What a discovery call is actually for
Knowing how to run a discovery call comes down to accepting what the call is not: it is not a pitch, not a demo, and not a price negotiation. A discovery call has exactly two jobs. First, qualify the client - figure out whether this project is real, funded, and worth your time. Second, gather the specifics that let you write a proposal so tailored the client feels understood before they see a number. Creators who treat the call as a performance talk for twenty minutes and learn nothing; creators who treat it as a diagnosis ask questions for twenty minutes and walk away holding everything they need to close.
The call sits at a precise point in the client pipeline. Your content, outreach, or referrals create the conversation - that part is covered in [how to get your first AI video client](/blog/how-to-get-your-first-ai-video-client). The discovery call turns interest into a scoped project. The proposal turns the scoped project into a yes, using the structure in [how to write an AI video proposal that wins clients](/blog/how-to-write-an-ai-video-proposal-that-wins-clients). Skip the call and your proposal is generic; skip the proposal and your call was a pleasant chat. The system works because each piece feeds the next.
The 20-minute structure: 5-10-5
Structure beats charisma on every call. Use three blocks: five minutes of listening, ten minutes of questions, five minutes of next steps. Open by asking the client to describe the business and what made them reach out, and then stay quiet - the unprompted first five minutes almost always contain the real motive, and the real motive is what your proposal needs to speak to. Then move into your questions, and close by telling them exactly what happens next. The structure also protects you: without it, discovery calls balloon into an hour of free consulting that never converts.
The 5-10-5 discovery call structure
| Block | Time | What happens |
|---|---|---|
| Listen | First 5 minutes | They describe the business and the problem in their own words; you take notes and do not interrupt |
| Diagnose | Middle 10 minutes | You ask the five scoping questions: purpose, placement, brand, deadline, budget |
| Direct | Final 5 minutes | You summarize what you heard, state what you will send and when, and name the one thing you need from them |
The five diagnostic questions in the middle block do the heavy lifting, and they are the same five that drive onboarding: what is this video supposed to accomplish, where will it run, what brand rules exist, when do you need it, and what budget range are you working with. The full question set - and how the answers flow into kickoff once the client signs - is in [how to onboard a new AI video client](/blog/how-to-onboard-a-new-ai-video-client). Asking them on the discovery call means your proposal, contract, and kickoff all draw from one conversation instead of three.
Qualifying: the signals that predict good and bad clients
Half the value of a discovery call is the projects you decline. A bad client costs more than no client - unpaid revisions, scope creep, and a testimonial you cannot use. The strongest positive signals: they answer the purpose question specifically, they name a real deadline tied to a business event, and they engage with the budget question instead of deflecting it. The strongest warning signs: they cannot articulate what the video is for, they want it as cheap as possible before hearing what it involves, or they describe firing their last three creatives with visible irritation.
- Green flag: a specific purpose - 'we are launching a second location in March' beats 'we want more content' every time.
- Green flag: a budget conversation held in ranges rather than refused outright.
- Red flag: 'this should be quick since the AI does the work' - a client who discounts the craft will discount the invoice.
- Red flag: no decision-maker on the call and no path to one - you are pitching a messenger.
- Red flag: urgency without a reason - manufactured deadlines produce rushed work and slow payments.
Budget deserves a direct question, asked without apology: 'so I can scope this properly, what range are you working with?' Most clients answer, and the ones who refuse have told you something too. What to do with the number - and how to price when the client asks you first - is covered in [how to price AI video services for clients](/blog/how-to-price-ai-video-services-for-clients). On the call itself, never quote a final price; anchor a range at most, and save the exact number for the proposal where your proof and plan frame it properly.
Closing the call and converting it
The last five minutes decide whether the call converts. Summarize what you heard in two or three sentences - purpose, placement, deadline - and let them correct you. That summary is quiet proof you listened, and it is the moment most clients relax. Then make one concrete commitment: 'I will send a one-page proposal by Thursday with the scope and price.' A named deliverable on a named day keeps momentum; a vague 'I'll put some thoughts together' is where deals go to die. Finally, ask for the one thing you need from them, usually brand assets or examples of work they admire.
Send the proposal within 48 hours while the conversation is warm, built from your call notes so every line reflects their answers. Once they accept, the agreement from [how to write an AI video contract that protects you](/blog/how-to-write-an-ai-video-contract-that-protects-you) locks the scope, and your revision policy from [how to handle client revisions on AI video](/blog/how-to-handle-client-revisions-on-ai-video) is already set in writing before the first draft ships. If the discovery call surfaced a monthly content need rather than a one-off, propose the retainer instead - the math and structure are in [how to sell AI video as a monthly retainer](/blog/how-to-sell-ai-video-as-a-monthly-retainer).
Inside the Higgsfield Income Club at higgsfieldincomeclub.com, $9/month, AI video creators share the discovery call scripts, qualifying questions, and follow-up sequences that are closing real clients right now, along with the red flags that cost members real money to learn. It is the fastest way to sound like you have run fifty of these calls when you are on your first five.
Short, practical drops on income paths, prompts that work, packaging for views, and getting paid. No spam, unsubscribe anytime.
Frequently asked questions
How long should a discovery call be?
Twenty minutes, stated up front when you book it. Use a 5-10-5 structure: five minutes letting the client describe the business and problem, ten minutes asking your five scoping questions, and five minutes summarizing and stating next steps. Short structured calls book more easily and signal that you run projects with the same discipline.
What questions should I ask on a discovery call for video work?
Five: what is this video supposed to accomplish, where will it run, what brand rules or assets exist, when do you need it, and what budget range are you working with. The answers qualify the client and give you everything a tailored proposal needs, so you never write from a template.
Should I give a price on the discovery call?
Give a range at most, never a final number. A price quoted mid-call lands without context, while the same number in a proposal lands after the client has seen your proof and plan. If pressed, anchor with 'projects like this typically run between X and Y' and commit to exact pricing in the proposal within 48 hours.
How do I know if a client is worth taking?
Look for a specific purpose tied to a real business event, willingness to discuss budget in ranges, and a decision-maker on the call. Decline or slow down when the client cannot say what the video is for, leads with cheapness, or expects the work to be trivial because AI is involved - those patterns predict unpaid revisions and payment friction.
What should happen after the discovery call?
Send a one-page proposal within 48 hours, built directly from your call notes, with the delivery day you committed to on the call. Once the client accepts, follow with a contract covering scope, revisions, ownership, and payment before generating anything. Call, proposal, contract, kickoff - each step feeds the next.
Last reviewed by David on September 21, 2026
