Meta Advantage+, TikTok Smart+ and LinkedIn's AI campaigns

0:00 Social platform automation

If you've built a Meta or TikTok campaign lately, you've noticed the old menus are gone or buried. Every social platform has moved the same way Google did. In this lecture you'll get a practical map of Meta Advantage plus, TikTok Smart plus and Symphony, and LinkedIn's Accelerate and predictive audiences, and you'll learn the one lever that matters most on each.

0:27 Why it matters

Why does this matter? Because the biggest waste in social advertising today comes from teams using automated campaigns with manual-era habits. Here's an analogy. It's like hiring a brilliant personal shopper and then handing them a list that says only visit these three shops, only buy blue, and only on Tuesdays. You've paid for intelligence and then removed its room to work. Automated social campaigns need room to explore, but they also need to know exactly what you're shopping for. So the craft is giving them freedom on the who, and precision on the what: the conversion goal, the value, and the creative.

1:12 Meta Advantage+

Start with Meta. Advantage plus shopping campaigns were renamed Advantage plus sales campaigns, and the same automated approach now covers app and leads objectives. In 2026 Meta made Advantage plus the default starting point for new campaigns. The key thing to understand is Advantage plus audience. Your interests and lookalikes become suggestions. The system can go beyond them. Only a few settings, like location and minimum age, act as firm controls.

1:43 What wins on Meta

So what wins on Meta now? Few campaigns, few ad sets, and many genuinely different creatives. Send conversions through both the pixel and the Conversions API, sharing an event ID so Meta can remove duplicates. If your purchase values are reliable, use value optimization. And define your existing customers with a custom audience, so you can see how much budget is going to people who already buy from you.

2:13 TikTok

Now TikTok. Smart plus is TikTok's automated buying: targeting, bidding, placements and budget. You can choose module by module how much to automate. Symphony is the creative side, generative tools for scripts, video editing, avatars, translation and dubbing, and TikTok has been building Symphony into the Smart plus flow. Two cautions. Native vertical video still beats a repurposed TV ad. And TikTok requires realistic AI-generated content to be labeled, so check the current policy before you use avatars or synthetic voices.

2:48 Simple example: Karachi streetwear (illustrative)

Let's do a simple example on TikTok. A Karachi streetwear brand runs Smart plus with two video concepts: a studio product showcase and a founder walking through the market explaining how the fabric is sourced. They let Smart plus handle targeting and bidding and simply watch results for ten days. The founder video, shot on a phone, gets far cheaper purchases. So next week they make three more founder-style videos with different hooks: the fabric, the stitching, and the price. The lesson isn't that founder videos always win. It's that the automated campaign found the audience for the native-feeling concept, and the team fed it more of what worked.

3:35 LinkedIn

LinkedIn is where automation meets long sales cycles. Accelerate builds a campaign from your website and goal. Predictive audiences model your best customers from your own data, like lead-form submissions or contact lists. BrandLink places video next to publisher and creator content. But here's the catch. LinkedIn conversions are rare and slow. If you only send form fills, the system will find you students and job seekers. Send qualified pipeline stages back through the LinkedIn Conversions API.

4:08 Example: London B2B SaaS (illustrative)

Here's an illustrative example. A London HR software startup ran LinkedIn Accelerate for demo requests, seeded predictive audiences with closed-won accounts, and sent sales-qualified-lead and opportunity events from the CRM. On Meta they tested an Advantage plus leads campaign with qualifying questions, also feeding CRM outcomes back. The breakthrough wasn't a clever audience. It was sending qualified stages back to every platform, so each model learned what a real customer looks like.

4:39 The three common mistakes

Let's talk about the mistakes I see most often when teams move into these automated campaigns. The first is structure. They bring across twelve ad sets split by age and interest, each with a tiny budget, so no ad set ever gets enough conversions to learn. The second is creative. They load ten versions of the same image with different button colors, and wonder why the system can't find new people. The third is feedback. They optimize for leads or add to carts without telling the platform which of those became revenue. Fix those three, structure, creative diversity and qualified feedback, and most automated campaigns improve before you touch a single bid.

5:28 Everyone else

A quick word on everyone else. Snap, Pinterest, Reddit, Amazon Ads and Microsoft Advertising all have automated campaign types and conversion APIs too. The rules transfer. Consolidate so models get data. Feed clean, deduplicated, qualified conversions. Diversify creative. And verify with an incrementality test before you scale. The most common failure is copying an old micro-segmented structure into an automated campaign and wondering why it never exits learning.

5:58 Mistakes + try this now

The common mistakes on social automation are predictable. Stuffing Advantage plus with dozens of interests, thinking they're restrictions. Launching Smart plus with one video and judging the whole approach on it. On LinkedIn, optimizing for form fills without feeding back which leads were qualified. And duplicating audiences across many campaigns so they bid against each other. Try this now: open your ad account and count how many active campaigns target the same country with the same objective. If it's more than two or three, write a consolidation plan: which campaigns merge, which creatives move, and which one conversion event they'll all optimize toward.

6:43 Watch me do it: LinkedIn setup (illustrative)

Watch me do it. I'll set up an illustrative LinkedIn campaign for a Riyadh cybersecurity consultancy so you can see the order of decisions. First, the conversion. Before building anything, I create two conversion rules: demo booked from the Insight Tag, and sales qualified lead fed from the CRM through the Conversions API. I set the SQL value using the historical close rate. Second, the audience. I upload a list of closed-won accounts to seed a predictive audience, and I add exclusions for current customers and competitors. Third, the campaign. I try Accelerate with the website URL and the goal of qualified leads, then read every suggestion it makes. It proposes a job-title audience that includes students; I remove that. Fourth, creative: four concepts, one for the security manager, one for the finance director worried about fines under Saudi rules, one for the CEO, one proof-led case study, each labeled with persona and motivation. Fifth, the check-in rule: I won't judge the campaign until it has either two weeks or a meaningful number of SQLs. Notice that targeting took two minutes. The conversion design took longest, and it matters most.

8:06 Recap and next step

Recap. Meta Advantage plus is the default, with audience inputs as suggestions. TikTok Smart plus automates buying while Symphony handles creative, with AI labeling rules. LinkedIn's automation needs qualified CRM feedback. Your next step: pick one platform, and write a one-page setup brief listing the campaign type, the hard controls you'll use, your creative plan, and the server-side events you'll send.