It’s a frustrating reality for business owners that you can spend an entire afternoon or more building Facebook and Instagram (Meta) campaigns and still attract the wrong people, while losing money in the process.
Meta needs varied creative (ad design) in order to learn what works across placements, and producing images, videos and copy takes real time. Meanwhile, AI promises speed, but generic outputs or ‘AI slop’ can leave your brand looking inconsistent or do more harm than good to your reputation.
The July 2026 Business Blueprint conference tackled both halves of that problem, with marketing specialists Loren Bartley and Stacey Hughes explaining how to combine Meta’s machine learning with human direction.
Their advice comes down to this: Give the platform clear goals, useful data and different ad creative and formats to try, then ‘train’ AI and let it support your efforts.
Stop Fighting the Machine and Feed It
Meta can now identify patterns across enormous numbers of actions, then show your ads to the people most likely to respond. So instead of relying on microscopic audience settings, give the system enough relevant material to test.
Supply it with variety:
- Image ads
- Carousels
- Short videos
- Face-to-camera clips
- Behind-the-scenes footage
- Voiceovers
One format may earn your first view from a new customer, while another reminds them to come back. You cannot predict the winning concept before launch, so starting with a variety of options gives the system options and stops your advertising becoming repetitive.
Choose the Objective You Actually Want
Meta is literal:
If you want enquiries, choose an objective that reflects this for your leads.
If you want purchases, choose a sales objective.
Boosting a post can certainly increase engagement, but remember that comments and views are not automatically commercial outcomes.
The objective you set with your ad tells the machine which action to optimise for, while the budget determines how widely it can search. A Meta Pixel, supported where appropriate by what’s referred to as the Conversions API, sends signals back about visits, clicks and purchases. Those results help the system recognise your buyers, then find more people who behave in a similar way.
Let the Message Help Target the Buyer
The look and feel of your ads is not decoration. As Stacey said, “The single biggest lever you can control is your creative and your ads.” Make the message specific enough that the intended buyer recognises it is meant for them.
Build the copy around three things:
- Person: Name who the offer serves.
- Problem: Describe what keeps them stuck.
- Promise: Show them the result that is available.
Relevance can beat a larger budget, because people respond much faster when an advertisement feels immediately useful to them.
If a campaign attracts the wrong engagement and doesn’t deliver the outcomes you were hoping for, review your message before assuming the targeting is broken. Then test your audience cues and compare broad delivery against audience suggestions rather than changing everything at once.
Train AI on Your Brand
AI works best when it understands your business. If you want your ads to look professional, give your AI platform information about your offers, your audience, the transformation you deliver, your customers’ beliefs, your brand voice and your proven examples. Develop a visual authority framework covering colours, imagery, style and the qualities the brand should convey.
For recurring characters or products, create visual references showing different angles, expressions and environments. Storyboards are another useful tool to define what happens from scene to scene. That guidance reduces inconsistency and makes a set of advertisements feel like a single recognisable campaign, even when every single execution is different.
Use a Workflow, Not a One-Click ‘Fantasy’
Creating effective ads takes time and strategy.
Begin with the tools you already have, adding specialised ones only when you need them: Develop a story arc, prepare your references, generate the short clips or images, then finish the options through editing, captions, music or voiceover.
AI doesn’t have to be perfect to be effective. The real test is if the creative gets the right person’s attention, explains the offer clearly and supports the campaign goal.
Read more: How To Explode Your Sales Using Free AI Marketing Tricks
Working With the Algorithm, Not Against It
The strongest system for Meta ads joins human judgement to machine scale. You define the buyer, the promise and the brand. AI speeds up production. Meta tests the delivery. Conversion data completes the loop, making each campaign more informed than the last.
To put this into practice, follow these six steps:
- Feed the algorithm variety to begin with.
- Choose the objective that matches the outcome you actually want.
- Build the message around the person, the problem and the promise.
- Train your AI on your brand before you ask it to create anything.
- Treat AI as one stage in a workflow. It’s never a one-click solution.
- Test in volume and let performance dictate the direction of your next round.
About Business Blueprint
The World’s #1 Business Program
Business Blueprint brings members the latest techniques and strategies to take control of their marketing, build better foundations and achieve the growth they are dreaming of.
If you’re tired of the business owner hamster wheel and sick of spending every day putting out spot fires, sign up for personal support, group accountability and access to the strategies that will maximise your profits and free you from the daily grind.
Membership includes:
- In-person conferences
- Weekly workshops (including AI news and training)
- Access to a library of over 1000 training videos (check out some samples here)
- Accountability groups
- One-on-one coaching calls
Want to know more? Book a call with a member of our team today.
