LinkedIn's AI creative tools: faster ads, but strategy still matters

Most B2B technology marketers know they should test more creative on LinkedIn. Few have the time to produce five versions of every ad. LinkedIn's latest set of creative tools is aimed squarely at that gap, and for lean teams at SaaS companies, MSPs and Microsoft partners, it's worth understanding what they do well and where human judgement still has to lead.

What's happening

LinkedIn has rolled out a group of AI-assisted creative features in Campaign Manager, positioned mainly for small and growing businesses.

Brand kit stores your colours, fonts, logo, tone of voice and key messages once, so anything generated later stays on-brand. Draft with AI produces ad copy from a URL, campaign goal and any context you supply, such as past top performers. Ads personalisation tailors messaging using attributes like job title, company and industry. AI ad variants create alternative headlines and introductory text from an existing ad. Flexible ad creation takes a pool of images, videos and copy, mixes them into multiple versions and shifts delivery towards the best performers.

The case for all of this is experimentation. LinkedIn's internal analysis found advertisers running five or more ad variants saw over 20% higher click-through rates than those running a single ad. It also reports that businesses using flexible ad creation generated roughly 7% more creative options.

What it means for B2B tech marketers

The bottleneck these tools address is real. In most Australian tech businesses we work with, LinkedIn ads are produced in a rush alongside events, partner commitments and sales requests. One ad goes live, it underperforms or doesn't, and nobody knows why. More variants mean faster learning.

But there's a risk worth naming. If every competitor in the Microsoft partner ecosystem uses the same AI drafting tool from similar landing pages, the output will converge. Expect a lot of ads promising to "unlock the power of Copilot". The tools speed up production; they don't supply the insight, the proof point or the specific customer problem that makes someone stop scrolling.

The smart approach treats AI as a production assistant for a strategy you've already defined. Decide the three or four distinct angles you want to test, such as a cost argument, a risk argument, a customer outcome and a point of view. Then use the tools to produce clean variations of each and let the data decide.

How to put it to work

• Set up brand kit properly. Spend an hour on tone of voice and key messages, including words to avoid. This is what keeps generated copy from sounding like everyone else.
• Feed Draft with AI real proof. Give it your strongest landing page and past high performers, then edit the output to add specific numbers, customer names you're cleared to use, or an Australian context.
• Test angles, not just words. Use AI ad variants for headline tweaks, but build separate ads for genuinely different messages.
• Load flexible ad creation with variety. Supply contrasting images and a short video, not five near-identical stock photos.
• Use personalisation sparingly. Industry-specific references can lift relevance; overdone personalisation can feel automated.
• Review weekly. Retire losing variants, keep the winning angle and brief the next round around it.

How UBE Marketing can help

UBE Marketing is a marketing agency in Sydney founded by Alvin Paronda, working with B2B technology companies on AI-assisted campaigns. Our AI marketing team helps B2B technology companies combine AI-assisted production with clear positioning and testing plans, so speed doesn't come at the cost of differentiation. If your LinkedIn creative needs a sharper testing rhythm, let's talk.