When I compare two AI production proposals, I first check whether they promise the same deliverables. Duration, resolution and a final figure can make different amounts of work look deceptively similar.
Procurement needs a fair comparison. The creative team needs something it can revise. I would put both needs into the same conversation.
Put the scope side by side
Imagine two proposals for a 30-second product ad. One includes a landscape master. The other includes vertical and textless versions, plus subtitles in two languages. This is a hypothetical comparison, not a real supplier quote.
The running time is the same, but the files to prepare and review are different. Even matching the version count leaves a question: is the vertical version a crop, or does it include revised composition and text placement?
I would ask both teams to explain the same items:
| Item | What needs to be clear |
|---|---|
| Shots | What is filmed, generated or composited |
| Product accuracy | Which approved assets are used, and who checks details |
| Versions | Language, framing, duration and intended use |
| Revisions | Approval stages and the scope of a revision round |
| Handover | Files and instructions included beyond the final film |
| Schedule | Client decisions and materials each stage depends on |
This does not prescribe how a team should price its work. It makes the work behind each promise visible.
Look at a test and its revision
A portfolio is useful. I also want to understand how a team would handle this project's difficult shot.
A test can address one precise question: does the product retain its shape when it turns, or does a hand interaction explain the operation? The scope and any cost of that test should be agreed beforehand.
If the test moves into revision, I would give a specific note and see what can remain stable. Slowing the hand while accidentally changing the packaging is a result worth discussing.
A supplier may instead explain that a shot would be better filmed, with a clear plan for the assets and post-production. I consider knowing when to change methods part of production judgement.
Discuss the difficult outcome too
“Revisions until satisfied” leaves too much unexplained. What is approved in the first round? What can change later? Which completed work is affected by a new direction?
An incorrect product shape and a rewritten campaign message are different problems. One concerns accuracy; the other may change the production scope. They need different conversations.
Handover deserves the same precision. Can text and sound be replaced independently? Are project files or original assets included? What software or linked materials would another editor need? Possessing a file is not the same as being able to continue working with it.
Make the choice around the project
I would mark the meaningful differences before judging their value. One job may depend on product stability; another may need to explore several creative directions quickly. The useful choice can differ.
This is a creative and production perspective. I am also the founder of ONCE, so it should not be read as an independent recommendation of my own company's services.
The questions in Preparing an AI Advertising Brief can help establish a common scope before approaching teams. That gives the discussion more substance than repeatedly asking whether the final figure can come down.