SLLC Intermediate common mistakes
SLLC Intermediate · Understand the traps and how to avoid them.
Treating a forecast-accuracy metric as an end in itself: reporting MAPE of 12% as 'good' without checking the forecast bias (mean signed error) or whether the errors concentrate on high-volume SKUs, then recommending a new forecasting model as the fix.
Candidates memorise that MAPE is a standard accuracy measure and stop there. MAPE is asymmetric and blows up near zero demand, and a low MAPE can hide a consistent under-forecast that drives stockouts on the items that matter. The exam often gives a table of actual versus forecast and expects you to compute bias and recognise that a systematic one-directional error points to a process or bias problem, not a model-choice problem.
What to do instead
For any forecast dataset, compute three things before concluding: MAPE or MAD for magnitude, mean signed error or tracking signal for bias, and the error broken out by SKU class or volume band. If bias is large and one-directional, state that the fix is process (sales incentives, judgemental override discipline, demand review) rather than a better algorithm. Practise interpreting a tracking signal drifting beyond plus or minus 4 as a signal to act.
Answering an inventory-policy question with an EOQ or reorder-point formula applied to the wrong cost structure, for example using EOQ where demand is lumpy or where quantity discounts and freight breaks change the total-cost curve, or ignoring that safety stock is driven by service level and demand variability rather than by average demand.
Formulas are the easiest thing to revise, so candidates reach for them first. Exam scenarios are usually built so that the textbook formula is only the starting point: the question hinges on which costs are relevant, or on the difference between cycle stock, safety stock and pipeline stock. Candidates who plug numbers without classifying the cost or stock type produce a plausible but wrong answer.
What to do instead
Before calculating, label the stock type the question is about (cycle, safety, in-transit, anticipation, decoupling) and list which costs are relevant and which are sunk. For safety stock, write the formula as z times the standard deviation of demand over the replenishment lead time, and check whether the scenario gives demand variability, lead-time variability, or both. If quantity breaks exist, compute total cost at each break point rather than using EOQ directly.
Confusing Incoterms responsibilities: assuming the seller always arranges main carriage and insurance, or stating that FOB means the buyer bears risk from the seller's door rather than from the point of shipment on board the vessel.
Incoterms are memorised as a list of three-letter codes without the risk-transfer point attached to each. Candidates mix up the cost-transfer and risk-transfer points, which differ under C-terms (seller pays main carriage but risk passes at shipment). Questions on landed cost, insurance obligation and customs responsibility are then answered with the wrong party.
What to do instead
Learn each rule as a pair: where cost transfers and where risk transfers. Anchor the C-terms by remembering that risk passes at the point of shipment even though the seller pays freight to destination. Know that only CIF and CIP require the seller to obtain insurance, and that DAP leaves import clearance with the buyer while DDP places it with the seller. When a question asks 'who bears risk at X', answer with the named place, not the country.
Designing a network by comparing transport cost per shipment in isolation, or choosing the lowest-cost single facility location, without netting the change against inventory, warehousing, service-level and lost-sales effects, and without testing sensitivity to demand or fuel assumptions.
Network-design cases present rich cost tables and candidates anchor on the largest visible line item, usually transport. Total landed cost and cost-to-serve thinking are what the exam rewards, but they require the discipline of building a full cost stack. Candidates also skip sensitivity analysis because it feels like extra work, yet many questions hinge on which assumption the recommendation is most fragile to.
What to do instead
For every network option, build a cost stack with at least five lines: inbound freight, outbound freight, facility fixed and variable cost, inventory carrying cost (including pipeline and safety stock changes), and service or lost-sales penalty. Then state which single input, if wrong by 20%, would flip the decision. If the question gives a service-level constraint, check it explicitly rather than assuming cost minimisation satisfies it.
Treating supplier selection as lowest-price bid selection, or evaluating suppliers only on price and lead time while ignoring total cost of ownership, capacity, financial health, quality history and compliance or ESG risk, and then failing to distinguish a supplier audit from a supplier scorecard.
Procurement questions often present a bid table where the cheapest supplier has a hidden risk, such as thin capacity, a single production site, or no quality certification. Candidates who rank on price alone miss the intended answer. The audit versus scorecard confusion arises because both are supplier-management tools and candidates do not separate their purpose, timing and output.
What to do instead
Score suppliers on a weighted matrix that always includes total cost of ownership, quality and delivery performance, capacity headroom, financial and geographic risk, and compliance or sustainability criteria. State the weights and justify them. Keep the tools distinct: an audit is a point-in-time verification of capability or conformance, usually before award or after a problem; a scorecard is a periodic performance review against agreed KPIs that feeds development or exit decisions.
Confusing warehouse layout and materials-handling choices: recommending automated storage and retrieval for a low-volume, high-variability operation, or choosing a handling method without matching it to unit load, throughput, order profile and building constraints, and without linking slotting to ABC velocity.
Candidates learn equipment categories as a menu and pick the most advanced-sounding option. Exam scenarios typically constrain the answer through throughput rate, SKU count, cube and weight, order profile (each-pick versus pallet), and available capital. Ignoring those constraints produces an over-engineered answer. Slotting is often overlooked entirely even though it is the cheapest lever in most warehouses.
What to do instead
For any handling or layout question, first write the operation's profile: SKUs, orders per day, lines per order, unit load, cube and weight, peak-to-average ratio, and building height and floor limits. Then eliminate options that fail a hard constraint before comparing cost. Always check whether the question rewards slotting by velocity (ABC) and whether pick-face replenishment or cross-docking would relieve congestion before new equipment is bought.
Answering customer-service questions by promising the highest service level everywhere, or treating on-time-in-full as a single number, rather than segmenting service by customer or product profitability and defining OTIF with explicit, stated tolerances.
Service level sounds like a virtue, so more is assumed to be better. In practice the exam tests cost-to-serve and segmentation: the right answer usually differentiates service by segment and recognises that raising service on low-margin tail items destroys value. Candidates also quote OTIF without defining the delivery window, the completeness rule and the measurement point, which makes the metric unfalsifiable.
What to do instead
Define OTIF precisely in your answer: measurement point (order, line or unit), the promised window, whether partial delivery counts, and how substitutions are treated. Then segment customers by profitability and strategic value and set differentiated service targets with the cost of each tier stated. Where a question offers a uniform 98% target, check whether the scenario's data supports tiering instead.
Treating a system or analytics question as a software-feature question: recommending an ERP or WMS module as the solution to a data-quality or process problem, or presenting a dashboard or AI forecast without naming the data source, the master-data owner and the decision the output changes.
Technology questions are easy to answer generically, and candidates default to 'implement a system'. The exam usually embeds a root cause such as duplicate item masters, inconsistent units of measure, or missing event data, which no system purchase fixes. Analytics questions also test whether the candidate can connect a metric to a decision, and feature lists do not do that.
What to do instead
Before recommending technology, name the decision being improved, the data required, its current source and quality, and the owner of that data. Diagnose master-data problems explicitly: item, supplier, location and unit-of-measure consistency. For any analytics proposal, state the baseline metric, the expected change, how it will be measured, and who acts on the output. If the scenario shows dirty data, the correct first step is data governance, not a new platform.
Handling risk and resilience questions by listing generic risks, or by treating a single-sourcing decision as automatically wrong, instead of quantifying exposure, identifying the specific failure mode and time-to-recover, and choosing a mitigation whose cost is proportionate to the expected loss.
Risk is the domain where candidates most often write broad statements ('diversify suppliers', 'hold more inventory') without connecting to the scenario. The exam rewards structured analysis: probability and impact, time to recover versus time to survive, and the trade-off between resilience cost and service benefit. Single-sourcing can be correct when capacity, quality or tooling economics dominate, provided a mitigation exists.
What to do instead
Use a time-to-recover versus time-to-survive comparison for each critical node, and map tier-two and tier-three suppliers where the scenario provides them. Quantify exposure as probability times financial impact plus service impact. Then choose mitigations that match the failure mode: dual sourcing for capacity loss, buffer or safety stock for lead-time variability, alternate routing for transport disruption, and contractual remedies for quality failure. State the cost of each mitigation against the exposure it reduces.
Ignoring sustainability and compliance obligations as 'soft' topics: omitting scope 1, 2 and 3 boundaries when asked to measure emissions, or answering a compliance question without identifying the specific regulation, reporting obligation, documentation and audit trail that applies, and who in the chain is responsible.
Candidates treat ESG and compliance as narrative sections and write aspirational statements. Exam questions in this area are usually precise: they ask you to classify an emission source, identify the responsible party for a customs or product-compliance obligation, or specify the records needed to evidence conformance. Vague answers score poorly because they cannot be checked.
What to do instead
For emissions, always state the boundary: scope 1 direct fuel and process emissions, scope 2 purchased energy, scope 3 upstream and downstream value-chain emissions, and note that logistics emissions usually sit in scope 3 for the buyer. For compliance, name the obligation, the responsible party, the documentation and the retention period, and the audit or verification step. Where the scenario involves cross-border movement, link the compliance answer to the Incoterms allocation of clearance responsibility.