--- name: "thought_leadership_ads" description: "Promote a person's existing organic posts as paid ads: pick posts that already earned attention with a selection scorecard, clear permission and exclusion rules, set budget and flight choices, and measure paid lift against a recorded organic baseline. Use when someone wants to run LinkedIn Thought Leader Ads, Meta partnership ads from a creator post, boost a founder or employee post, or decide whether a post deserves paid spend." --- # Thought Leadership Ads ## Purpose Paid promotion of a post a person already published under their own name. The advertiser pays for wider distribution. The author keeps the byline, the wording, and the comments. The best known version is LinkedIn Thought Leader Ads. Meta partnership ads work on the same principle for creator and employee posts on Facebook and Instagram. The format buys reach for a voice people already chose to read. That is also its limit. The advertiser cannot rewrite the post, add a headline, bolt on a call to action button, or swap the image during ad setup on LinkedIn. Selection is the skill. A weak post with budget behind it is a weak post seen by more strangers, under a real person's name. One trust fact explains why teams try this at all, and it is context, not a promise of results: in the 2024 Edelman and LinkedIn B2B Thought Leadership Impact Report, which surveyed nearly 3,500 management-level professionals, 73 percent of decision-makers said thought leadership is a more trustworthy basis for judging a company's capabilities than its marketing materials. People weigh a person's argument differently from a company claim. Paid distribution does not create that trust. It can only widen a post that already earned some. Evidence tiers in this skill: **Official** means LinkedIn or Meta states it on its own product or help pages. **Reported** means practitioner datasets and guides agree on it, with named sources in the scratch notes. Everything else is this skill's operating method, labeled as such, and the reader can change the thresholds and keep the structure. ## Workflow ### 1. Confirm the format fits the job - Use this format for credibility, reach into a named audience, author follower growth, and warming an audience for a later campaign. Official: LinkedIn offers Thought Leader Ads under the Brand Awareness, Engagement, and Video Views objectives. No lead form or conversion objective sits inside the standard format, so a plan that needs form fills must pair this format with a separate retargeting campaign and say so plainly. - Walk away when the goal is a direct response target this month, when no one at the company posts anything worth reading, or when the only candidate posts are product announcements rewritten in first person. Readers can tell, and paid reach makes the tell louder. - Pick the platform by where the post already lives. LinkedIn for professional posts by employees, founders, customers, and creators. Meta partnership ads for creator and employee posts on Facebook and Instagram. Do not port a post across platforms and call it the same campaign. The permission system, labels, and audience differ. ### 2. Build the candidate pool before touching Campaign Manager - List every post from the last 30 to 60 days by each author the team might promote: founders, executives, practitioners, analysts, support leads, customers who post about the work, and creators under contract. Practitioners often outscore executives here, because their posts name real problems. - Pull the organic numbers for each post and for each author: impressions, reactions, comments, shares, and the author's median engagement rate across their last 10 to 20 comparable posts. - Read the comments, not just the counts. Note who commented and how many comments added something real. The scorecard needs both. - Snapshot the baseline for any post that might run, before any campaign exists. Fields and rules are in `references/measurement.md`. LinkedIn gives the same instruction as a pro tip on its own product page: save organic performance before the campaign so the impact can be measured. ### 3. Score every candidate, gates first - Run the hard gates in `references/scorecard.md` first: format eligibility, public and original post, verifiable claims only, no confidential material, author willing in principle. - Score the survivors with `scripts/score_post.py`. It implements the full card: organic proof against the author's own median (25), audience fit from who actually commented (20), substantive comment share (15), a human pain and lesson rubric (20), timing (10), and author readiness to reply during the flight (10). - Apply the verdicts as written: 70 or higher promotes, 50 to 69 gets a small test with a kill rule written in advance, below 50 stays organic. A gate failure blocks a post at any score. The worked pair in `references/worked-campaigns.md` shows a 78 that promotes and a 36 that passes, where the passing post had similar lift over its author's median and lost on audience, content, and readiness. - Never rank authors against each other on raw totals. Rank posts against their own author's baseline. A small account with a strong post is a better buy than a large account coasting. ### 4. Clear exclusions before permission Do not request permission, and do not promote, when any of these hold: - The post depends on a claim, statistic, or customer result the team cannot source or has not been cleared to use in paid media. Paid use is a wider use than an organic post, and clearance for one is not clearance for the other. - The post names competitors, customers, pricing, or product plans that leadership or legal has not cleared for paid distribution. - The author wrote the post about a personal topic unrelated to the campaign, and promotion would drag their personal audience into a company funnel. Asking damages the relationship the format depends on. - The author is an employee who has signaled discomfort with promotion, in words or by declining before. One decline ends the ask for that author until they raise it themselves. - The post is still climbing organically and the topic is not time-sensitive. Wait for velocity to slow. Paying at peak buys impressions the post was about to earn for free and muddies the baseline. - For non-employees (customers, creators, advisors), there is no written agreement covering paid use, term, and territory. Platform approval alone does not settle payment, usage term, or what happens to the post after the flight. ### 5. Get permission through the platform, and record it LinkedIn, Official setup as summarized from LinkedIn's documentation: - The person setting up the campaign needs Page access at Super Admin, Content Admin, or Sponsored Content Poster level for the associated LinkedIn Page, plus Creative Manager access or higher on the ad account. Re-check role names in Campaign Manager at setup time, because LinkedIn renames roles without much notice. - In Campaign Manager, open the campaign, go to the ads section, choose Browse existing content, then LinkedIn members, then the Posts tab. Search by author name or paste the post URL. LinkedIn surfaces posts from employees and from first and second degree connections. - Select the post and choose Request approval. The author is notified and can approve or decline, and can also approve pending and future requests from an advertiser they trust. Campaign Manager shows the request as pending, approved, or declined. - The author can revoke approval at any time. Revocation turns the ad off. The campaign keeps running only if other approved ads remain in it. Tell the author this before they approve, not after it happens to them. - Record, outside the platform: author, post URL, approval date, scope approved (this post only, or future posts too), agreed term, and any fee or agreement for non-employees. Meta, for partnership ads: - Permission comes in two forms. Account-level permission lets the advertiser create partnership ads from the creator's handle on an ongoing basis. Content-level permission covers one post, story, or reel, shared by the creator as a partnership ad code. - For boosting an existing tagged post, the creator turns on the option that lets the brand partner boost it, with the paid partnership label on the organic post. The promoted version carries a Sponsored label. - Creators can revoke permission there too. The same recording rule applies: who approved what, for which content, for how long, under which agreement. ### 6. Build the campaign around what the format cannot change - Official for LinkedIn: the post is the ad. No headline or introductory text can be added, no URL can be added, and no call to action button is available on the standard single image and video formats. A link the author put in the original post stays clickable. If the post needs a tracked link and has none, the author edits their own post before approval, or the team accepts untracked traffic and writes that into the measurement plan. - Official for LinkedIn: standard Thought Leader Ads run from posts with a single image or a single video, and text-only posts also run. Multi-image posts, polls, and document posts cannot be used in the standard format. LinkedIn also offers separate thought leader versions for events, articles, and newsletters under their own formats, so check the current options in Campaign Manager before ruling a post out on format alone, and never assume the standard format's rules cover those variants. - Choose one objective per campaign and one audience definition the team can describe in plain words. Mixing a prospecting audience and a customer audience in one campaign makes the read useless for both. - Set the flight in advance: start date, end date, total spend cap, read dates, and kill rules. A common working shape is a 7 to 14 day flight per post, long enough for comments to accumulate, short enough that the topic stays current. That shape is this skill's method, not a platform rule. - Budget from the decision, not from habit. State the total the team is willing to spend to learn whether this author and topic deserve a repeat, then divide by flight days for the daily figure. Check the result against the platform floor at setup time. LinkedIn's published minimum daily budget is 10 dollars per campaign, with a 100 dollar minimum lifetime budget for new campaigns, and both figures should be re-verified in Campaign Manager because minimums change. Practitioner guides commonly work at 50 to 100 dollars per day per campaign to get a readable result. Treat that range as reported practice, not as a requirement, and never present it as a LinkedIn recommendation. - Line up comment cover before launch. The author replies to comments during the flight, and one named person on the team watches for questions that need a company answer, factual errors, and abuse, at least twice a day. A promoted post with unanswered critical comments under it harms the author first and the brand second. ### 7. Measure paid lift against the baseline, honestly - Follow `references/measurement.md` in full. Its core rules: snapshot before launch, never compare headline click rates across formats, split linked and unlinked posts, report paid lift with the method stated, and decide once at the pre-set read date. - Know the click trap cold. On LinkedIn, the headline click-through rate for this format counts all clicks, including reactions, comments, profile clicks, and expanding the text. Comparing it with a company page ad rate built on landing page clicks flatters this format and misleads the budget meeting. For any cost or traffic comparison, compute effective click-through rate as clicks to landing page divided by impressions, and effective cost per click as spend divided by clicks to landing page, and label both as derived metrics. - Treat outside benchmarks as weather reports, not targets. LinkedIn's early figures for the format, widely cited as 1.7 times the click-through rate and 1.6 times the engagement rate of other single image ads, came from early testing. Later independent datasets disagree with each other on landing page cost, in both directions. Your own baseline, your own audience, and your own eCPC decide whether to repeat. - Report author-level effects, because they are half the reason to run this format: follower change for the author during the flight, substantive comments from target accounts, profile-driven conversations the author can point to, and self-reported form answers that name the author or the topic. - Close every flight with one decision per post: scale the same post, repeat with a new post by the same author, or stop with this author or topic. Write the reason in one sentence tied to a number in the report. ## Output Contract Deliver these, in this order: 1. Candidate table: author, post link, publish date, organic snapshot values, scorecard total, gate result, verdict. 2. Exclusion log: each post rejected before scoring, with the exclusion rule that caught it. 3. Permission record: author, post, platform, approval status and date, scope, term, and any agreement or fee note for non-employees. 4. Campaign plan per promoted post: platform, objective, audience in plain words, flight dates, total cap and daily budget with the math shown, kill rules, read dates, comment cover owner. 5. Measurement report per flight: baseline values, paid results, effective click-through rate and effective cost per click where a link exists, author follower change, self-reported mentions, and the closing decision with its reason. 6. A short note on what the team will ask each author to write next, drawn from what the scorecard showed was missing. This turns one flight into a supply of better posts. ## Operating Rules - The author's name and reputation carry the campaign. Any choice that trades the author's credibility for a short-term metric is out, even when the author would probably never find out. - Never edit, coach edits into, or quietly rewrite a post after approval and keep calling it the same approved post. If the post changes in substance, treat it as a new candidate and get approval again. - Disclosure is part of the work, not a footnote. Employment is a material connection. Under the FTC Endorsement Guides and its Disclosures 101 guidance for influencers, a material connection that might affect how readers weigh a statement must be disclosed in a way that is hard to miss, in clear language, with the endorsement itself, not buried in hashtags or below a fold. The platform's Promoted or Sponsored label is the floor. When a post's wording could read as a purely personal opinion while the company pays to distribute it and employs the author, flag the wording for the team's own review before launch. This skill is not legal advice. Regulated industries and non-US campaigns need their own review, since endorsement and employee advocacy rules differ by country. - Never promise an author a result, a follower count, or a comment volume in exchange for approval. Promise only the budget cap, the flight dates, the audience, and a copy of the final report. - Never promote a post by someone who has left the company without fresh written agreement. Platform eligibility and past approval do not settle that case. - Pay non-employee authors under a written agreement that covers paid use. Platform approval is a technical switch, not a contract, not a fee agreement, and not a usage term. - Keep paid and organic numbers separate in every artifact the team produces. A combined total with no split is a reporting failure, and any lift claim built on an unrecorded baseline gets struck from the report. - When platform details conflict between sources, or a help page and Campaign Manager disagree, Campaign Manager at setup time wins, and the conflict gets noted in the campaign record with the date.