YOUR MERCHANT PLAYBOOK
A better plan starts
with a good routine.
Get the inputs right. Review the decisions. Measure what happened. Follow the process that fits your store and the modules you use.
This playbook is available before installation. Links to app screens require app access and the relevant subscription.
Process for Marketing + Inventory
BEFORE YOUR FIRST DECISION
Give the plan a trustworthy starting point.
Choose who will review decisions, then check the data and assumptions behind them. Product counts alone do not establish that your store is ready.
Choose your plan and a decision owner.
- Activate the module or modules you need on Plans. In Shopify, read and approve the subscription and trial terms.
- Agree who can approve advertising changes and supplier orders. Record your starting sales, contribution after ad spend, stock availability and purchasing commitments in your existing business records.
Before you move onAn owner is responsible for each decision, and you have a baseline to compare with later.
Check sales history, stock and product costs.
- On Data, check the latest successful sync and the period covered by sales history. In Shopify, correct missing unit costs and inaccurate inventory, then sync again.
- Investigate gaps, newly launched products and periods when products were unavailable. Zero sales during a stockout do not necessarily mean zero demand.
Before you move onYou understand the history available and have resolved missing costs or unexpected stock figures before relying on affected recommendations.
A local demo uses synthetic data. Shopify sync is available in the installed Shopify app.
Review the economics and supplier assumptions.
- Check the unit costs, selling-cost allowances, margin assumptions and supplier lead times used by the plan against your business records.
- For inventory planning, review the value assigned to leftover stock. The default is 30% of acquisition cost; the Shopify Settings page lets you change this assumption.
Before you move onThe assumptions reflect your business, or you have identified which recommendations need manual adjustment.
Most thresholds and supplier lead times are currently read-only in Settings. Resolve incorrect defaults with the person managing your app setup. Confirm exact delivery dates and planned promotions in your own operating records.
Settings are available in the Shopify appImport advertising data and confirm what each budget controls.
Marketing- Upload Meta and Google Ads CSV exports for matching complete periods and the store currency. Review import warnings and resolve unmapped campaigns on Ad spend mapping.
- On Marketing evidence, list each actual campaign, ad-set or shared budget once. Confirm its product mapping and the complete budget scope; do not count a shared budget and its child budgets twice.
Before you move onImported spend reconciles with your ad accounts, and each proposed change maps to a budget you actually control.
EACH WEEK, BEFORE ACTING
Turn the latest plan into considered decisions.
Start with freshness and exceptions. Review the reasons before making a change, and keep a record of what you actually apply.
Check the plan date and refresh your inputs.
- Check the week shown on Weekly plan and the last successful sync on Data. Update costs and inventory in Shopify as needed; import current ad exports if you use Marketing.
- Compare the saved plan with anything that changed after it was generated: a delivery delay, promotion, price change or recent budget move.
Before you move onYou know which data the saved plan reflects and which newer events still need a manual check.
The app creates a weekly plan on Monday when data is available. New imports do not automatically replace an existing plan; the page may show the most recent prior week when a current plan is unavailable.
Review stock risks before waiting for the next purchase cycle.
Inventory- On Inventory, review demand ranges, stock coverage, blocked products and purchase suggestions. Investigate missing costs, short histories and unusual demand.
- Check urgent shortages against stock you own and supplier-confirmed deliveries. Bring an order review forward when the stock situation calls for it.
Before you move onUrgent stock risks have a named next step and are not being left until the monthly review.
Product recalculation is a preview. It does not update the saved plan or the quantities on existing purchase orders.
Review the complete marketing recommendation.
Marketing- Read Hold reasons first and use Edit plan to inspect Watch and Keep items. Then consider Push and Clear opportunities, checking stock, margins and promotions against current conditions.
- Distinguish estimates from recommendations supported by eligible lift experiments. Review the whole proposed budget allocation, its current-budget baseline, constraints and uncertainty before deciding.
Before you move onYou can explain the proposed changes and their evidence, and the actual budget controls match what is running today.
Moving a product between lists does not recalculate the saved budget, profit or risk figures. An individual product label is not an executable campaign budget.
Approve deliberately and record what you apply.
Marketing- Record your approval in the app, then apply the chosen changes manually in Meta or Google Ads. Approval itself does not change an advertising account.
- Record the actual control ID, before-and-after weekly budget, execution date, reason and owner in your decision log. If execution differs from the proposal, record that difference and update the saved control inputs for future plans.
Before you move onYou can distinguish the recommendation, your approval and the actual change in the ad account.
MONTHLY, WITH URGENT EXCEPTIONS REVIEWED WEEKLY
Review the commitments behind the next period.
Use a monthly business review to revisit economics, evidence and purchasing. This is a suggested merchant routine; the app’s plan-generation schedule is weekly.
Revisit costs, demand changes and your priorities.
- Review changes in supplier prices, fulfillment costs, returns, selling prices and product mix. Update the source records that feed your next plan.
- Document upcoming launches, promotions or seasonal events, and check whether past sales are representative. Treat these as manual review inputs where the app does not yet capture them.
Before you move onYou have a current view of unit economics and the events that could make a historical forecast less useful.
Reconcile stock and outstanding orders before buying.
Inventory- Compare on-hand stock, open supplier orders, confirmed quantities and delivery dates with your own purchasing records. Check that received and committed stock is represented correctly in Shopify.
- Review proposed quantities against supplier minimums, pack sizes, storage limits and the cash you can commit. These purchasing constraints are not all included in the model.
Before you move onYou have checked for duplicate purchases and confirmed that quantities, timing and cash requirements are practical.
The current app does not pass confirmed receipt dates into the plan. Verify delivery timing yourself before acting on a recommendation.
Approve, export and confirm each supplier order.
Inventory- Review the saved inventory plan before approval. Approval prepares purchase orders; download the CSV and check supplier, SKU, quantity, cost and requested delivery before sending it yourself.
- Keep the supplier’s confirmation and actual delivery date in your purchasing records. Reconcile revisions before sending another order, and reflect placed or received stock in your operational inventory records.
Before you move onEach order has one clear owner, a confirmed quantity and delivery expectation, and a record of what was actually sent.
Reapproving a plan does not revise existing supplier purchase orders. A recalculated preview can differ from the saved order; reconcile that difference in your purchasing workflow before sending.
Improve the evidence behind future budget decisions.
Marketing- Review which decisions still rely on observational estimates and where uncertainty matters most. If suitable, work with your measurement specialist or ad platform on a randomized lift study.
- Enter only eligible, completed results with uncertainty and the required weekly population normalization on Marketing evidence. Allow outcomes to mature; do not enter attributed purchases as experimental lift.
Before you move onNew evidence is valid for the budget control and decision period, or the next plan remains clearly labeled as an estimate.
AFTER RESULTS HAVE HAD TIME TO ARRIVE
Close the loop with what actually happened.
Use consistent periods and your actual business records. Delivery lead times, returns and advertising measurement windows determine when a decision is ready to assess.
Keep a record of executed decisions.
- Use the downloadable decision-log template or your existing records. Keep the plan week, recommendation, actual action, owner, execution date, affected products and relevant budget or order quantity.
- Note stock availability, supplier dates, promotions and price changes. Set a review date that allows the outcome to mature.
Before you move onYou can trace a recommendation to an actual action and a suitable outcome window.
The app stores plans and approvals. A complete record of executed changes and their outcomes still needs to be maintained in your own records.
Compare actual outcomes with the original expectation.
- Review contribution after ad spend, stock availability and purchasing commitments alongside sales. Use consistent cost definitions, currencies and periods.
- Compare saved forecast ranges with later demand where demand was observable. Record stockouts and exceptional events; low sales during unavailable periods cannot establish forecast accuracy.
Before you move onYour review separates forecast expectations from observed results and explains missing or distorted comparisons.
Feed corrections into the next decision.
- Correct costs, inventory and incoming stock in the source records. Bring new ad-spend exports and eligible experiment evidence into the next plan where relevant.
- Review repeated misses and their causes with the person managing your planning process. A before-and-after change alone does not establish that the app caused a profit improvement.
Before you move onThe next review starts with better information and a clear record of unresolved uncertainty.
Automatic model selection and a complete outcome scorecard are still planned. More data creates opportunities to improve; better accuracy or profit is not guaranteed.
What to measure
Use these definitions in your own scorecard. They are review prompts, not live measurements from your store.
- Contribution after advertising
- Net sales less product costs, relevant variable selling and fulfillment costs, and advertising spend. Use one consistent accounting definition.
- Stock availability
- Record stockout days for important products alongside their sales. A sales decline needs context about whether customers could buy.
- Cash committed to stock
- Review the cost of on-hand stock and open purchase commitments, plus aging or excess units. Keep these distinct from cash already paid.
- Forecast versus observed demand
- Compare the saved prediction and range with later observations over the same horizon. Flag missing data, stockouts and exceptional events.
- Executed spend and experiment results
- Reconcile actual budgets and spend to the decisions recorded. Keep platform-attributed results separate from eligible randomized lift measurements.