Holly objections · 04 Holly playbook

Label the feeling, ask one calibrated question if needed, then answer plainly. Do not argue. More objections are in the Objection handbook.

Objection	Answer
"I don't trust AI to post for me."	Label: "It sounds like you've seen what AI does to LinkedIn and you don't want your name on it." Ask: "What would you need to see before you'd trust it with one post?" Answer: "That's the right instinct. Holly doesn't need your trust on day one. We recommend starting on Suggest: it proposes, you decide what goes ahead. You give it more only when it's earned it, and you can take it back."
"Will it embarrass me with clients?"	"That's the risk worth worrying about. Three things stand in the way: it writes in your voice, built from your own writing; your firm's rules check every post before it goes out; and we recommend a person approves every LinkedIn post. A bad draft costs you ten seconds."
"Isn't this just ChatGPT?"	"You'd get the writing free. Everybody has the writing now. What you won't get is a voice for each of your people, your rules checked before posting, approvals, a calendar, campaigns, reports, and an assistant that does all of it when you ask. Running it across a team, every week, is the job." If asked: "Claude, Anthropic's model. The value isn't the model. It's the voices, the rules and the work Holly does around them."
"Will it replace my marketing person?"	"No, and we'd never sell it that way. Holly takes the routine work off the people who already have the judgement. Your marketing person stops chasing twelve people for posts and starts running the plan."
"What if it gets something wrong?"	"It will get things wrong sometimes. That's why we recommend a person approves every post, why your rules check every post, and why every action is logged. If Holly gets a fact wrong in a draft, you catch it in the draft."
"Can I turn it off?"	"You set each person's level, down to Observe, where Holly is designed to watch and show you what it would do. You can also pause voice learning or mark a post 'don't learn from this'."
"Will it train on our content?"	"Holly learns how you write from your own writing, and you control that. For how data is handled with the model provider, I'll get you the written position from our legal team rather than paraphrase it." Never make your own written assurance
"Will people tell it's AI? Will LinkedIn flag it?"	"LinkedIn has said refining your own words with AI is fine. What it's pushing down is generic content. We won't tell you Holly is undetectable, because nobody honest can. Each person's own voice and the checks before posting reduce generic output, and a person who approves the post has read it and stands behind it."
"Do I have to tell people my posts are AI-drafted?"	"Inside Maktura, Holly tells you it's AI. On the posts themselves, that's a decision for your firm, and I'm not going to give you legal advice on it. Set up the way we recommend, every post is drafted with AI and approved by a person, and the record shows who approved it." Never promise a watermark, fingerprint or disclosure feature (D9)
"What does Holly cost on top?"	"Nothing. Holly is in every licence. The software is all in: every feature and Holly, in every seat." Then price per Commercial answers. Never say "no add-ons" flat
"Is it safe for a regulated firm?"	"Start Holly on Observe. Set approval for the whole workspace, with a chain if you need one: marketing, then compliance. Once the data processing terms are in place, upload your social media policy and it becomes rules that check every post: a second pair of eyes, not a compliance sign-off. You stay accountable for what goes out, and the record shows you did." Written security or data assurances come from Karl Foster
"My people won't use another tool."	Ask: "What happened last time?" Answer: "That's the problem Holly exists for. Nobody has to learn the tool. They get the morning list, pick a story, read Holly's draft and approve. If they can reply to an email, they can use Holly."
"It's a minimum viable product. Why trust it?"	"Fair. It is an early release, and you will see rough edges. That's why the founding rate is lower: you're using it early and helping shape it, and you have a direct line to the team. And it's why we recommend starting on Suggest with approvals on."
"Can Holly reply to comments for me?"	"No, and deliberately so. No automated likes, comments, connection requests or messages. You see who's engaging with your posts and reply yourself, from one place. Your name, your reply."
"Instagram, X, phone, voice?"	"LinkedIn first; ask us about other platforms and I'll get you a straight answer. And today you type to Holly. I'm not going to promise you anything else." Never list other platforms as live (D13). No mobile app, voice input or avatar
"What if Holly breaks?"	"It will have bad days; it's an early release. Set up the way we recommend, you see every post before it goes out, every action is logged, and Holly flags faults to our team itself." Never quote an uptime or service level