Brief
Greiskurv Swimwear is an invented maker of swimwear sold online. It builds demand through organic posts and video, a year-round creator programme and paid social on TikTok and Meta, and its summer season carries the year.
The finance director wants evidence of what TikTok and creators add before next year's budget. A hijacked account took days to recover because the only admin had left. A sale ad ran with a wrong price for a weekend. Nobody knows which creator videos may still run as ads, and a customer complained that the videos have no captions. The advanced social media level wrote a social strategy; this plan governs one.
Inputs and constraints
Greiskurv Swimwear: team, spend and orders
Team: the founder; a finance director, also the data protection owner; a head of marketing; a social media manager; a creator coordinator; a paid social manager; a freelance video producer; a freelance developer; a customer service lead with 2 agents, 4 in summer.
| Last year | Amount |
|---|---|
| TikTok spend | £84,000 |
| Meta spend | £96,000 |
| Creator fees and product | £38,000 and £6,000 |
| Orders | 41,000, of which 64% in May to August |
- Delivery postcodes fall into 10 regions: 3 pairs moved closely together for 2 years; the 2 coastal regions swing far more in summer than the rest. Both platforms can target each region separately.
- The monthly report sums TikTok's click-through and view-through conversions. Checkout asks customers nothing about where they first saw the brand, and branded search volume has never been pulled.
- Next year's budget is set in mid-March.
Greiskurv Swimwear: access register and last year's incidents
| Account | Admins today | Recovery details |
|---|---|---|
| Instagram and Facebook portfolio | Social media manager only | Her personal phone |
| Brand TikTok account | Shared login known to 4 people | The former agency's email |
| TikTok Business Center | Founder, paid social manager | Founder's email |
| Meta ad account | Paid social manager; the former agency still a partner | Business email |
- July: Instagram taken over; the only admin had left in spring. Recovery took 4 days, and followers heard nothing until day 3.
- August: a sale ad showed 40% off instead of 25% from Friday evening to Monday morning; 212 orders at the wrong price. Agents decided remedies case by case, and service heard from customers first.
- Both incidents sit in one shared note. Neither was reviewed.
Greiskurv Swimwear: creators, the TikTok account and a sample of posts
- Creators last year: 38; contracts for 22, of which 9 carry usage terms and end dates. 14 creator videos run today as Spark Ads or Meta partnership ads, 5 with known end dates. Gifted creators are untracked and their posts never checked for labels.
- TikTok: assets owned; browser and server purchases both counted, and TikTok reports 1.3 purchases for every shop order. Prospecting excludes nobody.
- The last 40 posts: 24 videos, 6 with captions, 4 of those generated and 3 misspelling product names; 16 images, 5 with alternative text; 7 posts give prices or sizes only inside the image. The captions complaint waited 9 days for a reply, and there is no route to report a barrier.
Constraints:
- No test, rescore or structural change between May and August; a change freeze covers the season, with the head of marketing deciding exceptions.
- Evidence for the budget must be read before mid-March.
- No new hires; a freelance analyst has days for one test at a time.
- No line of the plan forecasts a size: say which way you expect each result to move and what you would look at to check it.
- Compliance reviews offers, claims and accessibility commitments; nothing here is legal advice.
Task
Write one plan, in this order.
- Owners and calendar. Name the programme owner, an incident lead and deputy per channel, an accessibility owner and the owner of each audit and test, in a RACI matrix. Draw the year and date every change in the measurement change log.
- Contribution evidence. Keep view-through on its own line, start the outside signals in the self-reported attribution guide now, and choose between a geo holdout and a platform lift study by what must be measured, checked with the holdout test readiness checklist and its rule written in the experiment hypothesis template. Say what creators and organic video can claim this year and what waits.
- Incident readiness. Apply the marketing incident management playbook: a contact sheet with deputies who can act, an access register with recovery details from the tutorial on securing social media accounts, the procedure each kind of incident runs, levels and holding statementsA brief first public response in a crisis that acknowledges the issue, tells affected people what to do now and says when more will follow. from the social media crisis plan, review triggers, a rehearsal before the season and the freeze from the change freeze tutorial.
- Programme audits. Score the influencer programme audit and the TikTok Ads account audit on fixed windows, gates first, with one fix list, owners, rescore dates and a record of Usage RightsThe terms in a creator or influencer agreement that set how, where and for how long a brand may reuse the content, organically or in paid ads. for every creator video in paid use.
- Accessibility standards. Apply the marketing accessibility programme playbook: standards in the brand guidelines template and creative brief template, required caption and Alt TextThe text alternative to an image, read by screen readers and used by search engines. fields, a check owner per channel with the social post publishing checklist as the social gate, review of generated captions, monthly spot checks and a route to report a barrier.
- Learning across the year. Set the monthly read of the error log beside the risk register, reviews run on the method in the tutorial on reviewing a marketing error, and how their actions reach the access review and the next audit.
- Summary. The one-page marketing plan template, filled in for the founder.
Evaluation criteria
| Criterion | What a strong answer shows | Common gap |
|---|---|---|
| Contribution evidence | Keeps view-through on its own line, starts outside signals now, chooses a geo holdout or a lift study by what must be measured and books it before the budget decision | Last click or the platform's own column treated as the answer |
| Incident readiness | Names an incident lead per channel, a tested contact sheet with deputies who can act, an access register with recovery details and the procedure each kind of incident runs | One admin, no deputy and no rehearsal |
| Programme audits | Scores the creator programme and the TikTok account on fixed evidence windows, gates first, and orders one fix list with owners and rescore dates | Results scored before disclosure, tracking or ownership |
| Accessibility standards | Puts standards into templates and briefs, names the owner of the check in each channel, schedules spot checks and opens a way for customers to report a barrier | Captions added after each complaint |
| Learning across the year | Reads the error log monthly, reviews incidents that meet the triggers and carries their actions into the access review and the next audit | Incidents fixed and forgotten |
| A calendar that can be read | Orders tests, audits and rescoring so each is read alone and none falls inside the summer season | Everything changed at once just before the season |
Order decides most of the grade. The access gaps come first, because last summer's cause is still open: one admin on a personal phone, a shared TikTok login recovering to the former agency, the agency still a Meta partner.
The purchase count is the trap in the evidence. Until browser and server events count each order once, TikTok's own figures and any platform lift study read against them cannot be trusted, so the event fix runs first and settles for a month before a lift study starts; a geo holdout reads the shop's own orders, but it must still fall outside the summer season and be read before mid-March. A strong plan uses the readiness checklist to say whether the regions can carry a readable geo holdout on the orders left outside summer; if not, it books the lift study for the budget, names what the platform cannot see and books the independent holdout for after the season. Either choice scores well when the reasoning is written first. Creators and organic video get a range with the gaps named, never a figure implying IncrementalityConversions caused by marketing. nobody measured.
Creator videos without a known end date are paused until checked, and accessibility lands in templates before the summer content is made. Compare your plan with the exercises on evidence for a fishing tackle retailer, a hijacked account at a climbing centre and audits for an oral care brand.
Reference figures for self-checking
- Summer carries 64% of orders, so the months a test may run carry 36%: check readability before booking a holdout.
- Creator videos in paid use with a known end date: 5 of 14, so 9 are paused until their rights are found.
- TikTok over-count: at 1.3 reported purchases per order, about 0.3 ÷ 1.3 ≈ 23% of reported purchases were never separate orders.
- Captions on 6 of 24 videos (25%); alternative text on 5 of 16 images (about 31%).
- The wrong-price weekend: 212 orders at 15 points more discount, one remedy decision needed, not 212.
