Know what changed.
Know what to do next.

Identify exactly which business the report describes, which records belong to it, and what the database can establish. Record the available tables, keys, date coverage, operational logs, shared functions and exclusions.

Separate services that are configured, actually sold, fulfilled at this branch, fulfilled elsewhere, advertised, or merely proposed. A price-list entry does not prove that the store currently provides the service.

Determine the correct sales measure before calculating performance. Reconcile order totals with item values, quantities, discounts, fees, refunds and other adjustments.

Check the selected measure against the native point-of-sale report for identical dates. Reconcile receipts and account settlements separately. A matching sales total does not, by itself, validate the bank balance, tax accounts or physical stock.

Show recorded sales, positive-value orders, average order value, first-recorded customers and returning customers.

Then show paid-in-full order percentage, paid-value coverage, loaded/uncollected status, on-time completion or loading, and collection speed. Each measure needs its current result, comparison, definition, evidence coverage and next action.

Do not produce an overall score out of 100 without an approved, defensible scoring method.

Show the available full-year history, matched year-to-date comparisons, monthly performance and current-year weekly results. Include relevant rolling 7-, 28-, 30-, 90-, 180- and 365-day views.

Identify when deterioration or improvement first appeared, not just the final percentage change. Mark partial periods clearly and avoid comparing a partial month with a full month.

Assess the following 17 areas:

Demand: revenue momentum, customer activity, first-to-second visits and average docket.

Operations: trading hours, peak coverage, collection performance, promise dates, the ready-message-pickup sequence, stock control and roster fit.

Offer and growth: seven-point assessment, premium finishing, stains, price integrity, repair quotations and predictive customer recovery.

Each area receives evidence-based findings and actions—not an automatic pass because the feature exists in the software. Reassess the same definitions at 30 and 90 days.

Explain the arithmetic behind the change in sales. Separate the effect of more or fewer orders from changes in average order value and service mix. Reconcile ordinary counter work and account work.

Then determine the scale of any proposed recovery, including the additional orders, capacity and costs required. A sales shortfall is not automatically lost profit, and it does not prove that competitors took the customers.

Measure customer frequency, recorded spending, service mix and return intervals. Distinguish one-time, repeat and established customers.

Show second-visit conversion over appropriate follow-up periods. Customers must have had enough time to complete the relevant observation window. A one-off bridal customer should not automatically be classified as a failed retention outcome, and “first recorded” does not necessarily mean first-ever visit.

Trace the customer journey from enquiry to inspection, quotation, accepted order and repeat visit.

Measure which sources produce actual paying customers—not simply website traffic. Record how often the source is captured and distinguish an acquisition source from a transaction channel such as “Counter” or “Agent”. A click, view or telephone-button tap is not automatically a new customer.

Compare drop-offs, pickups and enquiries by weekday and time of day against staffed hours and roster coverage.

Include collections, breaks, production duties, queues and missed enquiries. Use actual paid hours when reviewing staffing. Transaction timestamps are useful evidence, but they are not a complete count of people entering the shop or all work performed by staff.

Identify activity outside the confirmed staffed hours. Separate Red Box activity, delivery events, authorised staff activity and delayed data entry.

Apply the correct time zone, holiday calendar and branch hours. Count orders once when they appear in more than one channel. An after-hours record is not automatically proof of additional business that would otherwise have been lost.

Report both the original payment choice and the current payment position. Pay Now, paid in full and bank settled are three different statements.

Classify eligible orders as fully paid, unpaid, part-paid or unresolved. Report both:

Paid-in-full order percentage = fully paid eligible orders ÷ all eligible orders.

Paid-value coverage = payments applied up to each individual order’s value ÷ total eligible booked value.

Show overpayments separately rather than allowing one customer’s credit to conceal another customer’s unpaid balance. Intake-month payment summaries describe the position at the extract date unless historical month-end balances have actually been reconstructed.

Compare original, listed and accepted prices. Identify docket discounts, individual item reductions, zero-price items and repeated edits.

Separate valid contracts, packages, specialist quotations, corrections and goodwill from unexplained exceptions. Record the reason and approving authority. A price below a stored reference price does not automatically establish theft, lost profit or an amount that can be charged retrospectively.

Review each repair type by demand, realised price, quoted scope, labour or contractor cost, materials and remake cost.

Confirm who has final pricing authority and whether the customer accepted the scope. Compare equivalent jobs: a simple hem, complex alteration and multi-item promotion may legitimately have different prices. Measure quotation acceptance and contribution before expanding promotions.

Answer directly: When was the last verified price change, which items changed, and by how much?

Distinguish a documented effective date from the first transaction observed at a different price. Compare equivalent items before and after the change, including quantities, reductions and service mix. A higher average docket alone does not prove an across-the-board price increase. Do not count an increase already achieved as another future benefit.

Assess each relevant item using realised price, processing cost, labour, materials, fees, capacity and like-for-like competitor evidence.

Give an explicit decision: HOLD, RECONCILE, ALIGN or TEST AN INCREASE. Protect accepted quotations and approved contracts. Review volume, customer response and retained contribution—not just the new price.

Press-only comparisons must use an owner-approved benchmark for equivalent work, not an assumed universal industry percentage.

Review the complete current price list and retain a separate register of legacy or unmapped items. Identify confusing descriptions, duplicate choices, inconsistent package units and specialist work requiring quotations.

Bring high-use items into a practical counter layout. Include best-seller analysis using relevant quantities, sales, accepted prices and costs where available. Test intake, printing, payments and reporting before changing live selections or hiding old entries.

Review the seven classifications: Dark, Light, White, Linen, Silk, Delicate, and Black & White.

Confirm eligible garments, approved pricing rules and actual transaction selections. Measure assessment coverage and investigate changes over time. A base garment description containing “Silk” is not necessarily evidence that a silk adjustment was selected or charged. Configuration, assessment, quotation and actual charging must remain distinguishable.

Map the relevant garment families to the checks and services they require: garment pieces, construction, fabric, colour, stains, trims, condition, requested finish and specialist risks.

For each family, define the quotation path, production provider, promise date, quality checks and packaging choices. Make measurement units explicit. The matrix must reflect real branch capability rather than every option stored in the system.

Use the actual approved name of the premium finishing service. Establish whether it is genuinely offered, which garments qualify and what additional work is performed.

Where recorded, report eligible, offered, accepted and declined work, together with additional fees, minutes, materials and rework. A dormant premium-service switch does not establish either staff use or untapped customer demand.

Answer four separate questions: Is stain assessment configured? Is it used on actual transactions? Are separate fees posted where appropriate? What treatment and outcome are recorded?

Show monthly and matched-period use, affected orders and the point at which behaviour changed. Distinguish included treatment from genuine additional work.

A stain tag is not a fee ledger, and the whole value of a stained garment is not “stain revenue”. Price differences may contain several adjustments.

Review the evidence recorded before accepting specialist work: care information, construction, trims, backing, prior damage, earlier treatment and condition photographs.

Identify the qualified assessor, production provider, custody arrangements and agreed scope. Record final quality checks and unresolved marks or damage. This section supports controlled acceptance and communication; it is not a treatment recipe or a blanket guarantee of results.

Map the actual journey for each service:

Intake → assessment → processing → finishing → quality control → loading/ready → customer handover.

For external work, add dispatch, provider receipt, expected return and actual return. Assign responsibility to the relevant store, plant, contractor or transport stage. On-site standard cleaning does not prove that every specialist job remains in the shop.

Measure whether customer commitments are met, while also showing how many records contain usable timestamps.

Include overdue open jobs, not just completed ones. Report median turnaround and the 90th percentile—the point within which 90% of measured observations fall. Separate ordinary, specialist and externally processed work. Missing stage timestamps limit the ability to assign responsibility for delays.

Trace the complete communication chain: ready event, provider acceptance, delivery or failure, reminder and actual collection.

Check recipient accuracy, notification failures, collection time and orders still awaiting pickup. A message marked “sent” is not proof it was delivered or read. Collection reminders must be suppressed once pickup is confirmed. Keep operational reminders separate from promotional campaigns.

Separate system records marked uncollected from garments physically verified as present.

Maintain four worklists: ready and paid; ready with a balance or payment exception; uncollected without a validated ready/load event; and other paths such as accounts, zero-value orders and credits.

Record docket, pieces, location, physical condition/status, checker and check time. Include zero-value garments. Age stock separately from intake and from ready date using explicit bands: 0–29, 30–59, 60–89, 90–179, 180–364 and 365+ days.

A paid customer needs a collection request—not another payment request. The report itself does not authorise disposal, storage fees or bulk closure of old records.

Determine which balances are valid, current, overdue, disputed, already settled elsewhere or unresolved.

Check receipts, credits, refunds, fees, adjustments and fulfilment status at order level before customer follow-up. Do not offset unrelated customer accounts. A positive “Due” field is a reconciliation starting point, not automatic proof of collectible debt or new profit.

Reconcile work accepted by the branch with invoices, receipts, credits and settlements managed elsewhere.

Establish which location owns acceptance, production, billing and collection. Identify account transfers or changes in billing pathways before declaring revenue or customer losses. Account work must not be mixed into ordinary unpaid-stock measures without validated mapping.

Distinguish entry corrections, duplicates, cancellations, quotation disputes, recleans, remakes, damage and loss.

Measure incident frequency against an appropriate completed-work denominator, together with actual remedial costs, resolution time and repeated causes. Administrative cancellation events must not automatically be reported as quality failures or evidence of staff misconduct.

Compare services and garment families by recorded value, relevant orders, quantities, accepted prices and customer demand.

Use the results to decide which services to KEEP, FIX, GROW or STOP. Examine whether apparent growth is driven by price, volume or a different mix of jobs. Service and garment-family views overlap and must not be added as independent sales totals.

Assess behaviours staff can control: correct item selection, garment assessment, quotation authority, promise dates, source capture and ready-order follow-up.

Use validated staff identities, actual paid hours, duties and opportunity exposure. Review examples with staff and document coaching outcomes. Raw sales attributed to a login do not establish who processed the work or whether staff had equal selling opportunities.

Identify active, warm, at-risk and dormant relationships using the customer’s established visit pattern and service requirements.

Review valuable overdue relationships individually and check duplicates, account movements, disputes and one-off needs. Historical customer spending is not a guaranteed recovery forecast, and absence from the store does not prove that a competitor won the business.

Create defined, relevant customer audiences rather than sending the same promotion to everyone.

Specify eligibility, permission, duplicate suppression, message purpose, budget, response window and stopping conditions. Use a comparison or holdout group where practical. Measure additional enquiries, accepted orders and contribution; an order placed after a message is not automatically caused by it.

Follow the full journey from enquiry and inspection through quotation, acceptance, cleaning, quality control, preservation and collection.

Distinguish garment and package scope, including fabric, layers, trims, train, stains and box or bag inclusions. Review accepted prices, quotation conversion and costs. Wedding-family docket counts are not necessarily gown counts, and included preservation packaging is not a second sale.

Compare household work by size, filling, service scope, accepted price, repeat interval and processing requirements.

Review drying and finishing capacity, packaging, storage, turnaround and rework. Any seasonal campaign should use a relevant audience and measured incremental results. A king feather doona should not be compared directly with a competitor’s unspecified starting price.

Verify whether pricing is per drop, metre, area or another defined unit. Record dimensions, lining, backing, condition and specialist assessment requirements.

Separate customer drop-off from removal and rehanging. Establish who quotes, who processes, who transports and what the customer accepted. Track quotation conversion and actual fulfilment costs before expanding the offer.

Check both explicit schoolwear codes and ordinary jacket, skirt or trouser categories before concluding that uniform work is absent.

Define the actual cleaning, repairs and turnaround offered. Verify local term dates and production capacity before scheduling campaigns. Capture useful garment information without collecting unnecessary information about children.

Review bags, boxes, tissue, garment-care products, vouchers and other relevant merchandise.

Separate optional purchases from packaging already included in a service. Measure units, stock costs, write-offs and contribution. Record voucher issues and redemptions distinctly and reconcile their accounting treatment rather than automatically counting both as additional earned sales.

Identify dependence on major customers, key staff, contractors, transport routes, machinery and critical systems.

Review absence cover, documented handovers, access controls and actual backup-restoration evidence. Customer concentration measures need consistent definitions because one customer identifier may represent an individual, household or commercial account.

Rank opportunities by evidence, expected impact, effort, cost, capacity and dependencies.

Protect supported strengths, correct control weaknesses and develop opportunities through controlled trials. Assign a manageable first-week workload. Do not change every price, service, trading hour and marketing channel simultaneously, or label every task urgent.

Build a supported baseline and clearly labelled downside, base and controlled-improvement scenarios.

Separate existing-volume price effects from additional orders, service development and customer recovery. Prevent the same order appearing in several projected benefits. Show revenue and contribution separately, with costs and response assumptions visible. Collecting old balances is not new forecast sales.

Every material finding becomes an accountable record:

Attention → evidence → current position → impact → known or possible cause → exact fix → named owner → deadline → budget → success test → verified result.

Track progress separately from business performance. An email sent, request raised or checkbox ticked does not prove that the underlying problem has been resolved.

Anchor the plan to the owner-approved start date.

Report preparation does not itself authorise database changes, messages, charges, price changes or garment disposal.