Brief
Lumbrook Lighting is an invented online retailer of ceiling pendants, lamps, large statement fittings and bulbs. It grew on Google Shopping, Meta and a storefront on a large marketplace, and the owner wants to double paid media within a year. Every dashboard reports a healthy return. The owner's doubts are specific: margin differs sharply by range, large fittings come back far more often than anything else, the marketplace takes a commission the dashboards ignore, and nobody knows which channel's sales would have happened anyway.
The plan assumes the cross-channel planning and target arithmetic of the intermediate performance marketing level. This capstone asks for the strategy that decides how far each channel may grow and what evidence each step needs.
Inputs and constraints
Lumbrook Lighting: an average month by channel
| Channel | Monthly spend | Revenue the channel reports | Reported return |
|---|---|---|---|
| Google Shopping, which also serves on brand queries | £18,000 | £108,000 | 6.0 |
| Google Search, brand | £3,000 | £45,000 | 15.0 |
| Meta prospecting and remarketing | £12,000 | £54,000 | 4.5 |
| Marketplace sponsored listings | £7,000 | £35,000 | 5.0 |
| Total | £40,000 | £242,000 |
Revenue actually recorded, every source: own site £160,000, marketplace £30,000.
Lumbrook Lighting: ranges, tracking, team and limits
| Range | Share of revenue | Contribution before returns | Return rate | Contribution after returns |
|---|---|---|---|---|
| Pendants and ceiling lights | 40% | 42% | 6% | 38% |
| Wall and table lamps | 25% | 45% | 5% | 41% |
| Large fittings and floor lamps | 20% | 34% | 18% | 20% |
| Bulbs and smart lighting | 15% | 22% | 2% | 21% |
- The marketplace charges 15% commission on its orders, on top of the costs above.
- Repeat purchase: 22% of new customers buy again within twelve months, mostly bulbs.
- Tracking: Google tag purchase event without enhanced conversions; Meta pixel only; the marketplace reports its own sales under its own rules. Advertising consent: about 70% of visitors.
- Team: a marketing manager and an executive, with a developer for two days a month. The owner signs off budget.
- Test risk budget: £15,000 across the year. Autumn and winter are Lumbrook's busiest quarters, with Black Friday and the January sale inside them.
The figures are invented to make the brief concrete; the reasoning is what is assessed.
Task
- Fix the signals first. Walk the purchase path with the tracking verification checklist. Specify enhanced conversions on Google, following the enhanced conversions guide, and a deduplicated Conversions APIServer-side event sharing with Meta. set-up on Meta, as in the boiler installer exercise. Decide what value each platform receives per order (contribution by range rather than revenue) and how returns are corrected. Order the work to fit two developer days a month.
- Set profit-based targets. Using the contribution margin and break-even guide and the break-even calculator, derive break-even and target returns by range and for marketplace orders after commission. Add the repeat value only within a payback limit you agree with the owner, as in the pet supplement exercise. Check current returns with the ROAS calculator.
- Plan the tests that gate the budget. For brand search, Shopping, Meta and the marketplace, choose a design from the incrementality testing guide: a geo holdout, a budget-increase test, or a switchback where nothing else is possible. Schedule them outside peaks within the risk budget, and write each one up in the experiment hypothesis template with its decision rule, as in the flower delivery exercise. Run them under the intake, sign-off and learning-log rules of the experimentation programme playbook.
- Assign roles and lay out the plan. Give each channel a role and a way of being judged, following cross-channel performance planning, and put the year on one page with the media plan template, including a protected test reserve.
- Write the budget path. Set quarterly steps towards the owner's goal, each released only when its gating test has passed. Following testing and scaling performance campaigns, name the marginal return at which a channel stops growing and the stop rules that pause a step. Forecast each step as a range with the method in the marginal returns and forecasting guide.
- Set the reporting rhythm. Report monthly in a marketing profit and loss view built from recorded revenue, contribution after returns and commission, and all marketing cost, using the monthly marketing report template. Name an owner for each review, the evidence it reads and the date targets are re-derived.
- Write the plan. A one-page summary for the owner, with appendices for signals, targets, tests, the budget path and the reporting calendar.
Evaluation criteria
| Criterion | What a strong answer shows | Common gap |
|---|---|---|
| Signals | Names the outcome each platform bids on, the fixes to tracking and the order of work | Scales spend before fixing the events bidding learns from |
| Incrementality | Gates each budget step behind a named holdout or geo test with a decision rule | Treats platform-reported returns as incremental |
| Profit targets | Derives targets from contribution margin, returns and payback by range or channel | Keeps one revenue return on ad spend target for everything |
| Budget path | Sets steps, stop rules and the marginal return at which spend stops | Plans a straight-line increase with no exit conditions |
| Reporting rhythm | Reports results in the marketing profit and loss with owners and review dates | Reports platform metrics only |
A strong submission notices what the inputs already say. The channels together claim far more revenue than Lumbrook recorded. Brand search reports the best ROASRevenue generated per unit of ad spend. for the demand least likely to need it. The marketplace's reported return sits below its own break-even once commission is counted, before any question of IncrementalityConversions caused by marketing.. And large fittings earn so little after returns that bidding them at the account average loses money. None of this says a channel is wasted; together it says the first quarter is for signals, values and tests, and only then for spend.
Lumbrook Lighting: the first figures a strong answer reaches
- Channels' claimed revenue against recorded revenue: £242,000 ÷ £190,000 ≈ 1.27.
- Blended contribution after returns: 0.40 × 38% + 0.25 × 41% + 0.20 × 20% + 0.15 × 21% ≈ 32.6%, so the Break-even ROASMinimum ROAS to avoid a loss. on own-site revenue is about 1 ÷ 0.326 ≈ 3.1.
- By range: pendants about 2.6, lamps about 2.4, large fittings 5.0, bulbs about 4.8.
- Marketplace orders: 32.6% − 15% ≈ 17.6% contribution, a break-even return of about 5.7 against a reported 5.0.
- Contribution from all recorded orders less paid media: 0.326 × £160,000 + 0.176 × £30,000 − £40,000 ≈ £17,440 a month, before fees and fixed costs.
