Pitches are decided by the diagnosis more often than by the idea. The room has usually seen three competent ideas by the time you present; what they have not heard is someone accurately describe their problem back to them. Lead with that, then show one idea three ways rather than three ideas once, and name the risk before they find it.
Lead with the diagnosis
Two minutes on what you think is actually wrong, in their language, with evidence. Not the brief restated, which signals you read it. Something they suspected and had not articulated.
If the diagnosis lands, everything after it is heard as informed. If it misses, the strongest work in the room is heard as decoration. This is the highest-leverage two minutes in a pitch and most teams spend it on credentials.
One idea, three ways
Counter to instinct, and it wins more. Three separate ideas make the room compare rather than commit, and they signal that you have no conviction. One idea shown across three formats or three executions shows range while making the choice for them.
Keep the other directions in the back pocket for the questions. "We explored a version where X, and here is why we set it aside" is a strong answer and a weak opening.
Show, do not describe
The change worth exploiting: producing finished-looking work for a speculative pitch used to be prohibitive, so pitches were words plus a mood film. Now three style frames on the prospect's actual product cost 18 to 63 credits on 8frame, roughly $0.18 to $0.63 at pack rate, and one animated beat is 54.
A pitch showing their product in your idea beats one describing it, and the cost is no longer the reason not to. Two practical rules: say clearly that it is concept fidelity, and do not over-finish, because a pitch that looks delivered invites the question of what exactly they are paying for.
Name the risk yourself
Every idea has a weakness the room will find. Naming it first converts it from a discovered flaw into evidence of judgment: "the risk here is that it depends on a spokesperson, and here is how we would de-risk that."
Teams that present flawless ideas are trusted less, not more, because everyone in the room knows there is no such thing.
The preparation that happens before the deck
Find out who decides, and whether they are in the room. Pitches lost to "we went another way" are frequently pitches presented to people who could not say yes.
Find out what they rejected before. If the incumbent tried something similar, you need to know.
Ask one question in advance. It signals engagement and often surfaces the real constraint.
Rehearse the first two minutes only. Beyond that, over-rehearsal makes teams brittle when interrupted, and good rooms interrupt.
FAQ
How many ideas should a pitch present? One, expressed several ways. Three ideas is a menu, and menus produce deferral or a hybrid nobody designed.
Should we show AI-generated work in a pitch? Yes, labelled as concept fidelity. It shows the direction on their product rather than someone else's, which is a stronger argument than a reference board.
How long should a pitch be? Shorter than the slot. Leaving time for questions is where the relationship starts, and the questions are where you find out whether you won.
Diagnose accurately, show one idea properly, and name the risk first. The 8frame canvas is free and unlimited, and generation is paid from $19/month.