is narrower than its subscription: one video of up to 15 seconds, with unlimited revisions during the 7-day trial ^[12]. Design Pickle publishes a 20 GB limit for short-form videos sent by file link ^[2] and states that the delivered file has a maximum duration of 10 or 30 minutes even when longer videos are submitted ^[2]. beCreatives caps raw footage per order at 50GB on its Scale plan and 100GB on its Pro plan ^[13]. Flocksy accepts uploads to the brief or through links ^[14] and states no size cap on that page.

Reviewing raw material is itself work. Design Pickle says that providing time stamps or pre-selected clips speeds final delivery ^[2]. The principle applies to any provider: the more selection the editor must do, the more of the allotment is spent before the first cut exists.

Adoption

Outsourcing is common but not dominant in the survey data: around a quarter of companies also outsource video creation. A buyer should compare an editing subscription with an in-house hire only on the same scope. BLS reports a median annual wage of $75,420 for film and video editors in May 2025; that is wage data and excludes software, management time and benefits, so it is not a like-for-like substitute for a subscription fee. BLS also notes that most videographers edit their own material ^[6], which matters when one hire is expected to shoot and cut.

Strengths and Limitations

The main strength is predictable recurring capacity when footage keeps arriving. The main limits are the supply of footage and the capacity model. EditClub observes that every provider limits capacity in some way ^[1]. Design Pickle's help center says its creatives work on one request at a time unless more subscriptions are added ^[15], and beCreatives states that it works on business days, one request at a time ^[13]. Flocksy instead meters work in hours: "Each business day, your dedicated team uses your available Daily Hours to move projects forward in the order you set." ^[14]

Handoff terms differ as well. Flocksy states working files and source delivery on its pricing page ^[16]. beCreatives says: "Under our SCALE plan, we are unfortunately unable to share our project files." ^[17] That is a plan-level term worth confirming before a team plans to re-edit in-house. Revision language also varies: Flocksy states unlimited revisions while the subscription is active ^[14], and Superside includes two rounds of revisions per project ^[18]. None of these terms makes a subscription suitable for a scene that has no recording behind it.

Route 2: Motion Design From Approved Assets

Capabilities

Motion design differs from editing because it can start without camera footage. Kimp's FAQ says it will design and animate from scratch ^[12]), and Superside defines brand-led animation as bringing existing brand assets to life through motion ^[19]. The word "approved" matters: animation is only as reliable as the script, storyboard, brand files, product images and licensed audio supplied. Penji's pricing page separates presentation and everyday design from video and motion. It lists Essentials at $995 a month ^[20], marks short-form video and motion as locked on that tier, and includes it on the $1,995-a-month Agency Partner tier ^[20]. The same table marks long-form video editing as limited on Agency Partner and included on Dedicated Team (^[20].

Adoption

No independent survey reviewed here isolates motion graphics subscriptions from video overall, so adoption claims are avoided; the publisher's own keyword estimates (170 a month on average for "motion graphics services") indicate niche interest at most.

Strengths and Limitations

The strength is control: every frame can be derived from approved material, which suits explainers, user-interface walkthroughs and typographic social cuts. The limitation is evidentiary. A 2D animation cannot show a real customer's reaction or a real product in use. Scope boundaries also vary by provider. ManyPixels' scope page lists complex 2D animation among work it does not do ^[21], and the publisher's own description in Table 2 should be checked for the same limits before purchase. These are provider statements, not a universal boundary, so the asset pack should be sent for an estimate rather than assumed to fit.

Accessibility applies even to silent pieces. WCAG 2.2 (the Web Content Accessibility Guidelines) asks that either an alternative for time-based media or an audio track be provided for prerecorded video-only content ^[22]. Any spoken line needs captions under success criterion 1.2.2, which asks that captions be provided for prerecorded audio in synchronized media ^[23].

Route 3: New Production With a Crew or Creator

Capabilities

Production is the capture purchase. Wistia describes three stages and defines pre-production as gathering what is needed for the actual shoot ^[7]. It recommends that a production schedule list items for each scene ^[7]. Adobe says a shot list tells you what you need for minimal coverage ^[24], and that B-roll (supplemental footage) can be shot by a second unit or smaller crew to save time and money ^[25]. BLS describes the division of labor: editors arrange footage shot by camera operators ^[6], and camera operators and videographers often shoot raw footage on location ^[6].

Providers differ on whether capture is offered at all. Superside's video page describes full-service video as "Script, cast, shoot, edit, and animate." ^[4] and says: "Only need the shoot? The edit? The voiceover? No problem." ^[4] Kimp's FAQ lists Video Shoots among requests it does not fulfil ^[12], and beCreatives says: "No, we only provide awesome editing services! We can add Stock videos from our libraries and use the footage you provide." ^[5] Nothing reviewed here shows that any other subscription provider films on location; each must confirm that in writing.

Adoption and Cost Structure

No single dependable average exists for a crew day. BLS reports a median annual wage of $74,990 for camera operators and $90,360 for producers and directors ^[26] in May 2025. Those are wages, not what a client pays a company. A union rate card shows how day rates are structured: ICG Local 600's 2025 to 2026 card lists a camera operator at $89.91 an hour and $719.28 a day ^[27]. It is union scale for features and episodic work, with rates published to 1 August 2026, so it illustrates structure only and is not a corporate-video price. Wistia's survey says almost 40% of companies spent under $5,000 on video production in 2025 ^[9], which describes budgets but not what a shoot costs.

Strengths and Limitations

A crew produces evidence of the real product and real people. The limits are planning load and rights. OWC, a storage vendor, states that fixing poor audio is significantly harder than dealing with bad video ^[28], and RODE, a microphone vendor, notes that if audio is recorded too hot and clips, that information is lost ^[29]. Releases belong before the shoot: a University of Hawai'i video services page advises that talent release forms be signed prior to the shoot ^[30] and that location permissions be obtained before any recordings proceed ^[30]. Ownership should be written down too. In a photography example, the U.S. Copyright Office explains that a hired photographer owns the copyright unless it is transferred ^[31]. Counsel should review contract terms for commissioned footage.

Feature Comparison

Table 1 compares the three routes, plus the mixed route, on the inputs they need rather than on price.

Table 1 shows that the routes are sequential dependencies, not rivals: editing and motion consume what production creates or what the buyer approves. A buyer whose scene list is mostly "missing footage" is shopping in the wrong row.

Table 2 lists what each provider's own pages state, as read on 2 November 2026. It is deliberately not a price ranking, because capacity units (requests, daily hours, credits, plan tiers) are unlike and scope differs. Prices appear only where a page showed a plan, a cadence and a number.

Table 2 supports three observations. First, footage rules and runtime ceilings vary by provider and by request type, so limits quoted for one plan cannot be assumed for another. Second, capacity is metered in different units, and dividing a monthly fee by an assumed clip count would compare unlike things. Third, the only reliable comparison is an estimate from the same asset pack sent to each provider. Tapflare's row is the publisher's own description and should be tested the same way: it fits steady recurring edit and motion work, not a single one-off job and not capture.

Scene Audit: Build the Evidence Inventory Before Pricing

Describe the finished video and separate the purchases

Start with the finished piece: channel, final runtime, aspect ratios, caption need, and the single claim each scene supports. Then split the budget into three purchases: pre-production (script, shot list, schedule, releases), capture (crew, talent, locations, screen recording) and post-production (editing, motion, captions, delivery). Wistia describes post-production as broadly covering everything between filming and publishing ^[7]. Keeping the three apart prevents an editing fee from silently absorbing capture risk.

Ask what evidence each scene must carry

Some scenes legitimately communicate with type, screen capture or 2D motion: a definition, a workflow, a plan structure. Others make an evidentiary promise: a customer says it worked, or a product is shown doing a real task. For those, the recording is the evidence, and an editor cannot supply it. The FTC's rule makes it an unfair or deceptive practice for a business to write, create or sell a consumer review or testimonial that materially misrepresents that a reviewer or testimonialist exists ^[8] or that they used or otherwise had experience with the product, service or business ^[8]. FTC staff guidance says the rule has no blanket prohibition on AI-generated avatars in marketing ^[34], and tells businesses to look for red flags that a testimonial is likely fake or false ^[34]; read the full Q&A for the conditions. The Endorsement Guides require endorsements to reflect the honest opinions, findings, beliefs or experience of the endorser ^[35], and an example states that if actors have been employed, that fact should be clearly and conspicuously disclosed ^[35].

Synthetic footage is not documentary evidence. Platforms require disclosure of realistic synthetic material. YouTube says it requires creators to disclose AI-generated or altered content that seems realistic ^[36], and its examples include altering footage of a real event or place ^[36]. TikTok's community guidelines address AI-generated or significantly edited content that shows realistic-looking scenes or people ^[37]. Content Credentials are an open technical standard for establishing the origin of digital content ^[38], and the Content Authenticity Initiative says they are cryptographically signed and tamper-evident ^[39].

Audit each scene

Table 3 is the copyable scene sheet. Fill one row per planned scene; the four rows shown are a template, not a real project.

Table 3 turns a vague brief into assignable gaps. The permissions column deserves the most attention. Adobe Stock states that assets marked Editorial Use Only may not be used for commercial purposes such as advertisements, promotions and endorsements ^[40], and that Enhanced licenses are the type available for Adobe Stock videos ^[40]. Shutterstock's help center describes a sensitive use license as covering uses where a recognizable model may appear to endorse, advocate or believe in a product, service, cause or opinion ^[41], and says sensitive use is not allowed under its Standard or Enhanced license ^[42]. Adobe Stock defines a model release as a legal agreement between contributors and the person depicted ^[43]. For music, Epidemic Sound states that platform limitations vary from plan to plan ^[44], so a music license should be checked against each destination. For anything else, the Copyright Office recommends obtaining permission in cases of doubt ^[31].

Audio and consent belong in the same row. For recorded webinars, Zoom states that it always notifies meeting participants that a meeting is being recorded ^[45]. That notice covers the recording, not permission to reuse a participant's words or face in an advertisement.

Check captions, ratios and file hygiene

Captions are a deliverable with an owner. WCAG 2.2 asks that captions be provided for all prerecorded audio content in synchronized media ^[23], and the Understanding document adds that captions identify who is speaking, not only dialogue ^[23]. The W3C states that WebVTT files provide captions or subtitles for video content ^[46]. YouTube accepts only basic versions of SubRip (SRT) files, with no style markup recognized ^[47], and warns that automatic captions might misrepresent the spoken content ^[48], so an automatic transcript needs a human check. Design Pickle states that captions are auto-generated using AI transcription from the original voice content ^[2], and beCreatives says it uses proofread AI-generated captions ^[33]. The accuracy owner therefore differs by provider and should be asked.

Final ratios multiply deliverables. YouTube states that the standard aspect ratio on a computer is 16:9 ^[49] and recommends uploading in the frame rate recorded ^[49]. TikTok recommends a vertical format for in-feed ads ^[50]. LinkedIn's ratio table lists a 4:5 vertical ratio ^[51]. Meta's Instagram Feed ad specification calls captions optional but recommended ^[52]. Apple describes ProRes 4444 formats as ideal for the exchange of motion graphics media because they are virtually lossless ^[53]. Destinations impose limits too: TikTok's ad specifications list a 500 MB ceiling for in-feed ad files ^[50].

Map input runtime against finished runtime

Table 4 records what the pages reviewed state; "not found" means no figure appeared on the pages read.

Table 4 shows the common failure: a long recording with a short finished cut still counts against the input cap and may count against review time. Count hook, crop and localization variants only where a provider actually prices them that way.

Route with a conditional decision tree

• Is there recorded footage for every scene that makes an evidentiary claim?
  • No: commission capture first (a production specialist, a creator, or an in-house shoot) and delay the subscription start.
  • Yes: continue.
• Is the audio usable and are releases signed?
  • No: fix or re-record; do not send the file for editing.
  • Yes: continue.
• Can the remaining scenes be built from approved scripts, brand files and fresh screen captures?
  • Yes, and the animation is simple: motion design from the asset pack.
  • Yes, but the style is complex (3D, advanced effects): check the provider's exclusions first.
  • No: return to pre-production.
• Is the work recurring?
  • Yes: a subscription can fit.
  • One job only: a project quote is the more suitable purchase.

A mixed route usually resolves to capture, then edit, then motion: the shoot feeds the edit, and motion fills the scenes the camera cannot.

Readiness Sheet and Vendor Questions

Copy the sheet below into a document and complete it before requesting estimates.

• Finished video: purpose, channel, finished runtime, aspect ratios, caption languages.
• Scene table (Table 3): one row per scene, with owner and route.
• Footage inventory: file formats, total size, source runtime, camera and microphone, date recorded.
• Audio check: dialogue intelligible, no clipping, music license platform and plan.
• Permissions: talent releases, location releases, quote approvals, stock license tier.
• Screen captures: date recorded, product version, personal data blurred.
• Brand assets: logo files, fonts, color values, existing templates.
• Missing inputs: each missing shot, with who will supply it and by when.

Send these questions to every provider, with the same asset pack:

• Scope: does the plan cover editing of supplied footage only, animation from approved assets, or both?
• Limits: what are the upload cap, accepted formats and finished-runtime limit for this request type, and do they differ in a trial?
• Review time: does reviewing raw footage count against the buyer's allotment?
• Captions: who creates them, in what file formats, and who checks accuracy?
• Files: which source, project or working files are delivered, and on which plan?
• Licenses: which licenses cover stock footage and music, and for which platforms?
• Voice: is voiceover recorded, generated or supplied by the buyer?
• Variants: are aspect-ratio versions or cutdowns priced separately?
• Commitment: what is the minimum billing commitment, and what happens to unused capacity?
• Capture: do you offer shoots, and is that written in the service terms?
• Exclusions: what is excluded, for example complex animation or 3D?
• Estimate: can the provider estimate from the actual asset pack before any commitment?

Data Analysis and Evidence

Total committed spend

The decision metric is the total committed to get one approved video, not the monthly fee:

Total committed spend = quoted capture + approved pre-production + editing fee for the required billing periods + separately quoted voice, music and licenses + buyer review labor.

Table 5 applies it with the reader's own quotes. This report supplies the structure, not the numbers.

The billing-period term matters. beCreatives lists $1,199 a month on month-to-month billing for its Scale plan and also $2,997 billed quarterly, so a team whose footage arrives in the second month still holds a three-month commitment under the quarterly option. Design Pickle states that all of its plans are on quarterly or annual billing cycles ^[32]. Neither is a criticism; each is a reason to schedule the subscription start after capture.

What the data do and do not show

• Search demand: US estimates for "video editing subscription" (260 a month on average; 10, 20 and 20 in July to September 2026) describe a small, spiky query. They do not show growth or conversions.
• Usage: Wyzowl reports 91% of businesses using video and 92% of marketers planning to spend the same or more ^[10], from a 266-respondent survey run by a company that sells video services.
• Budgets: Wistia says fewer plan to increase spend and almost half are keeping budgets flat ^[9], and almost 40% of companies spent under $5,000 in 2025.
• Challenges: HubSpot's challenge ranking says lacking time to create video content is the top reported challenge for video marketers (^[54]) and lists inadequate budget third (^[54]); the page is labelled 2023 and was updated in November 2024.
• Adjacent citation: A Modash guide says: "Research from Superside shows that short-form content is exploding" ^[55] Superside's linked post compiles other organizations' statistics, so no figure from it is reproduced here.

Table 5 and the list agree on one point: published market figures cannot substitute for a quote on the buyer's own assets.

Case Studies and Real-World Examples

Recorded webinar clips (Hypothetical Example)

A SaaS team has three recorded webinars and wants twelve short clips. The audit shows clean dialogue, a recording notice shown to attendees (Zoom states that it always notifies participants, per ^[45]) and slides at the current product version. Two guest speakers' clips need written approval for use as advertising. Route: remote editing subscription, after permissions. The buyer checks the input cap and finished-runtime limit against the longest webinar and asks who checks captions. With a long source and short finished cuts, review labor is the main cost.

Product demonstration with missing footage (Hypothetical Example)

An e-commerce brand wants a 60-second demo with a hands-on product shot and a customer saying it solved a problem. The audit finds an old screen recording, no hands-on footage and no recorded customer. Route: mixed. A production specialist or creator captures the hands-on shots and the customer interview, with releases; screen capture is refreshed; then an editor assembles. A subscription started first would sit idle while a quarterly commitment ran down. Presenting a stock clip or an AI-generated avatar as a real customer would risk misrepresentation under the testimonial rule, which looks at what the viewer is led to believe; counsel should review testimonial usage as well as contract terms.

Illustrated SaaS explainer (Hypothetical Example)

A SaaS team has an approved script, a brand kit, vector illustrations and a recorded voiceover, and the explainer shows workflows with type and 2D motion. No scene claims a customer outcome. Route: motion design from the approved assets. The buyer checks complexity against the provider's exclusions, asks for an estimate from the asset pack, confirms the caption file and aspect ratios, and ensures silent sections have a text or audio alternative.

Implications and Future Directions

Platform labeling is moving toward provenance. TikTok's guidelines and YouTube's rules already address realistic synthetic content, and TikTok says Content Credentials attach metadata to content (^[56]). Buyers who mix captured and generated scenes should record which is which.

Provider terms also change. The limits cited here differ by plan and request type, and the pages were read on 2 November 2026, so a buyer should ask for current terms in writing with the estimate. Competitor guides will keep comparing capacity; the more durable asset is a documented scene inventory that any provider can quote against.

For Tapflare readers, the implication is proportionate: a daily-hours design and development subscription fits recurring edit and motion work built from supplied or approved assets, and the project manager can confirm scope against the asset pack. It does not replace a shoot and should not be bought to fill a footage gap. Teams ready to test fit can start from the sheet above and ask Tapflare for an estimate. Tapflare is the publisher of this article.

Frequently Asked Questions (FAQs)

What is a video editing subscription service, and what does "unlimited" mean?

It is a recurring plan in which a provider edits or animates supplied or approved material. "Unlimited" usually refers to requests or revisions, with throughput bounded by capacity. EditClub's guide says some plans cap the number of active requests, some sell editor hours and some sell credits, and Design Pickle's creatives work on one request at a time (^[15]).

How much does a video production crew cost?

No dependable universal average is available from the sources reviewed. Crew cost sums roles, days, equipment, locations and rights. BLS publishes occupational wages ($74,990 for camera operators in May 2025), and a union card shows hourly and day-rate structure (^[27]). A written quote for the specific shot list is the only usable number.

Do you need a video production crew?

Yes, if a scene must show a real product in use or a real customer and no recording exists. No, if the scenes can be built from screen capture, licensed stock, typography or 2D motion without implying more than they show.

How to audit video footage before editing

Use Table 3: for each file, confirm format, size, runtime, intelligible audio, signed releases and license tier, then list the missing shots and an owner.

Video editing subscription versus hiring an in-house editor?

An in-house editor is a wage plus tools and management. A subscription is a capacity fee. Either way, an editor needs footage: BLS says editors arrange footage shot by camera operators.

Subscription video editing versus a freelance video editor?

A freelance editor is typically a project engagement, while a subscription suits steady volume. For a single one-off job, a project quote is usually the better fit.

What is a flat-rate video editing service?

A fixed monthly fee for a defined capacity, such as Flocksy's $1,699/mo for its 2 Daily Hours tier (^[16]), Kimp's USD 1397/month Video plan (^[3]) or Penji's $1,995/mo Agency Partner tier. Fees are not comparable until scope and cadence match.

Conclusion

A video editing subscription is the right purchase when recorded, rights-cleared footage keeps arriving and the open question is post-production capacity. Motion design is the right purchase when approved assets exist and the story can be told with type, screen capture and 2D motion. A production crew or creator is the right first purchase when a scene must show something real that has not been recorded. Provider pages confirm that rules vary: input caps, finished-runtime ceilings, caption responsibility, project files and even whether shoots are offered differ by provider, plan and request type. The practical method is the scene inventory in this report, an estimate from the same asset pack, and a total-committed-spend calculation using the reader's own quotes. Sometimes the correct result is to delay the subscription and commission the shoot first. That outcome is not a failure of the audit: it is the audit working, because it moves the spending decision from what a plan can process to what the project actually has. Readers who complete the scene sheet, send the same asset pack to each provider, and total their own quotes will have a defensible answer before the first invoice, whichever route it points to.