Chapter 6: Paid Social Advertising: Meta, LinkedIn, and TikTok
By the end of this chapter you will be able to:
Rupak NepaliPublished Nov 10, 2026
Learning objectives
By the end of this chapter, you will be able to:
- Explain how paid social differs from paid search, and when to use each.
- Set up tracking that paid social algorithms depend on, including server-side conversion APIs.
- Plan, launch, and optimize campaigns on Meta (Facebook and Instagram), LinkedIn, and TikTok.
- Build a creative testing system that feeds AI-driven ad delivery.
- Read paid social results honestly and judge true incremental impact.
- Use AI assistants to generate concepts, scripts, and analysis for paid social.
Paid search versus paid social
Paid search (Chapter 5) captures existing demand: people are already searching for what you sell. Paid social creates demand: it puts your product in front of people who weren’t looking but are likely to care. That difference shapes everything.
| Paid search | Paid social | |
|---|---|---|
| User mindset | Actively looking for a solution | Browsing, being entertained, connecting |
| Targeting basis | Keywords and intent | Interests, behaviors, lookalikes, and increasingly algorithmic prediction |
| Creative | Text-led, matching the query | Visual and video-led, stopping the scroll |
| Main lever | Keywords, bids, relevance | Creative and audience signals |
| Typical role | Conversion and demand capture | Awareness, consideration, demand creation, retargeting |
Most growing brands need both. Paid social fills the top of the funnel with new people; paid search and email convert them when they’re ready.
The foundation: tracking and signals
Modern paid social platforms rely on machine learning to decide who sees your ads. The algorithm learns from the conversion events you send it. If those events are missing, delayed, or wrong, the algorithm optimizes toward the wrong people.
Set up before spending a dollar:
- Platform pixel or tag on every page: the Meta Pixel, LinkedIn Insight Tag, TikTok Pixel.
- Server-side events: Meta’s Conversions API (CAPI), TikTok’s Events API, and LinkedIn’s Conversions API send conversion data directly from your server or e-commerce platform. They recover events lost to browser restrictions and ad blockers, and they improve match quality. Many e-commerce platforms (such as Shopify) and tag managers offer simple integrations.
- Standard events with values: ViewContent, AddToCart, InitiateCheckout, Purchase (with value and currency), Lead, CompleteRegistration.
- Deduplication so the same conversion from pixel and server isn’t counted twice (using a shared event ID).
- Domain verification and event configuration as required by each platform.
- Consent: only send data in line with users’ consent choices and your privacy policy (Chapter 9).
- UTMs on every ad so GA4 can report paid social traffic independently (Chapter 7).
Prompt 6.1: Paid social tracking checklist Act as a paid social technical specialist. Our website runs on [platform] and we advertise on [Meta / LinkedIn / TikTok]. Create a step-by-step tracking setup checklist covering pixels or tags, server-side conversion APIs, standard events and parameters, deduplication, domain verification, consent handling, and UTM conventions. Include how to test that each event is firing correctly, and the most common setup mistakes.
Toolkit: The Paid Social Tracking checklist covers pixels, conversion APIs, event deduplication, consent, and testing for Meta, TikTok, and LinkedIn.
Meta: Facebook and Instagram ads
Meta’s platforms reach a vast share of internet users and offer some of the most sophisticated ad delivery systems available. For most consumer brands, Meta is the first paid social platform to master.
Campaign structure
Meta uses three levels: campaign (objective and, optionally, budget), ad set (audience, placements, schedule, optimization event, and budget if not set at campaign level), and ad (creative and copy).
Objectives: Awareness, Traffic, Engagement, Leads, App Promotion, and Sales. Choose the objective that matches the business outcome. For an online store, that’s almost always Sales optimized for purchases.
The shift to automation and broad targeting
Meta has steadily moved advertisers toward automated, AI-driven campaigns:
- Advantage+ campaigns (for example, Advantage+ sales campaigns, formerly called Advantage+ shopping campaigns) automate audience, placement, and budget allocation using your creative and conversion data.
- Broad targeting: rather than narrowly defined interests, many advertisers now give Meta a broad audience (often just country, and sometimes age) and let the algorithm find buyers. Your creative tells the system who the ad is for.
- Advantage+ placements: delivering across Facebook, Instagram, Messenger, Threads, and the Audience Network automatically.
- Advantage+ creative: automatic enhancements such as cropping, brightness adjustments, music, and AI-generated text variations. Review these carefully for brand fit; turn off enhancements you don’t want.
Meta has described improvements to its ad retrieval and ranking systems that make creative diversity more important than ever: genuinely different concepts help the system find different pockets of customers.
A simple, effective Meta account structure
For a small to mid-sized e-commerce brand like Brightleaf:
- Prospecting campaign (Sales objective, broad audience or Advantage+): most of the budget, with several distinct creative concepts.
- Retargeting (optional, as Advantage+ often handles this automatically): site visitors and cart abandoners who haven’t purchased, with offer- and objection-focused creative.
- Testing campaign: a fixed portion of budget to test new concepts before promoting winners into the main campaign.
Avoid fragmenting budget across many small ad sets. The algorithm needs enough conversions per ad set to learn; Meta’s guidance has long been that ad sets exit the “learning phase” after roughly 50 optimization events within a week. Fewer, better-funded ad sets usually outperform many starved ones.
Meta formats
- Single image and single video (vertical 9:16 and 4:5 perform well in feeds, Stories, and Reels).
- Carousel (multiple products or steps).
- Collection and Instant Experience (immersive mobile shopping).
- Catalog ads (dynamic product ads from your product feed, ideal for retargeting).
- Partnership ads (running ads through a creator’s handle, with permission).
- Lead ads (forms inside Facebook and Instagram).
LinkedIn ads
LinkedIn is the leading paid social platform for B2B marketing because it lets you target by professional attributes no other platform offers as precisely: job title, function, seniority, company, company size, industry, skills, and more. That precision comes at a price: LinkedIn CPCs and CPMs are typically much higher than on Meta or TikTok, so it suits businesses where a customer is worth a lot.
When LinkedIn makes sense
- B2B products and services with a high lifetime value.
- Recruiting and employer branding.
- Professional education and events.
- Account-based marketing (ABM): targeting specific named companies.
Targeting
- Company lists (upload target accounts for ABM).
- Job attributes: titles, functions, seniority, skills.
- Company attributes: industry, size, growth.
- Matched audiences: website retargeting, contact lists, and people who engaged with your ads or page.
- Predictive audiences that use LinkedIn’s AI to find people similar to your best leads or customers.
Avoid audiences that are too narrow. A few thousand people may be fine for ABM, but most campaigns need tens of thousands to deliver efficiently.
Formats
- Single image and video ads in the feed.
- Document ads that let people read a PDF guide in the feed, with optional lead capture.
- Thought Leader ads that promote posts from employees or executives, often outperforming company-page ads because people engage more with people.
- Lead Gen Forms prefilled with LinkedIn profile data, which reduces friction dramatically.
- Conversation and message ads (use sparingly; they can feel intrusive).
- Event ads for webinars and conferences.
What works on LinkedIn
Valuable, specific, non-salesy content: original research, practical guides, benchmarks, webinars, and executive perspectives. Offer something worth a professional’s time, capture the lead, and nurture it with email (Chapter 13) and sales follow-up. Measure beyond cost per lead: track lead quality, pipeline created, and revenue.
Figure 6.2 [SCREENSHOT TO ADD]: LinkedIn Campaign Manager targeting. Capture audience targeting by job function, seniority, and company size, with the forecasted audience size panel visible.
TikTok ads
TikTok is an entertainment-first platform where discovery happens through its recommendation feed. Ads that look and feel like native TikTok content (creator-led, fast, authentic, sound-on) dramatically outperform polished TV-style ads.
Options
- In-feed ads: standard video ads in the For You feed.
- Spark Ads: boosting organic posts from your account or from creators (with their authorization), keeping the social proof of likes and comments.
- Automated campaigns (such as TikTok’s Smart+ campaigns) that automate targeting, bidding, and creative optimization.
- TikTok Shop ads and features for in-app shopping in supported markets.
- Creative tools: TikTok’s Creative Center (trends, top ads, sounds) and AI creative tools for scripts, editing, and translations.
TikTok creative principles
- Hook in the first one to two seconds.
- Use real people, preferably creators who match your audience.
- Show the product in use, with a clear benefit or transformation.
- Use trending formats and sounds (with appropriate commercial licenses).
- Add captions and on-screen text.
- Produce many variations; creative fatigue sets in faster on TikTok than on most platforms.
The creative testing system
On every platform, creative is now the biggest performance lever. Build a repeatable system.
1. Think in concepts, not tweaks
A concept is a distinct idea: an angle and a message aimed at a specific motivation. A variation changes execution details (hook, thumbnail, length, copy). Test concepts first, then optimize winners with variations.
Concept angles for Brightleaf:
- Problem and solution: “Your supermarket coffee was roasted months ago.”
- Demonstration: a pour-over in slow motion, with tasting notes on screen.
- Social proof: five customers react to their first bag.
- Founder story: why the roaster started Brightleaf.
- Comparison: Brightleaf versus a typical subscription.
- Offer: first bag 50% off, cancel anytime.
- Ritual and identity: “For people who take their morning seriously.”
2. Build a creative matrix
Map concepts against audience motivations (taste, convenience, ethics, value, gifting) and formats (UGC video, static, carousel). Fill gaps systematically rather than guessing.
3. Test with a disciplined process
- Launch three to five concepts at once in a testing campaign with equal opportunity to spend.
- Judge on the primary metric (cost per purchase or cost per qualified lead), with supporting diagnostics: thumb-stop rate (3-second views divided by impressions), hold rate, CTR, and conversion rate.
- Give each test enough spend and conversions to reach a fair conclusion.
- Promote winners into the main campaign; iterate on them with new hooks and variations.
- Log every test and its learning (Appendix D has a test log template).
4. Watch for creative fatigue
Rising frequency, falling CTR, and rising cost per result signal fatigue. Keep a pipeline of new concepts so you can refresh before performance drops. Chapter 11 shows how AI creative production makes this pipeline sustainable.
Prompt 6.2: Creative matrix Act as a performance creative strategist. For [product] sold to [audience], build a creative matrix: rows are 6 concept angles (problem/solution, demonstration, social proof, founder story, comparison, offer), columns are 4 audience motivations [list]. For each cell with potential, write a one-line concept, a 2-second hook, and the best format (UGC video, static, carousel). Then recommend the first 5 concepts to test and explain why.
Prompt 6.3: Platform-native scripts Turn this concept into three platform-native ads: a 20-second TikTok creator video script (hook, beats, on-screen text, sound suggestion), a Meta Reels ad (9:16, 15 seconds, captions, primary text under 125 characters, headline under 40 characters), and a LinkedIn Thought Leader post from our founder (120 to 180 words). Concept: [describe]. Facts we can use: [list]. Do not invent claims.
Budgets, bidding, and scaling
- Start with the economics (Chapter 1): know your affordable cost per acquisition.
- Budget for learning: give campaigns enough daily budget to generate meaningful conversions. A common rule of thumb is a daily budget of several times your target CPA per ad set.
- Bid strategies: most advertisers start with the platform’s default “highest volume” or lowest-cost bidding, then use cost caps or ROAS goals once they understand performance.
- Scale gradually: increase budgets in steps (for example, 20% to 30% every few days) rather than doubling overnight, which can reset learning and spike costs.
- Scale horizontally: add new winning concepts rather than only raising budgets on old ones.
Measuring paid social honestly
Paid social is where platform reporting and reality diverge most:
- Attribution windows: Meta’s default reports conversions within 7 days of a click and 1 day of a view. View-through conversions can include people who would have bought anyway.
- Platform versus GA4 versus back end: platform totals often exceed GA4 and actual orders. Know why, and choose one source of truth for decisions.
- New versus returning customers: check what share of attributed purchases came from existing customers. Meta and other platforms offer settings or reporting to focus on new customers.
- Incrementality: the gold standard is a lift test. Meta offers Conversion Lift studies; you can also run geographic holdout tests (Chapter 9). Even one well-run lift test per year recalibrates how much to trust platform numbers.
- Blended metrics: track blended CAC and MER (marketing efficiency ratio: total revenue divided by total marketing spend) to see the overall effect of all channels together.
Prompt 6.4: Paid social performance review Here is our paid social data for the last 30 days by campaign, ad set, and ad: spend, impressions, frequency, 3-second views, CTR, clicks, purchases, cost per purchase, ROAS, and new-customer share: [paste]. Identify winning and fatigued creatives, budget allocation problems, and audience saturation signs. Compare platform-reported purchases with this GA4 and back-end data: [paste], and explain likely reasons for differences. Recommend five actions for next week, separating evidence from inference.
Hands-on exercises
Exercise 6.1: Ad library teardown (30 minutes). Use the Meta Ad Library, LinkedIn’s ad library on company pages, and TikTok’s Top Ads in the Creative Center to collect five ads from one competitor or category on each platform. Classify each by concept angle and format.
Exercise 6.2: Creative matrix (45 minutes). Use Prompt 6.2 for Brightleaf or your business, then choose and refine the five concepts you’d test first.
Exercise 6.3: Tracking audit (30 minutes). Use the Meta Pixel Helper or TikTok Pixel Helper browser extension on a few e-commerce sites. Which events fire? Do any sites appear to be missing purchase tracking?
Strategic assignment: paid social launch plan
Deliver a Paid Social Launch Plan for one consumer platform (Meta or TikTok) and one B2B or secondary platform (LinkedIn, or the other consumer platform): tracking setup checklist; campaign structure and budgets based on unit economics; a creative matrix and scripts for five concepts; a 30-day testing plan with success criteria; a measurement approach including blended metrics and a lift test proposal; and an AI log.
Key takeaways
- Paid search captures demand; paid social creates it. Most brands need both.
- Tracking quality, including server-side conversion APIs, determines how well the algorithms perform.
- Meta rewards broad targeting, consolidated structures, and creative diversity; LinkedIn offers unmatched B2B targeting at a premium; TikTok rewards native, creator-led video.
- Creative is the main lever. Test distinct concepts, then iterate on winners, and refresh before fatigue sets in.
- Treat platform-reported results with skepticism. Use blended metrics and lift tests to find true impact.
Author of four Opencart book.
The recent are Opencart 4 developer book and Opencart 4 user manual
