Gemini for SaaS Ad Copywriters: Match Client Voice in Social Posts

You are an expert SaaS brand voice and social media copywriting consultant with 11 years of experience helping agency copywriters extract and replicate client voice from existing content samples with enough precision that the client's internal team cannot distinguish the agency-written posts from posts the founder wrote personally. Help me create a social post series from an article so I can publish more consistently and stop losing 4–6 hours per client per week to voice revision cycles that delay every scheduled post. My situation: - My client's SaaS product and the audience they speak to: [e.g., Stackify — a developer productivity platform — their audience is CTOs, VP Engineering, and senior engineering managers at companies between 50 and 500 employees] - The existing content samples I have access to for voice extraction: [e.g., 12 LinkedIn posts the founder wrote personally in the past 6 months — I have access to all of them — I can summarize the patterns I observe: the founder opens with a specific team failure observation rather than a success story, uses short punchy sentences under 12 words, cites a specific number in every post, and never uses the word "leverage"] - The article I am repurposing into social posts: [e.g., a 1,800-word technical blog post titled "Why your engineering team's sprint velocity is lying to you" — the article argues that velocity is a local optimization metric that can improve while overall product output quality declines — it challenges the standard Agile practice of using velocity as a health metric] - The client voice problem I am currently facing: [e.g., my previous 3 social post batches have been returned with 6, 8, and 11 comments respectively — the client says the posts "sound like marketing" rather than like someone who has managed engineering teams — specific feedback: too many hedge words like "can often," "may be," "in some cases" — the client uses declarative statements, not hedged observations] - My publishing target and the client's tolerance for off-voice posts: [e.g., 4 posts per week, 2 LinkedIn and 2 Twitter/X — the client has said that one more off-voice batch will result in them bringing the work in-house — I need this batch to be right without a revision cycle] - The specific voice attributes I have identified from the 12 founder posts: [e.g., opens with an observation about what a specific role (CTO, VP Eng) does wrong — cites a specific number or percentage — uses a 2-word or 3-word sentence at least once per post as a deliberate rhythm break — ends with an opinion, not a question — never uses the word "important" or "crucial"] - The campaign constraint: [e.g., all 4 posts must be derived from the same article — no external data sources — each post must be new information for a follower who has read the previous 3] Deliver: 1. Write a voice extraction document — a one-page reference covering the 8 specific voice rules extracted from the 12 founder posts, with one example from the founder's actual posts illustrating each rule and one example of how the rule applies to the engineering manager audience. 2. Write 2 LinkedIn posts derived from the article — each 120–160 words — that follow the voice extraction rules from item 1, open with an observation about what a CTO or VP Engineering does wrong, cite at least one specific number from the article, and end with a declarative opinion rather than a question. 3. Write 2 Twitter/X posts derived from the article — each under 280 characters — that carry the same tonal authority as the LinkedIn posts, use at least one sentence under 10 characters as a rhythm break, and do not duplicate the insight used in the LinkedIn posts. 4. Write a voice compliance checklist — 6 yes/no questions I run against every post before submitting to the client — covering hedge word count, sentence length variation, declarative ending, specific number presence, forbidden word absence, and the opening observation specificity. 5. Write a revised version of the weakest post from my previous batch — identify from the description of the 11-comment batch which type of error occurred most frequently, rewrite one representative example from that batch to comply with all 6 rules from the voice compliance checklist. 6. Write a voice drift early warning system — a 3-question test I apply to my first draft of any new post before writing the full batch, designed to catch the hedge word and "sounds like marketing" patterns before I write 3 more posts in the same failing register. 7. Write a client voice brief template — a 1-page document I complete for every new client before writing a single word of social copy, covering the 6 voice attribute fields most predictive of client approval: sentence length pattern, forbidden words, opening structure type, ending type, numerical evidence requirement, and the single phrase that most captures the founder's editorial personality. **Write both LinkedIn posts and both Twitter/X posts as complete final-draft copy ready to schedule — not drafts for review, not frameworks — every post must pass all 6 voice compliance checks before I receive it, and I should be able to send this batch to the client without any revision.**