Brief
Mossgrove Ceramics is an invented business: a pair of potters who sell handmade mugs, bowls and plates from their own website and take commissions for custom tableware. Most sales happen in the days after a "kiln opening", an email to subscribers announcing a new batch. The owners want three things from their marketing this year: more online sales at full price, a bigger subscriber list for kiln openings, and a steady flow of commission enquiries.
They have GA4Google's event-based analytics platform. on the site, but nobody trusts it. A freelancer installed it some time ago, nobody has looked at the settings since, and the revenue it shows never matches the shop platform. They have asked you to sort it out properly and to tell them what to read each month.
Inputs and constraints
Mossgrove Ceramics: what you are given
| Input | Detail |
|---|---|
| Last month from the shop platform | 164 orders, £9,840 revenue, so £60 average order value |
| Last month in Google Analytics | 131 purchases, £7,210 revenue |
| Subscribers | 2,300 on the email platform; about 90 new last month |
| Commission enquiries | 11 last month, arriving by email from a contact form |
| Analytics property | Created in the freelancer's own account; the owners have viewer access only |
| Tag on the site | A direct Google Analytics snippet pasted into the theme; no Tag Manager container |
| Checkout | Card payments complete on a payment provider's domain, then return to a thank-you page |
| Data layer | The shop platform pushes purchase with transaction_id, value, currency and items on the thank-you page |
| Campaign links | Instagram bio link and kiln-opening emails are untagged; a stockist's blog links to the shop |
| Consent | A consent banner is live and connected to the platform's consent settings |
Constraints: the owners can give a few hours a week and have no developer, so the setup must be something one person can maintain through Tag Manager. There is no budget for paid tools. Any test order must be recognisable and removed afterwards.
Task
Produce one document with the sections below, in order. Link or attach the working files.
- Measurement plan. Use the measurement plan template. Write each objective in the owners' words, give each goal one KPIA metric tied directly to an objective., record the baseline source for each target, and name the source of truth for every KPI. Say which numbers come from the shop platform or the email platform rather than from analytics, and why.
- Property decisions. Work through the GA4 setup checklist for this property. Decide what to do about ownership, and list every setting you would change, with a reason. The payment provider, internal traffic, data retention and enhanced measurement options all need a decision.
- Key events. Using the key events and conversions guide, choose which events to mark and which to leave as ordinary events. For each key event, give the counting method and the source of its value, and explain why the steps towards a purchase are not marked.
- Tag Manager specification. Specify the container in GTMTag management system.: the Google tag and its trigger, each event tag with its trigger, variables and parameters, the exception for the owners' own visits, and what happens to the direct snippet on the day the container goes live. The tutorial on installing GA4 through GTM is the reference.
- Campaign tagging. Write the UTMURL tags identifying traffic source. rules for this shop's links, following the guide to UTM and naming conventions: the Instagram bio link, kiln-opening emails, and the stockist's link. Say which links should stay untagged.
- Verification record. Run the tracking verification checklist with a test order, a test sign-up and a test enquiry. Record each item as passed or failed, with the evidence you would keep and how the test data is removed.
- First monthly report outline. Using the monthly marketing report template, outline the report the owners will read after the first full month: the results table, the channel table, the definitions section, and the standard reports you will pull each figure from, drawing on the reports and dimensions guide. Include how you will compare analytics purchases with shop platform orders, and what gap you would treat as needing investigation.
The kind of gap the report should explain
In the inputs, analytics shows 131 purchases against 164 platform orders: 33 orders, about 20%, missing. A strong submission does not guess at a cause. It lists the candidates the setup has already exposed, such as the payment provider taking credit for sessions, consent refusals and the old snippet, and says which check will separate them in the first month after the new setup.
Evaluation criteria
Judge your own submission against this table, or ask a colleague to. A strong answer meets every row.
| Criterion | What a strong answer shows | Common gap |
|---|---|---|
| Plan starts from the business | Objectives in the owners' words, one KPI per goal, a source of truth for each | KPIs chosen from what Google Analytics already reports |
| Sources are honest | Subscriber growth and commissions read from the systems that hold them | Every KPI forced into analytics |
| Ownership resolved | A property in the business's own account, with the reasoning for moving rather than repairing | Settings fixed on a property the owners do not control |
| Settings deliberate | Payment provider in unwanted referrals, internal traffic tested before activation, retention and enhanced measurement decided | Defaults left unchecked |
| Key events few and meaningful | Purchase, sign-up and enquiry only, with counting methods that match each outcome | Add to basket and checkout starts marked as outcomes |
| Container maintainable | Google tag on Initialisation, a constant for the ID, triggers on confirmed events, an exception for the owners | Click triggers on buttons, the ID typed into each tag |
| No double counting at launch | The direct snippet removed in the same release as the container goes live | Both running for a period |
| Tagging consistent | Lower-case rules, controlled mediums, the stockist link left untagged | Bio link tagged with a free-text campaign |
| Verification evidenced | Each checklist item tested with a real action, recorded with a date, test data removed | Verification read from configuration screens |
| Report decision-ready | Results against targets, definitions stated, a planned comparison with platform orders | A page of charts with no targets or explanation |
Submission checklist
Tick each line before calling the capstone done.
If any row of the evaluation table is weak, the earlier lessons on the beginner analytics level cover each part: the goals-to-KPIs exercise for the plan, the sign-up form specification for the container, and the UTM scheme exercise for the tagging.
