Organic social was never free, but for most brands it is now a weak route to new audiences. As Module 1 showed, recommendation-led feeds distribute content on predicted interest rather than on who follows whom, so organic posts reach mainly people already engaged with the account. That is valuable for community and for testing ideas, but unreliable for reaching a defined audience at a defined time. Paid social buys that reliability: delivery to people the brand chooses, in placements it chooses, as often as it chooses. In the paid, earned, shared and owned (PESO) model (Dietrich, 2014), paid media is the part a brand can switch on, scale and schedule.

What a brand should buy with it depends on how brands grow. Sharp (2010) argues that brands grow mainly by recruiting and refreshing large numbers of light buyers rather than by deepening loyalty among a few heavy ones, and that this depends on mental availability (being easily thought of in buying situations) and physical availability (being easy to find and buy). For paid social, reach among category buyers who do not yet know the brand matters more than repeated exposure to existing followers. Frequency still has a role, since one exposure in a fast-scrolling feed may not register, but Krugman (1972) argued that a few exposures can be enough for an advertisement to do its work, which is a reason to cap frequency rather than maximise it. The planning question is the media planner's: how many of the right people, how often, over what period.

Binet and Field (2013) add the distinction between long-term brand-building and short-term sales activation, introduced in Module 1. Paid social can do either, but platforms make activation easy to buy and easy to report, so budgets drift towards it, and a plan made only of conversion campaigns aimed at warm audiences tends to harvest demand created elsewhere rather than create it.

Paid social is the wrong lever when the product, landing page or checkout cannot convert the attention it buys; when the job needs someone else's credibility, the influencers' role set in Module 4; when creative has not yet been shown to stop the right people; and when the margin cannot support the cost of acquiring a customer, a calculation Lesson 6.3 makes explicit. Amplifying a weak idea only makes it fail faster.

For Moorfield, organic Instagram and TikTok reach mainly existing followers, but the FY2027 goal is to grow new customers from 48,000 to 60,000. Paid social, with £260,000 of the £720,000 digital budget, is the line that can reach people with reactive skin who have never heard of the brand.

Every campaign starts with an objective: the advertiser tells the platform what to optimise for (reach, video views, traffic, leads or purchases), and the delivery system seeks the people most likely to take that action. Purchase optimisation finds people likely to buy, some of whom would have bought anyway, so platform-reported conversions can overstate what advertising caused. Gordon et al. (2019), comparing large randomised experiments at Facebook with common observational methods, found the latter often misestimated the true effect, sometimes widely. Choose the optimisation event that matches the job and expect conversion-optimised reporting to flatter itself; Module 7 covers the tests.

The signals behind audience strategy have weakened. Apple's App Tracking Transparency, introduced with iOS 14.5 on 26 April 2021, requires apps to ask permission before tracking users across other companies' apps and websites. Where users decline, platforms lose data behind detailed targeting, retargeting, lookalike audiences and conversion reporting, and lean on modelled results and data advertisers send from their own servers. Targeted audiences shrink and measurement becomes less certain; size the effect only from a brand's own data.

The platforms' answer is automation: broad targeting, often only country, age and exclusions, letting the system find buyers. As at 2025–26, Meta's automated suite is branded Advantage+ (Advantage+ shopping campaigns became Advantage+ sales campaigns in early 2025) and TikTok's is Smart+; names change often. Broad delivery trades control and transparency for scale and cost, putting the weight of targeting on creative.

In a broad audience the ad itself decides who stops scrolling. A hook that names a reactive-skin problem in the viewer's own words selects people with that problem; a generic beauty shot selects no one. Distinct hooks built on the pillars from Module 3 do the segmentation that targeting settings used to do. Specificity should match the decision stage: Lambrecht and Tucker (2013) found dynamic retargeted ads showing products people had browsed were on average less effective than generic ads, unless browsing showed that preferences had narrowed.

Test with structure: a hypothesis, one changed variable (hook, format, message or creator), a fair budget and time per variant, and a decision rule set in advance, because early results are noisy and delivery systems shift spend to early winners.

Creator amplification runs creators' content as ads: Meta's partnership ads run from the creator's handle, or both identities, labelled "Paid partnership"; TikTok's Spark Ads boost an organic post through an authorisation code the creator generates for a limited period. Both need the creator's permission and a contract granting paid usage rights (Module 5): the platform switch is not the licence, and amplified content is still advertising that must be disclosed. Moorfield's 90 days of paid usage covers the six-week launch.

Every paid social budget rests on a chain of assumptions; making it explicit is the best defence against plans that cannot work. Spend buys impressions at a cost per thousand (CPM): impressions = budget ÷ CPM × 1,000. A share of impressions become clicks (click-through rate, CTR), a share of clicks become orders (conversion rate) and each order is worth the average order value (AOV). These yield revenue, return on ad spend (ROAS = revenue ÷ spend) and cost per acquisition (CPA = spend ÷ orders).

ROAS leads most dashboards and is the most often misread, because it ignores what it costs to make and deliver the product. The business needs to know whether the media paid for itself out of margin. Gross profit is revenue × gross margin; contribution after media cost is gross profit minus spend. Two break-even points follow. Break-even ROAS is 1 ÷ gross margin: at a 68% margin, about 1.47x. Break-even CPA is AOV × gross margin: at £42 and 68%, about £28.56. A ROAS of 1.3x is "positive", since revenue exceeds spend, yet at a 68% margin it loses money on every order. If the margin figure excludes fulfilment, payment fees and returns, true contribution is lower still.

The assumptions are fragile. CPMs vary by platform, placement, audience, season and competition; click-through and conversion rates with creative, offer, landing page and product. Published benchmarks differ widely by source, sample and date: use them as a sense check, not a forecast, and prefer your own data and small tests. The model is also linear, while campaigns are not: as spend rises, delivery reaches less responsive people and frequency climbs, so returns usually diminish. And platform-attributed orders are not orders the advertising caused, a gap Module 7 quantifies.

Sensitivity analysis shows which assumption the decision depends on: in a tightly balanced model, a small change in one rate turns profit into loss. Change one input at a time and note how far each must move to reach break-even; the inputs with the smallest margin for error are the ones to test first and to monitor closely in flight.

First-order profit is not the whole case. An acquisition campaign can justify a thin or negative first-order contribution if the customers it recruits come back, but that is a claim about future behaviour, to be stated as an assumption and tested rather than used to excuse any result. Binet and Field (2013) also remind planners that brand-building effects arrive over longer periods than a campaign report covers.

Amara's model for the £60,000 launch budget shows all of this in one line: a ROAS of 1.58x, "positive", leaves just £4,260 of contribution after media cost.

A campaign is integrated when its channels work as one plan rather than run in parallel. Schultz, Tannenbaum and Lauterborn (1993) argued that communication should be planned from the customer outwards, with every contact coordinated, rather than bought channel by channel, an early statement of integrated marketing communications (IMC). Integration is often mistaken for consistency; shared cues do build memory, but the stronger idea is complementarity: each channel does the job it is best at, and the channels reinforce one another. Batra and Keller (2016), reviewing how traditional and digital media combine to affect consumer decisions, argue that plans should be designed around where consumers are in their decision journey and how media effects interact.

Practical integration rests on four shared elements: one idea, which every channel expresses natively; one audience definition; one calendar, so activity is sequenced rather than colliding; and one measurement plan, agreed before launch, so that each channel is judged on its role and the same sale is not counted several times by different platforms.

Two frameworks help assign roles. Chaffey's RACE framework (Chaffey, Ellis-Chadwick and Abed-Rabbo, 2025) organises digital activity around the customer lifecycle: Reach (build awareness and visits), Act (prompt interaction and leads), Convert (turn interest into sales, online or offline) and Engage (build long-term loyalty), often preceded by a Plan stage. Kaushik's (2013) See-Think-Do framework groups audiences by intent: See is the largest addressable audience; Think, those with some commercial intent; Do, those ready to buy now; a later fourth stage, Care, covers existing customers. RACE suits planning objectives by stage; See-Think-Do-Care is a sharper check that content and targeting match intent, and it exposes plans that spend everything on Do.

In a social-led campaign, organic social sustains community and tests ideas cheaply. Influencers supply credibility and native content the brand cannot make for itself. Paid social buys reach at scale and amplifies proven content. Search, especially paid search on brand and problem terms, captures demand the other channels create; it rarely creates it. Email reaches people who have given permission, cheaply. Owned community and the website hold the relationship and the sale. This module treats search and email only in these roles.

Integration fails in known ways: silos with separate budgets and targets, email going out before the creators have posted, platforms each claiming the same order, and messages drifting as each team adapts them. The cure is as much organisational as creative: one owner of the plan and one channel-role map everyone works from.

For Moorfield, the idea is "Strong Skin Season", the audience is people with sensitive and reactive skin, and 1 March to 11 April 2027 is the shared calendar; the worked example shows Amara's channel-role map.

Plans fail in execution more often than conception: assets arrive late, claims go unchecked, budget sits in a weak ad because no one knew who could move it, and a comment thread goes wrong on a Saturday. An execution plan settles, before launch, who does what, by when, and what happens when things change.

The launch plan works backwards from go-live, listing every asset and task with an owner and deadline, and identifies the critical path: the dependent tasks that set the earliest possible launch. In social and influencer campaigns it usually runs through creator content, claims approval and platform ad review, all slower than expected, so build buffer before launch.

A RACI matrix names, for each task, who is Responsible (does the work), Accountable (owns the outcome and signs off; one person only), Consulted before a decision and Informed after it, exposing two common faults: tasks with no accountable owner, and tasks with so many people consulted that nothing moves.

In skincare, every ad, caption and creator script describing what a product does must be checked against the claims the brand can substantiate, and paid ads face the same advertising codes as any other ad. Agree which claims are pre-approved, who signs off anything new and the turnaround time, so approval is a scheduled step, not a bottleneck.

In flight, a campaign needs a small dashboard and pre-agreed rules. The dashboard tracks leading indicators for each channel's role (reach and frequency for paid social, creator posts live and compliant, branded search volume, email clicks, landing-page conversion) against the plan's assumptions. Optimisation rules set in advance what triggers a change (a creative losing budget, spend shifting between platforms, creative rotating as frequency rises), guarding against two opposite errors: reacting to daily noise, especially during delivery systems' initial learning period, and leaving money in something clearly failing. Base them on the economics of Lesson 6.3, such as CPA against break-even CPA, remembering that platform attribution overstates causal effects.

A risk register lists each risk with its likelihood, impact, owner, preventive action and response. Typical launch risks include a creator post that breaks disclosure or claims rules, rejected ads, stock-outs, site failure under traffic, and criticism or misinformation in comments. Community management is part of execution: a launch brings questions, complaints and claims into comments, so the team needs staffed hours, escalation routes and approved answers, including for anyone asking whether a product treats a medical condition. The moderation principles in Module 2 apply with extra force while ads drive traffic.

For Moorfield, the partnerships audit found two gaps, no process for checking micro creators' Stories and no written affiliate-code policy, and both belong on the launch risk register until closed.