Get five LinkedIn ad copy variants testing different angles, written within platform character limits, labeled for A/B testing with run-first recommendations.
Provide the client's URL, their campaign goal, and the target audience on LinkedIn. The flow returns five complete ad variants with intro text, headline, and CTA for each, character counts included, and a recommendation for which 2-3 to run first in Campaign Manager.
Each brief includes five complete variants, each testing a different angle: statistic-led hook, pain-point opener, social proof, curiosity gap, direct benefit statement, and others depending on the product and audience. Every variant has intro text (up to 600 characters), a headline (up to 70 characters), and a CTA. Each is labeled with its testing hypothesis so the team knows what it's designed to prove before putting spend behind it. The brief includes character counts per element and a recommendation for which 2-3 variants to run first, based on the campaign goal and audience segment.
LinkedIn truncates intro text after 150 characters on mobile, placing a "see more" cut before the reader sees the rest of the ad. That makes the first two sentences carry the full weight of the hook. Unlike Meta, where a strong visual can carry weak copy, or Google Search, where intent already exists, LinkedIn ads open in a passive scroll environment where attention has to be earned before the truncation point. Every variant is written with the 150-character break in mind.
Same funnel stage and format, different direction. Use this after reviewing the first set of variants when you want to explore alternative angles before committing to a test.
Lead with customer results instead — things like "cut office setup time by 40%" or "same-day delivery on orders over $500." Replace the two weakest variants with data-driven angles.
LinkedIn ads need different copy at awareness versus conversion. The context from the previous prompt carries over, so you get a new set of variants without re-briefing the client.
Now write awareness ads for the same client — top of funnel, targeting facilities managers who aren't aware of IKEA Business yet. Recommend a format and write five awareness variants.
Once the copy is confirmed, generate an ad visual matched to the format dimensions. The context and variant copy carry over automatically.
Create an ad visual for variant A. Professional and clean — it needs to stop the scroll in a LinkedIn feed without being loud.
LinkedIn ad copy gets written on repeat: new campaigns, new offers, same client voice and audience. A Juma project stores the brand context and performance history so every round of variants starts from what's already known about the client, not from scratch. If the client project already exists, add these items to it.
Brand Voice Guide
How the client sounds in ads: tone, vocabulary, what to avoid. With this in the project, copy drafts land in the right voice from the first pass without rounds of corrections.
Audience Profile
Who the ads target on LinkedIn: job titles, seniority, company size, and pain points. This context shapes which angles the AI leads with and which CTAs it recommends for the audience segment.
Ad Performance History
Past campaign results: CTR by variant, format, and hook type. The AI uses this to weight recommendations toward what has worked and away from what hasn't for this specific audience.
Output Examples
Two to four high-performing ads from past campaigns, with brief notes on what each was testing. These serve as structural references so new variants match the format and length the team knows performs.
Add a short description to each knowledge item in the project's info field so Juma knows what each file contains and when to use it. For example:
