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

E-commerce Marketing › Intermediate

Capstone: An E-commerce Plan from Acquisition to Retention

The capstone of the intermediate level: a marketing plan for an invented homeware store whose sales have stopped growing. It joins feed and shopping ad structure, margin-based targets, order value levers, retention flows and loyalty into one plan, with a pirate metrics view of where each part should move the numbers, judged against explicit criteria.
Capstone projectIntermediate levelModule 3: Retention, loyalty and lifetime value7 minUpdated

Brief

Bramblewood Home Goods is an invented online store selling candles and diffusers, cushions and throws, tableware, small furniture and seasonal decorations from its own website. It has grown for several years on Google and Meta ads and an email list, and its revenue has now been flat for a year while ad costs have crept up. The owner's instinct is to spend more on ads. Your job is to show where the store's growth is actually stuck and write a plan that fixes it.

Start from the growth equation in the e-commerce marketing hub: traffic, conversion, order value and repeat multiply, and margin against acquisition cost decides whether growth makes money. Use the AARRR view for e-commerce sales to price each stage, and the repeat purchase and loyalty guide for the retention design.

Inputs and constraints

Everything about Bramblewood is invented for the exercise.

Illustrative example — not a real result

Bramblewood Home Goods: last 12 months by category

Margin is after product cost, delivery, payment fees and returns. The repeat column is the share of customers whose first order was in that category who ordered again within 180 days.

CategoryProductsRevenueMarginAd spendReported ROASRepeat within 180 days
Candles and diffusers120£210,00052%£22,0004.624%
Cushions and throws260£260,00044%£38,0003.915%
Tableware180£150,00038%£24,0003.412%
Small furniture60£190,00024%£44,0003.25%
Seasonal decorations140£90,00030%£16,0003.09%

Total revenue £900,000, the same as the year before. Ad spend £144,000, with about 6,800 new customers from paid channels, so roughly £21 per new customer. Blended reported ROAS 3.6.

Illustrative example — not a real result

Bramblewood Home Goods: ads, baskets and retention

Ads: one Performance Max campaign across the whole catalogue with a store-wide target ROAS of 3.5. Feed titles are copied from page titles such as "The Ellis – our favourite"; no custom labels; product type is filled. Meta runs catalogue retargeting only, on the whole catalogue. Conversion tracking was last verified before a checkout change in the autumn.

Baskets: average order value £46. Free standard delivery over £75; the store pays £5.20 per standard parcel. Furniture ships by courier, charged at cost.

Basket valueShare of orders
Under £3030%
£30 to £49.9933%
£50 to £64.9921%
£65 to £74.997%
£75 and over9%

Retention: 15,000 customers ordered in the last 12 months; 14% of first-time customers order again within 180 days. Flows: a welcome series and an abandoned cart flow, nothing else. A points programme gives 1 point per £1 and a £10 voucher at 1,000 points; 80% of points are never redeemed, and members order no more often than non-members.

Constraints:

  • The team is the owner, a marketing manager and a freelance designer.
  • The monthly ad budget stays where it is; the owner will only raise it once targets per category are in place and being met.
  • No new software beyond the existing email platform, which can send SMS, though no SMS consent has been collected yet.
  • A website change freeze runs from mid-November to early January.
  • Small furniture is the owner's favourite range and the hardest to cut.

Task

Work through the steps in order. Each produces a section or a linked document for the plan.

  1. Diagnose. Price each AARRR stage from the inputs and name the weakest term of the growth equation. Say why spending more on ads under the current structure would not fix it, using the category table.
  2. Feed and shopping structure. Set out the feed fixes (titles first) and the custom labels you would add, then the campaign structure: which segments go to Performance Max, which to Standard Shopping and which get no paid shopping. Use the catalogue segmentation exercise as the model, and say what changes in Meta catalogue retargeting.
  3. Targets. Work out the break-even ROAS (Break-even ROAS) per category with the break-even calculator, set a target above each and use the ROAS calculator to check what the budget can buy at those targets. Make a written call on small furniture, including whether its repeat purchase justifies a lower first-order target.
  4. Order value. Choose the levers that fit this catalogue, propose a free delivery threshold from the basket distribution, show how it moves the break-even line, and write the test on the experiment hypothesis template, as in the coffee roaster exercise.
  5. Retention flows. Map the post-purchase, replenishment, cross-category, win-back and sunset flows, with trigger, timing, channel and exit, following the e-commerce email and SMS flows guide and the spice merchant exercise. Include how SMS consent will be earned before any SMS is sent.
  6. Loyalty. Redesign or replace the points programme so that the first reward is reachable on the second order, state its cost, and say how a holdout will measure its effect.
  7. Dashboard. Write the AARRR stage definitions Bramblewood will use, with retention windows by category, today's baseline, the direction each part of the plan should move each stage, and lifetime value (LTV) read as contribution per customer at fixed ages.
  8. Sequence and budget. Order the work across the half-year around the change freeze and the team's capacity. Run the tracking verification checklist before restructuring campaigns, record monthly spend by segment in the media plan template, and keep a small reserve for tests.
  9. Assemble. Put the plan on the one-page marketing plan template, with risks and assumptions in the last row and the detail linked rather than pasted.

Evaluation criteria

Judge your plan, or a peer's, against this table before calling it finished.

CriterionWhat a strong answer showsCommon gap
DiagnosisStages priced; repeat purchase and margin by category named as the constraint, not traffic"Spend more on ads" with no stage-level reasoning
Feed and structureTitles fixed first; a few documented labels from data; segments split by target, with no product in two campaignsCampaigns split by category with one target each, or labels typed by hand
TargetsBreak-even per category, targets above it, a clear call on furniture with any repeat assumption written downOne store-wide target kept, or furniture protected without numbers
Order valueThreshold chosen from the distribution, effect on postage and break-even shown, test judged on contribution per visitorThreshold set from the average and judged on average order value (AOV)
Retention flowsEach flow has trigger, timing, channel and exit; timing comes from category cycles; SMS waits for consentA list of flow names, or one delay for every product
LoyaltyFirst reward reachable on the second order, cost stated, holdout plannedMembers compared with everyone else, breakage treated as success
DashboardStage definitions fit a shop, windows by category, baselines and expected direction for each part of the planGeneric funnel metrics with no baseline
SequenceTracking checked first, work ordered by impact and capacity, nothing launched during the freezeEverything starts in the first month
The pageOne page, decisions visible, detail linkedA long document with the decisions buried

Submission checklist

  • The diagnosis prices each AARRR stage and names the weakest term of the growth equation.
  • Break-even ROAS is shown for every category and each target sits above it.
  • The furniture decision is explicit, with any repeat purchase assumption written next to the target.
  • Feed titles, custom labels and the campaign split are specified, and no product sits in two campaigns.
  • Meta catalogue retargeting has product sets and a stated purpose.
  • The free delivery threshold comes from the basket distribution and has a written test with guardrails.
  • Every flow has a trigger, timing, channel and exit, and SMS starts only after consent is collected.
  • The loyalty change has a reachable first reward, a cost and a holdout.
  • The dashboard shows baselines and where the plan should move each stage.
  • Tracking verification comes before any campaign change, and the calendar respects the change freeze.
  • Risks and assumptions are listed, including what would show the plan is wrong.

When you are done, compare your plan with the one-page marketing plan template example, then revisit any module of the intermediate e-commerce level where your plan showed a common gap.

Pages this lesson applies

Terms used on this page

Also related

Related