Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Meta Ads › Advanced

Capstone: A Meta Ads Scaling Strategy for a Bedding Company

The closing project of the advanced Meta Ads level. You write the scaling strategy for an invented online bedding company whose account is profitable at modest spend but rests on weak signals, untested remarketing and irregular creative, covering the signal plan, a holdout test, the creative pipeline, a budget path on marginal returns and the guardrails that stop scaling when it stops paying.
Capstone projectAdvanced levelModule 3: Creative systems and cross-platform frequency6 minUpdated

Brief

Eiderholm Bedding Company is an invented online seller of duvets, pillows and bed linen, with a small trade service supplying guest houses. Its Meta account is profitable at modest spend and investors want growth. The owner suspects the account is less healthy than its reports suggest: event matching is weak, most of the budget sits in remarketing, creative is refreshed when someone remembers, and nobody has ever tested what Meta adds beyond demand that already exists.

The owner wants a plan that grows new customers without betting the business on reported return. One marketing manager runs the account with a freelance designer; the owner signs off budget changes. The brief assumes the audit and scaling mechanics of the intermediate Meta Ads level; this capstone asks for the strategy those mechanics serve.

Inputs and constraints

Illustrative example — not a real result

Eiderholm Bedding: account structure and results

Average month over the last quarter, as reported by Meta:

CampaignMonthly spendPurchasesRevenueView-through share of purchasesAttribution setting
Remarketing: visitors and baskets, catalogue ads£13,200610£80,52048%7-day click, 1-day view
Prospecting: broad and lookalike, 3 ad sets£7,000150£16,80022%1-day click
Brand video, reach objective£1,80020£2,400n/an/a
  • Two prospecting ad sets are limited by too few conversions to exit learning.
  • Remarketing does not exclude purchasers from the last 30 days.
  • 3 concepts live; the newest is 10 weeks old. No video beyond the brand film.
Illustrative example — not a real result

Eiderholm Bedding: signals, history, economics and limits

  • Purchase event: pixel and server plugin share an event ID. Rating low. Server coverage: email 61% (guest checkouts omitted by the plugin), phone 0%, click ID 3%.
  • Trade enquiries from guest houses arrive by form; the CRM records qualified accounts and first trade orders. None of it reaches Meta.
QuarterMeta spend a monthShop orders a monthNew customers a month
Q1£16,0002,0501,020
Q2£19,0002,1101,040
Q3£22,0002,1401,050
  • Contribution margin after product, delivery and payment fees: 48% of revenue. Returns refund about 10% of revenue (bed linen higher, pillows lower), leaving about 40% after returns.
  • Owner's target ROAS: 3.0 on reported revenue, window not stated.
  • The owner will risk up to £6,000 over the two quarters on tests.
  • The plan period includes Black Friday and the January sale.

The figures are invented to make the brief concrete; the reasoning is what is assessed.

Task

  1. Fix the account before scaling it. Run the Meta Ads account audit checklist against the inputs and score the result with the account audit scorecard. List the structural fixes that must precede any budget increase, including the purchaser exclusion, the starved prospecting ad sets and the mixed attribution settings.
  2. Write the signal and attribution plan. Order the event matching fixes for the Conversions API integration, with verification for each. Plan which trade outcomes the CRM sends back and which one the trade campaign would optimise for, as in the jewellery attribution exercise. Choose one attribution setting for delivery across purchase campaigns, with reasons from the attribution settings guide, and state the window the owner's target is judged under.
  3. Calculate the economics. Break-even and target return after returns, the return of each campaign under a consistent view, and the marginal cost per extra order and per extra new customer for each spend step in the history. Use the break-even calculator and the ROAS calculator to check your figures.
  4. Design the holdout. Test the largest spend line with the remarketing holdout test: holdout share, exclusions across every campaign that could reach the group, primary metric from the shop's own orders, guardrails, window and timing around the peaks, and the cost of the test against the owner's risk budget. Write it up in the experiment hypothesis template with the decision rule for each outcome.
  5. Set the creative pipeline. Following the paid social account management playbook and the activewear pipeline exercise: concepts needed a month, briefs from the creative brief template, rotation, fatigue triggers, a Frequency guardrail for remarketing and a learning log.
  6. Plan the budget path. Quarter by quarter: what moves from remarketing to prospecting depending on the test, the size and spacing of budget steps from the scaling campaigns guide, the marginal limit each step is judged against using the marginal returns and forecasting guide, and a forecast given as a range for each scenario.
  7. Set guardrails and reviews. Stop rules for any step, what happens during peaks, and the weekly, monthly and quarterly reviews with owners and the evidence each one reads.
  8. Write the strategy. A one-page summary for the owner and investors, with appendices for the audit, the test plan, the pipeline and the budget path.

Evaluation criteria

CriterionWhat a strong answer showsCommon gap
SignalsRaises match quality and feeds CRM outcomes back, with a reasoned attribution settingChanges the attribution setting to flatter reports
IncrementalityDesigns a holdout for the largest spend with a decision ruleScales on reported return alone
Creative systemSets a pipeline with briefs, rotation, fatigue triggers and frequency limitsRelies on one winning ad
Budget pathScales in steps against marginal return with stop rules and reviewsDoubles budget in one move
Account healthUses the audit checklist to fix structure before scalingScales a broken structure

A strong submission also notices what the history already says. Orders and new customers rose much less than spend in the last step, and the remarketing line's reported ROAS leans heavily on view-through credit among people who had already visited. Neither proves remarketing is wasted; together they are the case for testing Incrementality before moving money, and for expecting the budget path to shift spend towards prospecting rather than simply adding to the total.

Illustrative example — not a real result

Eiderholm Bedding: the first figures a strong answer reaches

  • Break-even ROAS after returns: 1 ÷ 0.40 = 2.5. A target of 3.0 implies keeping about 6.7% of revenue as profit after ad cost (0.40 − 1 ÷ 3.0).
  • Spend step Q1 to Q2: £3,000 for 60 more orders (£50 each) and 20 more new customers (£150 each).
  • Spend step Q2 to Q3: £3,000 for 30 more orders (£100 each) and 10 more new customers (£300 each).
  • Remarketing reported ROAS: £80,520 ÷ £13,200 = 6.1, with nearly half of its purchases credited to views.

Submission checklist

0/11

Pages this lesson applies

Templates and checklists used

Terms used on this page

Also related

Related