What this playbook covers and when to run it
The playbook covers one planned peak from the decision to run it to the review afterwards: a seasonal gifting period, a sale event the business runs every year or a one-off clearance that will stretch the whole team. It joins the work that today sits with separate owners, the promotion plan, stock and delivery, merchandising, search pages, paid media, email, affiliates, customer service, the change freezeA set period around a peak when changes that are hard to reverse are stopped while routine, reversible work continues through an agreed route. and measurement, and gives the whole sequence one owner.
It does not teach any of those parts. The guide to planning promotions decides the mechanic, eligibility, the break-even and where the promotion sits in the calendar. The change freeze guide, the service briefing guide, the campaign merchandising guide, the guide to seasonal sale pages, the Meta guide to sales and seasonal peaks, the Google Ads guide to sales and tracking outages and the guide to sale and launch emails each own one team's part. The guide to measuring promotions judges the result. The sale promotion readiness checklist is the go-live gate for a single sale; this playbook is the programme around it.
Run it when the store has a peak it repeats, when last year's peak ran over stock or service, or when each team prepared its part well and the parts still did not meet. The e-commerce industry guide sets out three planning habits for seasonality and the trading calendar: build stock, feed and creative changes before a peak, read results across the whole cycle, and plan retention for the weeks after. This playbook puts owners and gates on those habits. A product launch is a different motion with no sales history behind it; the product launch playbook runs that instead. Like every playbook in the playbooks library, this is a starting shape that a small store can run with two or three people and a larger retailer can split across several teams.
Roles and responsibilities
- Trading lead. Owns the peak from the promotion decision to the review. Chairs every gate and the daily stand-up, holds the decision logA running record, one row per decision, of what changed and why, what was expected, who decided and when the result will be checked. and is the one person who may change the plan during the window. Names a deputy.
- Paid media leads. Own search, shopping and social campaigns for the peak: structure built in time, budgets and Budget PacingHow fast a campaign spends its budget across the flight compared with the plan. for the window, and the platform-specific checks in the Meta and Google Ads peak guides.
- Email and CRM lead. Owns who hears first, the announcement, reminders only to people who have not bought and the post-peak flows for new customers.
- Site and merchandising owner. Owns the landing collection, homepage and menu slots, the permanent sale pages, one price everywhere and the take-down. Usually also owns the Product FeedA structured file of a store's products, prices and availability used to build shopping and catalogue ads. or works with whoever does.
- Customer service lead. Receives the service brief, prepares saved replies and the escalation routeA short written route naming who front-line staff contact, how and how fast, when a problem is beyond what they can decide alone., and reports what customers are asking at each stand-up.
- Operations and stock contact. Confirms what can be sold, packed and shipped, the delivery promise the business can keep and the cut-off dates, and calls stock risks early.
- Analyst and finance partner. Sets the break-even and the baseline before the peak, verifies tracking, watches recorded orders against the business's own order system during the window and leads the measurement afterwards.
A phase-gate review gives the programme its shape: each phase ends with a short review where the trading lead decides go, change or stop on the evidence from each owner. In a small store one person may hold the trading, merchandising and email roles at once. The roles still need to be written down, because the unowned one, often the stock contact or the service lead, is the one that surprises the team mid-peak.
Cadence, inputs and outputs
The rhythm below is relative to the business's own peak, not to a calendar borrowed from elsewhere. Set the dates from past peaks, from how long the team needs to build and test, and from how long paid campaigns usually take to settle, and adjust them after each review.
| Cadence | Owner | Inputs | Outputs |
|---|---|---|---|
| Well before the peak | Trading lead, finance partner | Last peak's review, margins, the promotion calendar | An agreed promotion, its break-even, eligibility and the peak's dates |
| Build window | Channel owners, merchandising owner | The promotion brief, stock and capacity check | Pages, feed, campaigns, emails and partner offers built and tested; the service brief |
| Freeze | Trading lead | The freeze notice and exceptions route | Only agreed changes, each recorded and tested |
| Daily during the peak | Trading lead with every owner | Stock, pacing, recorded against actual orders, service queue | A short decision log: what changed, who decided, what to watch |
| After the peak | Analyst, trading lead | Orders, baseline, margin, new customer list | A promotion scorecard, a post-mortem and second-order flows switched on |
Peak figures also belong in the team's normal reporting. The reporting cadence playbook sets who sees which numbers when; during the window the daily stand-up replaces nothing in it, it only adds a short operational read.
The playbook in phases
Decide the promotion and its break-even
The job here is a promotion with a purpose, an eligible group, honest prices and deadlines, and a break-even agreed before anyone agrees the discount. The guide to planning promotions sets the job, the mechanic and who is eligible, and protects the full price across the year. It records that in the UK the Competition and Markets Authority enforces consumer protection law that covers reference prices never genuinely charged, untrue urgency and fees revealed late, and turns that into three checks before any promotion is briefed: a "was" price must have been a real price, a deadline must be real, and the advertised price must be one people can pay. The honest persuasion guide applies the same tests to countdowns, stock messages and add-ons on the site. None of this is legal advice; whoever handles compliance reviews the offer. The break-even comes from the guide to contribution margin and break-even, which shows how a discount lowers the margin on every order and how to derive Break-even ROASMinimum ROAS to avoid a loss. from what is left; the break-even calculator turns the promotion's own costs into the orders needed. The gate: the trading lead signs off the mechanic, eligibility, price evidence and break-even, or changes the promotion.
Check stock, capacity and delivery promises
This phase establishes what can be sold, packed, shipped and answered, and the delivery promise the business can keep. The guide to marketing within capacity treats stock, staff and fulfilment as planning inputs and sets out the choices when demand will outrun them: shift, slow, queue or stop it. The delivery promises guide sets a promise operations can keep, states it the same way on pages, feeds and ads, and covers cut-off times and last order dates for a peak. The operations contact brings the stock position for the products the promotion will carry; products that cannot be replenished are marked so the build phase can plan around them. The gate: the product list, the delivery promise and last order dates are agreed, and any product with too little stock is taken out of the promotion's advertising before it is built in.
Build pages, feeds, campaigns and flows
Everything the peak needs is built before the window: landing collections, permanent sale pages, product data, paid campaigns, emails and partner offers. The campaign merchandising guide chooses the products, builds the landing collection and homepage slots and keeps one price on the site, feed, ads and email. The seasonal sale pages guide keeps one permanent URL per event and prepares it ahead of the peak. The product lifecycle guide for Shopping and Performance Max labels seasonal ranges, clearance lines and stockouts so each is advertised on the right terms. The Meta peak guide launches the peak structure early so the ad sets relied on in the window have left learning, and prepares creative, audiences and the catalog in advance. The sale and launch emails guide plans who hears first and reminders only to people who have not bought, and the affiliate promotions guide gives partners the offer calendar, codes and notice in time. Stores that advertise a large range through product campaigns run the Shopping and Performance Max playbook alongside, so feed fixes and lifecycle labels are in place before the window, and creative for the peak goes through the creative production playbook. The gate: every owner reports built and tested, or names what will not be ready.
Brief service and freeze changes
Customer service is briefed on the offer and its rules, and a freeze notice says what may still change and who decides. The service briefing guide gives the support team the offer and its terms, likely questions, saved replies, the escalation route and the end date on one page. The change freeze guide names the period, lists the kinds of change frozen, keeps routine reversible work moving, finishes and records changes before the start, writes the exceptions route and tells agencies and developers. In this playbook the trading lead is the freeze's decision-maker, so exceptions and in-window decisions go through the same person. The gate: the brief is acknowledged by the service lead and the freeze notice is confirmed by everyone who can change the covered systems.
Pass the readiness gate
Every team's checks are done and tracking is verified before the promotion goes live. The sale promotion readiness checklist covers the offer and price claims, site and stock, media budgets and pacing, email and measurement. The tracking verification checklist confirms the tracking plan, tag preview, analytics debug view, conversion status in the ad platforms and consent signals. A failed item goes back to its owner with a fix and a recheck; the trading lead decides whether the promotion goes live, goes live with a named risk or waits. The decision is recorded.
Run the daily peak routine
During the window a short daily stand-up covers stock, pacing, tracking and service, and protects automated bidding from broken days. The Google Ads guide to sales and tracking outages sets out, from Google's documentation, when a seasonality adjustment suits a short sale and which changes to hold back during an unusual period; the ROAS calculator checks what an expected change does to return on spend. The Meta peak guide lists what to touch and what to leave in the window and the order of checks when something breaks, and reminds the team that the most recent days are always incomplete. If recorded conversions fall away from the business's own orders, the tracking outage response SOP runs the incident. When a delay or fault affects orders, the proactive service messages guide tells affected customers before they ask. If the service queue reported at the stand-up has grown beyond the business's own reply standards, the tutorial on clearing a support backlog sets the order for working through the customers waiting. The stand-up ends with the decision log updated.
Measure the peak and keep its customers
Incremental orders, pull-forward and margin are read against a baseline, the campaign is wrapped up and new customers get a plan for their second order. The guide to measuring promotions builds the baseline, separates incrementalConversions caused by marketing. orders from pull-forwardPurchases a promotion moves earlier in time rather than creates, as customers who would have bought later at full price buy during the offer., stock-up and cannibalisation, counts the margin given away and follows who the promotion brought in. The campaign wrap-up checklist switches everything off and reconciles the final numbers, and the campaign post-mortem template records what worked, the root causes and the agreed changes. The repeat purchase and loyalty guide plans the second order for the new customers the peak recruited. If the review shows customers have learned to wait for the sale, the troubleshooting guide for customers waiting for sales diagnoses why. If the peak sold well below its plan, the diagnosis for a sale or promotion underperforming separates a target built on the wrong comparison from the offer, campaign, stock and demand causes. The gate: the trading lead brings the scorecard and the agreed changes to next year's promotion plan.
One decision-maker during the window: what the trading lead decides
A peak goes wrong fastest when several people make the same decision differently on the same morning. During the window, channel owners run their channels within the plan; anything outside the plan goes to the trading lead, or the deputy, and is announced once to everyone at the stand-up or in the peak channel.
| Decision | Who raises it | Who decides | Guide behind it |
|---|---|---|---|
| Extend, shorten or repeat the sale | Any owner | Trading lead, with the compliance check | Planning promotions |
| Pull a product from ads or the landing collection | Operations or merchandising | Trading lead | Product lifecycle |
| Budget beyond the agreed plan | Paid media lead | Trading lead with finance partner | Sales and tracking outages |
| An exception to the freeze | The person who found the problem | Trading lead as freeze decision-maker | Change freeze |
| Tell customers about a delay or fault | Service or operations | Trading lead with service lead | Proactive service messages |
| Treat a tracking gap as an outage | Analyst | Trading lead, then the outage procedure | Tracking outage response |
Two rules keep this workable. Decisions that were planned, such as a budget step written into the media plan, need no one's approval in the window; the table covers only what the plan did not foresee. And every decision goes in the log with the time and the reason, because the review will need to tell a decision's effect apart from the season's. The academy exercise on running a winter peak and its close-down for a jigsaw puzzle shop practises setting a peak up as one programme, from the promotion's break-even and the gates to the daily stand-up, who decides what in the window and the close-down in every channel.
A pet supplies shop runs its winter peak
Habrusk Pet Supplies, an invented online pet supplies shop with a team of 6, runs its winter peak as one programme for the first time after last year's sale sold out of its best-selling heated bed on day 2 while the ads kept running. The trading lead sets the promotion as 15% off bedding and toys for 5 days, with the reference prices checked against the last 3 months of selling prices. With a contribution margin of 40% of the full price, the break-even sales uplift on promoted lines works out at 0.15 ÷ (0.40 − 0.15), or 60%, and the team agrees the promotion is worth running only because past sales in the same category cleared that comfortably. The stock check shows only 180 of the hero heated bed, so it is kept on the landing collection but taken out of the shopping and social ads. The freeze starts once the pages, feed and campaigns are built and tested. On day 3 of the window the stand-up catches a price mismatch: one blanket shows the sale price on the site but the full price in the feed. The merchandising owner raises it, the trading lead approves the fix as a freeze exception, and it is logged. After the peak, orders against the baseline show a dip in the 3 weeks afterwards, read as pull-forward, and 410 first-time customers are put into a second-order flow. At 60 days the review compares their repeat rate with full-price first-time customers before deciding next year's mechanic.
Measuring the playbook
The playbook works when the peak is decided on its economics, built in time, run without surprises and judged honestly afterwards. Operating signals show whether the programme held: whether the break-even was signed off before the discount, whether every gate was held on its date, how many freeze exceptions were needed and why, whether each daily stand-up ran and logged its decisions, how long a tracking gap took to be noticed, and whether the readiness checklist was complete before go-live rather than during the first morning.
Outcome signals come from the promotion scorecard rather than from the headline: incremental orders and margin after pull-forward and cannibalisation, the delivery promises kept, the service contacts the brief should have prevented, the new customers recruited and how many came back for a second order at full price. Read them against the store's own previous peaks, not against figures from other businesses, and bring them to the next promotion decision so each peak starts from the last one's lessons. The tutorial on measuring whether delivery promises are kept builds the delivery promises kept figure from the order, dispatch and delivery records, with peak weeks read apart. The academy checkpoint on peaks, promotion close-downs and full-price protection tests decision rights in a peak, the stock gate and freeze exceptions beside the order of a promotion close-down and a test of fewer coded offers.
Common mistakes
- Choosing the discount before the break-even. A discount picked first and justified afterwards rarely survives an honest margin calculation. Set the break-even, then the depth.
- A sale page built new each year. A fresh URL starts from nothing every peak. Keep one permanent page per event and prepare it in advance.
- Advertising products the store cannot ship. Stock and delivery promises are checked before the build, and a product that runs low is pulled from ads by one decision, not argued over.
- Launching paid structures on the first day. New campaigns and ad sets spend the most expensive days learning. Build early, then leave working structures alone.
- No single decision-maker. When every channel owner can extend the sale or change budgets, the window fills with conflicting changes. Name a trading lead and a deputy.
- A freeze without an exceptions route. Something will break. A route that says who decides keeps a needed fix from becoming an argument or an unrecorded change.
- Judging the peak on the headline. Sales during the window include demand borrowed from the weeks around it. Read the baseline, pull-forward and margin before calling the peak a success.
- Forgetting the new customers. The weeks after the peak hold the store's largest group of new buyers. Without a second-order plan, the peak's acquisition cost is never recovered.
