Fundle Answers
Retail AI, answered
in a few lines.
Short, honest answers to the questions retail and mall teams ask Fundle Brain — loyalty, campaigns, marketplaces, MBO/SIS/EBO, social commerce, malls and governed AI. Every answer links to the source article.
Loyalty (30)
What data does a retailer need before using AI for loyalty?
AI-powered loyalty needs reliable customer identities, item-level transactions, returns, reward movements and current channel permissions. Start with one decision, such as identifying eligible first-time buyers for follow-up. Reconcile its source data, document exclusions and test the proposed audience before activation. A large customer database is less useful than a smaller, traceable dataset with clear ownership and refresh rules.
Do we need several years of purchase history before starting?
Not for every decision. A descriptive pilot can start with a shorter, reconciled period if its limitations are visible. However, identifying a genuinely first-ever purchaser or estimating seasonal repeat behaviour requires a longer observation window. State the historical coverage explicitly and avoid calling a customer new when you only know they are new to the available dataset.
Should unmatched store transactions be deleted from the loyalty dataset?
Keep them in the transaction record and report them as unmatched. They still contribute to store sales and coverage analysis. Exclude them from person-level activation until identity is resolved, rather than forcing a match. Maintaining the unresolved records allows later reconciliation without rewriting the original sales history or concealing the size of the gap.
What is the difference between a customer table and a reward ledger?
The customer table describes identities and attributes. The reward ledger records individual balance movements with reasons, times and rule references. A current points balance alone cannot explain how that balance arose. Keeping movements separately lets operations investigate reversals, duplicate earning and disputed redemptions without guessing from a customer profile snapshot.
How should an AI answer handle a late POS feed?
It should identify which stores or periods are incomplete and limit the conclusion accordingly. For example, a repeat-purchase audience may need to wait until the missing transactions arrive. An analyst can still review the available subset, but the interface should not present that subset as complete enterprise coverage or label missing activity as customer churn.
What should we bring to a Fundle data-readiness walkthrough?
Bring anonymised example records, a field dictionary, reconciliation totals, programme rules and a small set of known exceptions. Include one partial return and one duplicate import. Ask the team to explain how the proposed workflow handles them and which steps require configuration, integration or human review. Agree the pilot acceptance criteria before discussing wider rollout.
How should a retailer design a second-purchase loyalty journey?
Begin with a clearly defined first qualifying purchase, then choose a useful follow-up based on the category, fulfilment status and customer permissions. Measure second purchase within a fixed period from that first purchase, comparing customers at the same journey age. Test service, advice or relevant benefits before assuming that every new buyer needs a discount to return.
Should the second-purchase offer always be a discount?
No. Start with the customer need created by the first purchase. Care guidance, fitting support, relevant availability or an appropriate reminder may be more useful than a price incentive. Compare the additional intervention with the existing experience, and judge the result using contribution and customer response rather than assuming that a higher redemption rate means stronger loyalty.
How long should the second-purchase window be?
Choose a window that fits the category and the decision you are evaluating. Review observed purchase intervals, delivery timing and seasonality, then specify the window before analysing the test. Use equal follow-up for all customers. A short window can measure immediate response, but it cannot by itself establish whether the journey improves longer-term retention.
Can an exchange count as a second purchase?
It depends on the agreed definition, but a replacement or size exchange should normally be identified separately from new demand. Link it to the original transaction so it does not inflate the repeat-purchase metric. If an exchange includes an additional paid item, define beforehand whether that incremental purchase qualifies and apply the same rule across comparison groups.
What happens when a customer buys again before the next message?
Remove them from the first-to-second-purchase promotion and reassess their next useful step. A queued discount reminder can be confusing immediately after a purchase. Preserve the original assignment for experiment analysis, while recording that the operational journey correctly stopped or changed. This separates responsible execution from selective removal of successful customers in the final measurement.
How can we evaluate a journey with a small customer base?
Use fewer variants and a clearly defined outcome so the available sample is not fragmented. Estimate the uncertainty around the result and avoid declaring success from a handful of repeat purchases. You may need a longer enrolment period or several comparable cohorts, while keeping the same rules and allowing each customer the full observation window.
How can retailers measure the incremental value of a loyalty programme?
Estimate loyalty incrementality by comparing outcomes for otherwise comparable customers who are assigned different eligible programme experiences. Preserve promised benefits, define the intervention precisely and measure all assigned customers over equal periods. Evaluate repeat purchase and contribution after programme costs. Member spending, redemptions and attributed revenue describe activity, but they do not by themselves show what the programme caused.
Does higher spending by members prove programme success?
No. Members may have been stronger customers before joining, so the difference can reflect selection. Use member spending to understand the programme population, then evaluate a defined change with a credible comparison. If only observational analysis is possible, disclose its assumptions and avoid presenting the full spending difference as revenue caused by loyalty.
Can we measure incrementality without removing existing benefits?
Yes. Test an additional benefit, a new communication or a staged enhancement while preserving the base programme and its promises. Define exactly what differs between groups. The result then measures that added experience for the eligible population, rather than claiming to estimate the value of every historical component of the programme.
Why should non-redeemers remain in the analysis?
Whether a customer redeems is an outcome of the experience, not a neutral eligibility rule. Removing non-redeemers selects a more responsive group after assignment. Keep everyone assigned in the main comparison and use redemption as a diagnostic measure. This answers what happens when the business offers the intervention to an eligible customer.
Should we report incremental revenue or incremental contribution?
Both can be informative, but they answer different questions. Revenue shows additional sales value on the agreed basis. Contribution also considers product economics, returns and relevant variable programme costs. State what is included in the calculation and keep the treatment consistent. A campaign can produce positive revenue lift while reducing contribution after rewards.
What if the programme changes halfway through the test?
Record the change and assess whether it affects both groups or only one. A material change can alter the intervention being evaluated and make the original result difficult to interpret. Do not silently combine different programme versions. The team may need to close the original analysis window and start a newly specified test.
How can a retailer design loyalty rewards without eroding margin?
Estimate the additional contribution a reward can create, then subtract its expected redemption, fulfilment and contact costs across all eligible customers. Check existing discounts, returns and product availability before approval. Compare the proposed reward with a no-additional-reward baseline. High redemption is useful operational information, but a reward succeeds commercially only when it changes behaviour enough to cover its costs.
Is a high redemption rate good or bad?
It depends on the objective and economics. High redemption can show that customers understand and value the benefit, but it also increases fulfilment cost. Compare contribution with the existing experience and examine whether behaviour changed. A low redemption rate is not automatically profitable either, because it may indicate a confusing or irrelevant programme.
Should complimentary products be costed at retail price?
For the operating contribution calculation, use the incremental cost of providing the product, including relevant handling and fulfilment. The advertised customer value is a separate concept. Confirm the cost basis with finance and consider stock opportunity cost where inventory is scarce. A high ticket price does not necessarily mean an equally high programme expense.
Why include reward cost on customers who would buy anyway?
Because those redemptions still reduce the contribution the business retains. The commercial comparison must include every relevant cost incurred by the offer group. If the reward also subsidises baseline purchases, that cost can outweigh the value from additional purchases. Ignoring those customers makes an expensive offer appear more efficient than it is.
Can we use expected point expiry to fund a richer programme?
Treat expiry assumptions cautiously and distinguish operational planning from accounting policy. Forecasts should be checked against actual cohorts and redemption conditions. Do not rely on customers misunderstanding or being unable to use a promised benefit. Review programme value, economics and communication together, with finance responsible for the treatment of outstanding obligations.
How should staff discounts interact with loyalty offers?
Define the combination explicitly and test it in the actual selling systems. Staff may have manual discount permissions that behave differently from standard promotion rules. Provide a short operating guide and an escalation route for unclear baskets. Record overrides so the commercial team can see whether unexpected combinations are changing reward costs.
How should retailers connect customer identities and consent across channels?
Connect records through verified identifiers and retain the evidence behind each link. Store permissions separately by person, channel, purpose and time, then check the applicable current state before activation. A matched customer record does not grant marketing permission. Ambiguous identities, shared contact details and conflicting preferences should trigger a controlled review or suppression rather than an automatic merge and send.
Is a hashed phone number sufficient for identity resolution?
Hashing changes how an identifier is represented; it does not prove that the identifier belongs to one person or that a marketing action is permitted. Shared and reassigned numbers remain ambiguous after hashing. Use verified linkage evidence, preserve source context and evaluate the applicable permission independently before placing a customer in an activation audience.
Should a website and a store always share one customer ID?
A common internal key can be useful when the relationship between records is verified and the use is appropriate. Keep the original source IDs so the connection remains traceable. Do not force every anonymous or ambiguous interaction into that key. Some records should remain separate until the customer or another reliable process confirms the relationship.
What should happen when two linked records contain conflicting preferences?
Apply a documented rule using scope, timing and evidence rather than choosing whichever value permits contact. A recent withdrawal for the relevant channel should not disappear during a merge. Where the conflict cannot be resolved reliably, suppress the proposed activation and route the record for review. Preserve both events so the decision can be explained later.
Does consent mode store a full customer permission history?
No. Google documents consent mode as a mechanism for communicating consent states to its tags, and its guide says the mechanism does not itself save choices. The retailer needs an appropriate preference record and update process around it. That record should support the channels and purposes actually used by the business, beyond website tag behaviour.
How can teams test withdrawals without using real customer details?
Create controlled test identities with known preferences and follow them through scheduling, export and dispatch checks. Change one preference after the message is queued and verify the expected suppression. Include more than one channel so the team can confirm scope. Keep a written expected result for each case and inspect the downstream logs.
Campaigns (18)
How should AI help a retail team choose a campaign audience?
Use AI to prioritise a clearly eligible population, not to override permissions, service exclusions or commercial rules. Define the campaign objective, freeze the available customer signals and distinguish purchase likelihood from responsiveness to the proposed action. Select the candidate pool first, then assign a valid comparison group. Evaluate contribution and customer response before expanding the audience or message frequency.
Should we always target customers with the highest purchase probability?
Not necessarily. Some highly likely buyers would purchase without additional contact, so an incentive may mainly subsidise existing demand. Match the ranking method to the campaign objective and test the intervention within the selected population. Purchase propensity can help describe likely behaviour, but it is not the same as the incremental effect of a message.
Can AI override a contact-frequency limit for a valuable customer?
A frequency limit should operate as an explicit policy constraint unless an authorised business process changes it. A high score alone should not bypass it. Review competing messages and choose the most useful eligible action, which may be no contact. Keep any approved exception visible so the team can evaluate its effect on customer response.
Why are lower-ranked customers a poor control group?
They differ from the targeted group by design, often on the same characteristics that predict purchasing. Comparing their outcomes therefore mixes campaign effect with selection differences. Choose the intended candidate pool first and assign comparison groups within it. This makes the experiment answer whether the campaign helps customers like those the business actually plans to contact.
What evidence should accompany an AI audience recommendation?
Include the audience definition, data cutoff, qualifying signals, exclusion counts, score meaning and relevant uncertainty. Show examples that operators can inspect, including records close to the selection boundary. Preserve the rule and model versions so a later review can reconstruct the decision instead of depending on a narrative generated after the campaign ran.
How do we handle customers with little purchase history?
Use a deliberately limited approach rather than pretending sparse data supports precise personalisation. A general but relevant service message, an optional preference question or no campaign may be appropriate. Keep these customers identifiable as a separate evidence condition, and test any proposed approach fairly. Do not manufacture preferences from weak or unrelated signals.
How do retailers run a reliable campaign holdout experiment?
Define one campaign change, establish eligibility and randomly assign customers to treatment or an appropriate comparison before activation. Keep assignment stable, protect the comparison from the tested campaign and measure outcomes for all assigned customers over equal periods. Choose sample size around the effect worth detecting, then inspect data quality and uncertainty before interpreting lift or making a rollout decision.
What percentage of customers should be held out?
Choose the allocation from the measurement need, rather than a universal rule. Consider baseline purchase rate, the smallest effect worth acting on, expected sample size and outcome variability. A small holdout can leave the result too uncertain. Ask an analyst to size the design before committing the campaign calendar and contact budget.
Should customers with failed deliveries be removed from treatment?
Keep them in the original assigned group for the primary intention-to-treat comparison. The business decision concerns offering the campaign through the actual delivery process, including its limitations. Analyse delivery separately to diagnose operations. Removing failed deliveries after assignment can create a selected group and exaggerate the effect of the campaign.
Can a holdout still receive ordinary store promotions?
Yes, if that is part of the defined comparison environment and applies consistently across groups. The study then measures the extra tested campaign above those ordinary promotions. Document overlap so the result can be interpreted. If unrelated treatment differs systematically between groups, you may be measuring a bundle of changes rather than the intended intervention.
How is an A/B creative test different from a no-contact holdout?
An A/B test between two messages estimates which message performs better under the tested conditions. A no-additional-contact holdout estimates the effect of contacting customers compared with the existing experience. Either can be useful, but they answer different questions. A winning creative does not by itself prove that sending either version creates incremental value.
What should we do when the experiment result is inconclusive?
Report the uncertainty and the range of commercially plausible outcomes. Check whether the test had adequate sample, clean assignment and complete follow-up before interpreting the lack of clarity. Depending on the decision, continue a properly planned enrolment period, simplify the test or revise the hypothesis. Do not turn an inconclusive result into a positive claim.
How should a retailer decide the next best campaign action for a customer?
Define a small set of useful actions, including no promotional contact, and filter them through permission, service, stock and commercial constraints. Compare the remaining actions using a consistent objective and evidence available at decision time. Record the chosen action, rejected alternatives and outcome. Begin with reviewable rules, then test more adaptive decisioning only when feedback and evaluation are reliable.
Is next best action the same as choosing the next best offer?
No. An offer is only one possible action. A useful decision may be a reminder, service intervention, appointment invitation or no promotional contact. Define the available actions around customer needs and operational capability. Restricting the catalogue to discounts encourages the system to solve every situation through price, even when another response would be more appropriate.
Why should no contact be an explicit option?
It provides a meaningful baseline and prevents the workflow from assuming that every customer needs a message. A recent purchase, high contact pressure or weak evidence may make no promotional contact the best choice. The business should still maintain any necessary service process. Record the reason for restraint so it can be reviewed alongside active recommendations.
Can a purchase prediction identify the best campaign action?
A purchase prediction alone describes likelihood under its modelling conditions; it does not compare the causal effects and costs of different actions. To choose among campaigns, the team needs consistent action-specific evidence or a carefully tested decision rule. Include the existing experience as a baseline and avoid equating a high forecast with a large incremental opportunity.
What does reward mean in a contextual-bandit approach?
It is the feedback signal the learning process seeks to improve, such as a defined commercial outcome or a measured cost. It does not necessarily mean a customer incentive or points award. Choose that signal carefully because optimising clicks, immediate purchases and longer-term contribution can lead to different actions and different customer experiences.
When should a next-best-action recommendation require human review?
Require review when the proposed action exceeds approved operating limits, depends on uncertain identity or data, changes a customer commitment or carries a material commercial cost. The exact boundaries should be agreed by the business. A clear review record should show context, alternatives and reasons, so approval is an informed decision rather than a routine confirmation click.
Marketplace (30)
How should a retail brand reconcile marketplace settlements?
Reconcile marketplace settlements by matching order-line events, refunds, charges and adjustments to a settlement period, then matching the resulting payout to the bank receipt. Separate amounts not yet due from unexplained differences. Preserve each source reference, classify exceptions, assign an owner and close a case only when the supporting correction or payment is recorded.
Why can order value and payout differ even when nothing is wrong?
Orders, delivery eligibility, refunds, charges and bank receipt dates can fall in different periods. A payout can also include adjustments unrelated to new orders. Rebuild the settlement from its included financial events before comparing it with the bank. Comparing the bank deposit directly with booked orders mixes commercial activity with cash timing.
What should we do with charges that have no order ID?
Retain them as account-level events until their nature is established. Some charges legitimately apply to a period, service or adjustment rather than an individual order. Record the source reference and classification, then decide whether an allocation is useful for management reporting. Never invent an order match just to make every row look complete.
How do we prevent duplicate recovery claims?
Create a case key from the seller account, source event or adjustment reference, reason and disputed amount. Check both open and closed cases before submission. Link later credits to that case and preserve partial recoveries. A re-exported deduction should update evidence for the existing investigation rather than create another claim.
Should finance close a case when the marketplace approves it?
Record approval as a separate stage. Financial closure should follow the organisation's policy and include the actual credit or payout evidence, including any remaining difference. Keep approved but unpaid cases visible with an expected next event. This prevents operational approval from being reported prematurely as cash recovered.
Can reconciliation start without an API integration?
Yes. A controlled export process can establish the matching rules, source coverage and exception categories before any automation investment. Record file versions and retrieval dates, restrict access and test repeat imports. A later integration should reproduce those controls and expose failures; its availability must be verified for each account and platform.
How do you calculate marketplace SKU contribution after fees, returns and advertising?
Calculate SKU contribution from net realised sales on a defined basis, then subtract the product cost of retained units and attributable marketplace, fulfilment, return and advertising costs. Reconcile refunds and recovered stock separately to avoid double counting. Show the period, return maturity and allocation method so a profitable-looking SKU is not supported by incomplete costs.
Is contribution margin the same as gross margin?
Not necessarily. Gross margin commonly compares revenue with product cost, while a channel contribution view can also deduct variable fulfilment, marketplace, return and advertising costs. Agree the cost boundary with finance and display it clearly. Comparing figures with different boundaries can make one channel appear unfairly better or worse.
How should resellable returned inventory affect the calculation?
Separate the customer refund from the physical stock outcome. A fully resellable return may restore inventory value, while reverse freight and inspection still cost money. A damaged return may create an additional product loss. Use your finance-approved costing policy, and avoid treating every returned unit as both permanently consumed and fully recovered.
Which denominator should we use for contribution per unit?
Choose the denominator that matches the decision and label it. Contribution per shipped unit captures failed deliveries and returns within the original workload. Contribution per retained unit describes the economics of units customers kept. Show both when return rates vary materially; otherwise a change in denominator can disguise deteriorating operations.
Can a positive return on ad spend still mean a losing SKU?
Yes. Advertising return measures attributed sales relative to advertising expense, not all costs of serving those sales. Product cost, fees, fulfilment, discounts and returns can consume the remaining value. Review contribution after those costs and distinguish attributed revenue from demonstrated incremental revenue before treating a strong advertising ratio as commercial success.
What if marketplace costs arrive after the monthly review?
Show provisional contribution with separately identified estimates and their source. Reconcile those estimates to actual postings when they arrive, then explain the revision. Keep the original cohort intact so late costs do not disappear into an unrelated month. Managers should know whether a recommendation rests on mature evidence or incomplete cost recognition.
How do you map marketplace catalogue listings to internal retail SKUs?
Create a governed crosswalk linking each marketplace account and listing identifier to an internal sellable variant, with GTIN where applicable, unit quantity and effective dates. Validate size, colour, pack and bundle composition before publishing the match. Quarantine ambiguous mappings, preserve historical versions and reconcile mapped quantities to source totals before using the data for decisions.
Can we use product titles as the main matching key?
Titles are useful supporting evidence but are unreliable primary keys. Sellers can change wording, abbreviate sizes or use similar names for different packs. Prefer approved identifiers combined with variant and unit checks. When only text is available, treat the match as a review candidate and retain the evidence and person who approved it.
How should bundles be represented?
Record the bundle as a sellable item with an explicit component list and quantities when component-level analysis is required. Distinguish fixed bundles from configurations that can vary. Validate how returns and replacements affect the components. A pack factor alone is insufficient when a bundle contains different colours, sizes or product types.
What is the difference between a SKU and a GTIN?
A SKU is an organisation's internal product identifier, while a GTIN follows a standard for identifying trade items. They can both appear in a product record but should not be treated as interchangeable strings. Preserve the original values, validate their intended level and avoid manufacturing a GTIN from an internal SKU.
Should we overwrite historical mappings when a listing changes?
Usually you should preserve the old version and add an effective-dated change. Historical transactions need the product meaning that applied when they occurred. If the old mapping was an error, correct it through a documented reprocessing process. Separate a genuine product change from an administrative correction so downstream reports remain explainable.
How do we prioritise thousands of unmapped listings?
Rank them by recent transaction value, unit volume, returns exposure and operational urgency. Add a separate queue for identifier conflicts that could contaminate already mapped records. Review inactive long-tail listings later. Track verified coverage on actual business activity, because a high percentage of mapped catalogue rows can still leave important sales unresolved.
How should retail teams distinguish returns and RTO when measuring ecommerce economics?
Treat return to origin as an unsuccessful delivery journey and a customer return as a post-delivery reversal. Measure them against their own eligible shipment or delivered-order populations. Track refunds, transport, inspection, inventory condition and recovery separately. This prevents one blended return percentage from hiding whether the underlying problem is delivery, product fit, expectations or fulfilment quality.
Is a refused cash-on-delivery parcel a customer return?
If the parcel never completed delivery and is sent back, treat it as RTO within the shipment journey. A post-delivery customer return is a different event. Preserve the carrier status and customer communication because refusal, failed contact and address problems need different follow-up even when they end in the same RTO category.
Can we add the RTO rate and customer-return rate together?
Not when their denominators differ. RTO is often measured against dispatched shipments, while customer returns are measured against delivered units or orders. Calculate a combined non-retention measure directly from one common mature cohort if needed. Continue showing the separate rates so managers can distinguish delivery problems from post-delivery product or service issues.
When should returned stock become available again?
Only after the required receipt and inspection steps confirm it is suitable for sale under your operating policy. A carrier delivery scan or customer refund is insufficient evidence. Record resellable, repairable, damaged and unresolved quantities separately, and link the decision to the warehouse record so recovered value can be explained.
How can we compare courier performance fairly?
Compare similar lanes, service levels, payment types, parcel characteristics and cohort maturity. Show shipment counts and unresolved outcomes alongside rates. A courier handling a harder delivery mix can look worse in an unadjusted total. Investigate specific exceptions and response times before concluding that the carrier alone caused the difference.
Which costs belong in a return economics report?
Include the cost boundary relevant to the decision: outbound and reverse logistics where applicable, handling, inspection, unrecovered product value and related customer-service expense. Show refunds and revenue reversals separately from costs. Identify excluded advertising or overhead so users do not mistake a narrow reverse-logistics calculation for full order profitability.
What should a marketplace super intelligence morning decision brief include?
A useful marketplace morning brief shows a small set of material exceptions, the evidence behind each one, its commercial exposure and the next accountable action. Include data freshness, confidence, constraints and approval needs. Treat super intelligence as an operating ambition: recommendations should support human decisions, with execution permissions and outcome measurement explicitly defined.
How many items should the morning brief contain?
Use the number the team can actually decide and assign within its meeting, rather than a fixed universal target. Keep supporting detail and the remaining backlog accessible. If every item is urgent, review the thresholds and ownership model. The brief should clarify priorities without hiding unresolved risks that still require monitoring.
Should we rank every issue by estimated rupee value?
Rupee exposure helps but is insufficient. Include urgency, evidence quality, customer impact, reversibility and documented deadlines. Separate potential contribution, disputed cash and hypothetical opportunity because they are not equivalent amounts. An imminent evidence deadline or unreliable data feed may deserve attention before a larger but weakly supported growth estimate.
What does confidence mean in this workflow?
Confidence should describe the evidence and assumptions behind a recommendation: source coverage, freshness, sample size, matching quality and stability of the comparison. It should not be an unexplained percentage that users treat as certainty. State what additional information could change the recommendation and when the decision should be deferred.
Can AI execute actions from the brief automatically?
Only where an actual implementation supports the action and the organisation has defined permissions, limits, monitoring and exception handling. Some recommendations require human approval or manual execution. Evaluate each workflow separately. The morning-brief framework does not establish that Fundle or any connected marketplace already provides autonomous execution capabilities.
How do we know whether the brief is improving operations?
Track whether accepted items receive owners, whether evidence is complete, how quickly decisions are made and whether recurring exceptions decline. Then examine commercial outcomes with appropriate comparisons and costs. Do not count more generated recommendations as success; a smaller queue with reliable follow-through can be more useful than a large volume of alerts.
MBO (12)
How should brands distinguish MBO sell-in from sell-through?
Measure sell-in as the defined movement from the brand or distributor into the multi-brand outlet, and sell-through as verified sales from that outlet to consumers. Connect both to receipts, returns, transfers and stock balances at variant and location level. Evaluate replenishment using consumer demand and available stock, while showing outlet reporting coverage and timing gaps.
Can a dispatch file be used as secondary sales data?
A dispatch file establishes movement from the sender, not a consumer sale. It can support sell-in and stock-in-transit analysis when its definitions are clear. Secondary or outlet consumer-sales analysis needs the corresponding outlet evidence. If that evidence is unavailable, label the limitation rather than presenting dispatch growth as consumer demand.
Why can a growing outlet still need less replenishment?
Growth can coincide with excess receipts, a slowing recent sales pace or an imbalanced size curve. Review available stock, incoming commitments, lead time and demand at variant level. More total sales do not automatically justify more of every item, particularly when slow variants already occupy working capital and selling space.
Which sell-through formula should an MBO network use?
Choose a documented formula that fits the planning decision and can be calculated consistently from partner data. State the period, treatment of returns, transfers and opening stock. Different software reports use different denominators, so never average their percentages blindly. Recalculate from comparable quantities where possible and disclose exceptions.
How should we handle an outlet that reports weekly instead of daily?
Keep its reporting cadence explicit and mark the latest complete period. Do not distribute weekly sales evenly across days unless the result is clearly labelled an estimate for a specific purpose. For daily decisions, combine that limitation with current stock or partner confirmation and avoid implying visibility that the data cannot support.
What makes a zero-sales alert credible?
The outlet must have submitted a complete record for the relevant trading period, with functioning product mapping and no known reporting failure. Check whether the store traded and whether stock was available. A missing file or unmapped product should create a data exception, while a verified zero can trigger an operational investigation.
How should brands reconcile MBO schemes and partner claims?
Reconcile an MBO claim against the approved scheme version, eligible transactions, exclusions and previous settlements. Separate the partner's requested amount from the calculated entitlement, approval and actual credit. Investigate differences with source evidence and preserve partial resolutions. This creates an auditable commercial workflow without assuming every deduction is invalid or every accrued incentive is already paid.
What is the difference between a scheme accrual and a claim?
An accrual represents an estimated or calculated incentive obligation under the organisation's accounting and commercial process. A claim is the partner's request for compensation or credit. They can differ because of timing, eligibility or calculation. Reconcile both to the approved scheme and let finance determine the appropriate accounting treatment.
How should we handle a claim that is partly correct?
Split the claim into supported and unresolved components while preserving its original reference. Record the basis for any partial approval and the amount actually settled. Continue investigating the residual balance. Rejecting or accepting the whole claim can obscure a valid entitlement and make later credits harder to match.
Can sales approve a scheme change retrospectively?
That depends on the organisation's authority policy and the commercial agreement. The reconciliation workflow should require an authorised amendment with its effective scope and reason. Preserve the original calculation and show the revised entitlement separately. A conversational assurance should not silently overwrite a rule that finance and the partner previously relied on.
How do we prevent the same sale earning duplicate incentives?
Define whether schemes can overlap and apply that rule at transaction level. Keep a link between each eligible event and every entitlement it generates. Review overlapping periods, duplicate partner submissions and prior credits. If overlap is contractually allowed, record it explicitly rather than treating every repeated transaction as an error.
Which metrics help manage the claims backlog?
Track requested value, supported value, unresolved balance, case age, evidence waiting time and the distribution of reasons. Show partial settlements separately. Measure recurring root causes such as unclear scheme rules or missing transaction references. A high closure count is insufficient if large residual balances or repeated disputes remain unresolved.
SIS (12)
How can a brand improve sales and stock visibility across shop-in-shop locations?
Connect each shop-in-shop location's consumer-sale lines to variant-level receipts, transfers, returns and stock states. Reconcile the movement ledger to reported or counted stock, show feed freshness and distinguish physical possession from ownership. Use available stock and consumer demand for decisions, while flagging missing reports and unresolved variances before recommending replenishment or redistribution.
Does shop-in-shop inventory always belong to the brand?
No. Ownership depends on the commercial arrangement, which may be outright purchase, consignment or another model. Confirm the contract and finance policy before interpreting stock balances or liabilities. Keep ownership separate from physical location in the data so operational possession is not mistaken for an accounting conclusion.
Why is available stock different from physical stock?
Some physically present units may be reserved, damaged, awaiting inspection or otherwise unavailable to sell. The exact states depend on the operating system and policy. Establish a shared definition and check whether partner figures already exclude restricted quantities. Without that step, teams can overstate availability or subtract the same restriction twice.
How should we handle stock moving between concessions?
Record a transfer reference with dispatch and receipt events, quantities and responsible locations. Keep the unresolved quantity in transit until receipt is confirmed. Investigate shortages or timing differences without duplicating stock at both ends. A transfer request or approval alone should not establish that the destination can sell the goods.
What can we do when the host provides only weekly sales files?
Use the weekly period honestly and show its latest complete date. Combine it with authorised current stock confirmations when a decision cannot wait. Avoid distributing sales across invented daily records. Agree a practical exception process with the host for urgent availability issues and separately evaluate whether more frequent reporting is feasible.
When should a stock variance trigger a commercial claim?
First verify the count, movement cutoff, receipts, transfers and sales coverage. Then identify the responsible party and applicable contractual process. A variance can arise from data timing or mapping rather than physical loss. Preserve evidence and approval before moving from an operational investigation to a financial claim.
How should brands reconcile shop-in-shop concession settlements?
Reconcile a concession settlement by connecting the host's eligible sales and returns to the applicable commercial agreement, authorised charges, previous adjustments and bank receipt. Keep stock ownership, sales recognition and payout timing distinct. Explain every difference, preserve contract versions and assign unresolved items to an owner instead of treating the host's net statement as self-explanatory.
Should every SIS settlement use sales minus commission?
No. The agreement may include different eligible-sales definitions, fixed charges, minimum guarantees, promotion funding or other adjustments. Establish the actual commercial model and approved calculation before comparing statements. A simplified commission formula is useful only when it accurately reflects that concession's terms and the period being reviewed.
How do we treat a customer return from an earlier month?
Link it to the original transaction and apply the agreement's return and settlement rules. Preserve both the original sale period and the date of the reversal or credit. This allows the team to explain period trading and current cash separately, while avoiding a second deduction when the same return appears in another export.
What if the statement total matches the bank but still seems wrong?
That proves the stated payout reached the bank, not that the underlying commercial calculation was correct. Reconcile eligible sales, returns, rates, charges and credits to source evidence and contract terms. Bank matching is the final cash check; it cannot replace validation of what the host should have paid.
How should minimum guarantees be handled?
Model the approved rule and its period explicitly, including whether it replaces, exceeds or combines with other charges. Check how closures, partial months and amendments are treated. Do not infer the formula from a statement label. Have the commercial and finance owners resolve ambiguity before using the calculation to approve or dispute a settlement.
When is an unresolved deduction a recoverable amount?
Only after evidence supports the brand's entitlement and the relevant resolution process establishes what can be recovered. Initially it is an unexplained or disputed difference. Keep claim submission, approval, credit issuance and cash receipt as separate stages. This prevents the exception dashboard from overstating both leakage and realised recovery.
EBO (12)
What should an EBO daily performance dashboard show?
An exclusive brand outlet dashboard should show comparable sales, completed bills, units, returns, traffic coverage and customer identification alongside operational exceptions. Every number needs a stated period, store set and denominator. Use the dashboard to assign a next action, such as investigating a size gap or fixing a missing feed, rather than to rank stores using incompatible totals.
Should an EBO dashboard use gross sales or net sales?
Use both only when they answer different questions and their definitions are visible. Gross merchandise activity helps explain demand and discounting; a net measure includes the agreed reversals and adjustments. Finance should approve the definitions. Avoid replacing a recognised finance measure with a convenient dashboard total that uses different dates or exclusions.
How should yesterday’s dashboard handle a return from last month?
Show the reversal in the operating period where the source records it, while preserving the original sale link. For purchase-cohort analysis, allocate it back to the relevant cohort under an explicit rule. Label these views separately so the manager can distinguish today’s selling performance from the period’s return burden.
Can bill count divided by footfall be called conversion?
It can be a defined bill-to-entry conversion measure if both datasets cover the same store, hours and population. Explain repeat entries, staff exclusions, counter outages and multiple bills per shopper. Without stable counting rules, the percentage is an operating proxy rather than a reliable estimate of unique shoppers buying.
Which dashboard issue should the regional manager fix first?
Prioritise an issue that changes the business decision: a missing store feed, duplicated bills, an incorrect return join or broken traffic coverage. Cosmetic chart improvements can wait. Once the evidence is dependable, investigate the largest comparable commercial movement and assign an owner with a clear follow-up question.
Does an AI summary remove the need for a metric dictionary?
No. A fluent summary can still mix incompatible sales periods, stock snapshots or customer denominators. Require the proposed interface to state the selected scope, definitions and evidence limitations. Keep the metric dictionary available to the analyst and request a walkthrough of corrections before using generated explanations in operating reviews.
How should EBO teams use AI recommendations for clienteling?
Use AI recommendations as preparation for an associate’s conversation, grounded in permitted purchase history, current product information and the shopper’s expressed need. Separate a suggested product from confirmed stock, an inferred preference from a customer choice, and an assisted sale from proven incremental impact. Evaluate the recommendation, the handover and the final customer outcome as separate stages.
What is the difference between clienteling and a product recommender?
A recommender proposes items or content. Clienteling includes the associate, the customer’s current need, product clarification, store availability, service promises and follow-through. Evaluate whether the recommendation improves that whole interaction. A technically plausible product match can still be poor service when it ignores timing, fit or a recent issue.
Should an associate tell the customer why an item was recommended?
Give a useful, ordinary explanation based on the current request or an appropriate known preference. Avoid revealing unnecessary internal data or presenting an inference as a fact about the customer. The associate should be able to explain the product’s relevance and invite correction without sounding like a profiling system.
Can a previous size purchase be treated as the customer’s current size?
Treat it as context, not certainty. The purchase might have been a gift, a different product fit or an item later returned. Ask the relevant fit question when it matters. Preserve whether the size was explicitly confirmed by the customer and when that information was last updated.
How do we test clienteling when the store network is small?
Begin with recommendation-quality and service-process checks, then design a comparison that fits your appointment volume. Random assignment can help, but contamination across associates and uneven product availability need attention. Report uncertainty honestly. A short pilot can expose operating problems without proving a commercially reliable sales effect.
Should recommendation acceptance become an associate incentive?
Avoid rewarding acceptance alone because it can encourage associates to use unsuitable suggestions. Assess service quality, factual correctness, customer feedback and completed outcomes together. Rejections with clear reasons are valuable evidence. A good operating process lets the associate challenge the recommendation while preserving accountability for the next service step.
Social commerce (18)
How should retailers govern opt-in and follow-up messages in social commerce?
Retailers should distinguish a service enquiry from permission for future marketing, retain evidence of the customer’s choice, and check eligibility again before sending. Connect preferences and withdrawals to every relevant audience, define the owner of each message, and keep platform rules separate from internal campaign decisions. A stored phone number or previous purchase does not by itself establish permission for every follow-up.
Does a customer enquiry permit future promotional follow-ups?
Do not infer a broad future marketing permission from an enquiry. Preserve the original service purpose and check the applicable customer choice and current channel conditions for the proposed next message. The practical test is whether the team can explain why that particular recipient should receive that particular communication at that time.
What evidence should a store retain for an opt-in?
Retain enough to identify the customer and contact point, the relevant purpose, the choice made, when and where it was collected, and the wording or notice version presented. Protect access to that evidence. A bare yes field is difficult to interpret when the collection journey or message purpose later changes.
How should a retailer handle a withdrawal received by a store associate?
Provide a simple recording route that preserves the customer identity, request time and affected communication. Carry the update to every relevant sender and queue, and track failures. The associate should know who owns unresolved propagation. Avoid treating the request as complete merely because one local contact list was edited.
Why check permissions again after the campaign is approved?
The customer’s preference can change between audience preparation and dispatch. A current check prevents an earlier approval from becoming a stale justification for contact. Record the final sendable population and explain material differences from the approved audience. Keep commercial exclusions and channel eligibility current at the same decision point.
Can one vendor setting establish messaging compliance for the business?
A setting can support part of the process, but the organisation still needs appropriate collection journeys, notices, purpose definitions, suppression propagation and operating ownership. Evaluate the actual connected workflow and obtain the relevant internal review. Do not describe a product label or a checked box as proof of compliance across all communications.
How can retailers turn social conversations into reconciled orders?
Create a traceable handoff from conversation to an authoritative order record, then connect payment, fulfilment and refund events using stable identifiers. Keep enquiry, quotation, checkout, paid order and completed sale as separate states. Reconcile values and exceptions at the appropriate grain so an enthusiastic reply, payment screenshot or repeated webhook cannot silently become duplicated revenue.
Is a customer payment screenshot enough to mark an order paid?
Use the approved payment source to verify the event and its relationship to the order. A screenshot can help investigate a customer question but may be incomplete, duplicated or refer to another transaction. Define who can change payment status and retain the verified reference, amount, currency and event time.
What if one conversation creates two separate orders?
Keep one conversation record linked to two distinct order identifiers, with the relevant enquiry or quotation context for each. Do not force a one-thread-to-one-order model. This preserves service continuity while allowing payments, fulfilment and refunds to be reconciled correctly for each separate commercial transaction.
How should cash-on-delivery orders appear in the social funnel?
Give them a distinct payment and fulfilment state supported by the business process. An order placed is not collected cash, and a shipped order is not necessarily delivered. Reconcile delivery confirmation, collection and any provider remittance separately. Use a completed-outcome definition that reflects those differences when comparing channels.
Why might a payment-provider payout be lower than paid orders?
The payout can cover a different event window and include refunds, fees, reserves or other documented adjustments. Reconcile the included payment events before treating the difference as an error. Preserve the provider reference and calculation so the reviewer can reproduce the expected amount and classify any remaining unexplained variance.
Can a social campaign receive credit when checkout happens on the website?
It may receive attribution credit under a documented rule when the relevant source and identity context are available. That marketing interpretation should not create another order or change verified revenue. Preserve the source evidence and attribution settings, and distinguish credited orders from sales proven to be incremental to the campaign.
What is the difference between attribution and incrementality in social commerce?
Attribution assigns credit for an observed order according to a chosen rule; incrementality estimates how many outcomes happened because of an intervention rather than without it. Social-commerce teams need both. Use reconciled orders and explicit attribution settings for channel reporting, then a suitable comparison design to evaluate additional customers, contribution and the effect of campaign cost.
Can attributed revenue be higher than incremental revenue?
Yes. Attribution may credit orders from people who would have purchased even without the intervention. Incrementality estimates the additional outcome against a comparison. The totals can differ substantially without either report being internally wrong. Check definitions, outcome windows and uncertainty before using either number in a commercial decision.
Does a unique coupon prove that a campaign caused the order?
No. A coupon identifies a redemption path but does not show what the customer would otherwise have done. Existing demand can also use the offer. To estimate additional impact, compare appropriately defined outcomes against a credible alternative condition while accounting for eligibility, overlapping promotions, returns and the cost of the incentive.
Should the holdout contain only customers who would open the message?
No. Defining the comparison through a response that occurs after assignment can bias the result. Start with the eligible population and a suitable assignment method, then analyse according to the predefined design. Non-openers and non-buyers are part of understanding the effect of offering the intervention to that audience.
What if our audience is too small for a reliable lift estimate?
Report the uncertainty and use the pilot to learn about delivery, identity, offer relevance and operational friction. Consider a longer suitable observation period or a different design with expert review. Avoid converting an unstable point estimate into a confident revenue forecast simply because the dashboard displays a positive percentage.
How should social messages and paid ads be tested together?
Define whether the question concerns the combined journey or the separate contribution of each intervention. A combined test cannot automatically isolate the effect of the message versus the ad. Track exposure overlap and choose a design that matches the decision, with enough volume and consistent outcomes across the relevant groups.
Malls (12)
How should a mall measure repeat visits from its loyalty programme?
Define a qualifying visit, identify a mature visitor cohort and measure the share returning within the same observation window. Deduplicate bills from the same visit and show participating-tenant coverage. Keep member repeat behaviour separate from total mall footfall and from causal programme impact. Use a suitable comparison to evaluate whether rewards created additional visits rather than merely recording existing shoppers.
Can every bill upload be counted as a mall visit?
No. Several bills may come from the same shopping day, and upload time may differ from the transaction date. Define the visit proxy your evidence supports and deduplicate accordingly. Retain bill, member, tenant and date references so the team can explain the difference between uploaded receipts, shopping days and identified visitors.
Why should recently enrolled members be excluded from a 30-day return rate?
They should remain visible in a separate immature cohort until they have had the full opportunity to return within the defined window. Including them too early changes the denominator unfairly. Once their window matures, calculate the same outcome using the same event definition and reporting cut-off rules.
Does strong member retention prove that mall loyalty works?
It shows observed behaviour among the members represented by the data, but enrolled shoppers may already be more engaged. To assess the effect of a specific benefit or invitation, use an appropriate comparison design. Keep the descriptive retention result separate from claims about additional visits caused by the programme.
How do tenant additions affect the reported repeat rate?
New participating tenants can make more existing shopping activity visible. That can raise measured repeat days without changing total visitation. Show coverage changes and compare stable tenant sets where useful. Keep a full programme view as well, so the business can see both observed growth and the limits of like-for-like comparisons.
What should a mall do when it cannot link entry counts to loyalty members?
Maintain separate reports with clearly labelled populations. Entry counts can describe traffic under the counter’s rules; validated member activity can describe identified programme behaviour. Do not assume one is a subset with a known conversion rate unless the matching and coverage support that calculation. Improve the data contract before combining the measures.
How should a mall improve tenant data quality before analysing sales?
Start with an effective-dated tenant master, a clear reporting calendar and explicit states for missing, rejected, corrected and validated submissions. Preserve source files and sales definitions, then compare equivalent tenants and periods. Show reporting coverage beside commercial movement. A clean-looking total is not dependable trading intelligence when absent tenants, duplicate files or changed definitions remain hidden.
Should missing tenant sales be filled with zero?
No. A missing report means the value is unknown under the current collection process. A zero-sales day is a confirmed reported state with its own evidence. Keeping them separate prevents an absent file from being misread as a trading collapse and preserves the correct owner for the next follow-up.
What is the difference between received coverage and validated coverage?
Received coverage counts expected submissions that arrived. Validated coverage counts those that also passed the agreed checks. A rejected file improves the first measure but not the second. Show both with the same expected-report denominator so management can distinguish a collection delay from a source-quality problem.
Can comparable sales be reported when some tenants are missing?
Yes, for an explicitly defined subset with suitable records in both periods. Label the tenant coverage and exclusions, and keep the incomplete network total separate. Do not imply that the subset establishes the property’s complete movement, especially when missing tenants are concentrated in a large category or a major anchor.
How should a corrected tenant submission affect yesterday’s brief?
Update the current reporting view through a controlled version process while preserving the original submission and the earlier conclusion. Record who approved the correction and why. If the change alters a material management interpretation, communicate the amendment with enough context for the commercial team to revisit the relevant action.
Can AI decide that an unusual tenant value is wrong?
It can be evaluated for flagging an unusual record or explaining why review is needed, but a statistical outlier is not proof of error. Keep the source evidence and agreed validation rules. Financial corrections should follow the authorised review process, with the original value and the reason for any change retained.
Governed AI (6)
How should a retailer evaluate Fundle Brain as a governed AI decision interface?
Evaluate Fundle Brain against a defined retail decision, representative source records and explicit acceptance cases. Check whether the proposed deployment preserves scope, shows evidence limitations, respects user roles and separates explanation from authorised action. Treat connected data, integrations and execution capabilities as items to demonstrate, not assumptions inferred from AI positioning or a polished interface.
Does an AI-powered interface imply that it can execute retail actions autonomously?
No. Explanation, preparation, approval and execution are different capabilities and permission states. Ask which operations are supported in the proposed deployment and require a concrete demonstration of the transition. A generated recommendation or attractive product tour does not establish authority to change rules, send messages or alter financial records.
What should be included in a Fundle Brain evaluation dataset?
Include representative ordinary questions and material edge cases from the intended retail workflow. Useful examples include returns, duplicate rows, stale stock, incomplete tenant reporting, changed preferences and restricted store access. Attach expected evidence and allowed actions to each case so reviewers can assess correctness, usefulness and boundary handling separately.
Is one overall accuracy percentage enough to approve deployment?
No. An aggregate can conceal critical failures within a small but important category. Review numerical correctness, knowledge quality, source limitations, permissions and action handling separately. Define severity and acceptance criteria before testing. The organisation should be able to explain why the remaining errors are acceptable for the particular operating scope.
How should the interface respond when a required data source is missing?
The desired behaviour is to identify the missing scope, avoid inventing the answer and propose an appropriate clarification or handover. Test this explicitly with representative records. The reviewer should understand which conclusion remains supported and which part cannot be completed until the source or definition is corrected.
When should a retail AI workflow be evaluated again?
Rerun the relevant acceptance cases when a change can alter behaviour, such as a source format, business rule, model, knowledge set or permission scope. Keep version and ownership records so the comparison is meaningful. Also review recurring operating exceptions; a stable interface can still become unreliable when its underlying evidence changes.