Two speakers tailor a keynote to the same industry. The results shouldn't match.

An empty Brand Kit does not break the tool. It just leaves that paragraph out

If you have never branded your hub, this section of the prompt is blank, and the generation falls back to industry analysis alone: three to five challenges the industry faces today, then a rewrite that leans on that analysis instead of on you. That is a usable keynote variant. It is also the generic one, the kind that reads like it could have been written for any speaker who covers the same topic.

Filling in a Brand Kit takes one visit to Settings. Add your website there and Speaker Hub pulls in your name, topics, bio, and voice automatically. No website to point at, or want more control than that, type the fields in yourself, including the voice and tone description and the optional profile paragraph meant for exactly this: telling an AI tool who you are in your own words instead of letting it guess.

The audience field is not a second industry field

Below the industry input sits a smaller one asking for a specific audience. It is optional, and plenty of speakers leave it blank because the industry field already feels like it answered the question. It did not answer all of it. "Healthcare" covers a hospital board and a floor of night-shift nurses, and a keynote that lands with one will not land the same way with the other.

Fill in the audience field and the prompt adds a second, narrower line naming the specific audience within that industry. The model still analyzes the industry as a whole, but the rewrite aims at the room you actually named. "Healthcare" plus "hospital administrators deciding next year's budget" pulls on a different set of pressures than "healthcare" plus "frontline nurses six months from burnout," even though both start from the same master keynote and the same industry pick.

The two inputs stack, they do not replace each other

Your Brand Kit and the audience field are not competing settings. One tells the model who is speaking. The other tells it who is listening. A variant built from both is doing real work: a specific person's voice, aimed at a specific room, inside an industry the model has already thought through. A variant built from neither is still a tailored keynote. It is just the version anyone who typed the same industry would also get back.

If you have not tailored a keynote yet, how to tailor a keynote walks through the four layers worth changing before you touch this tool at all. And once a variant comes back, read Keynote Tailoring does not fact-check your talk before it goes anywhere near an organizer, since a fuller Brand Kit changes the voice of a variant, not whether every claim inside it is true.

• keynote
• speaking business
• audience

Back to all posts