Waste and Root Causes in a Drive-Thru Peak
At a busy cafe drive-thru, average order-to-handoff time during the 11:30-13:00 peak is 7.5 minutes against a 4-minute target. A Yellow Belt timed 240 orders and found that 38% waited for food after paying, remakes were 6%, and fries were batch-cooked every 8 minutes. Two staff take orders, three work the kitchen and two handle payment and handoff, but during rushes the window person often leaves the window to fetch items. The menu has 46 items, and several share similar packaging. A partial 5S effort has left some prep items without labeled homes. The shift manager says, 'The team just needs customer-service training and to move faster.' A fishbone diagram lists people, method, machine, material, measurement and environment branches, but the team has not verified which causes matter. The regional manager wants a solution list next week and has already asked whether a new fryer would fix the problem.
1. Classify the main wastes visible in the drive-thru and prioritize which one the team should investigate first, using the 240-order data.
Read the worked answer
Wastes: waiting - 38% of orders wait after paying; overproduction - fries batch-cooked every 8 minutes regardless of demand; defects - 6% remakes; motion - window person leaves the window to fetch items; inventory - prep items without labeled homes and possibly excess prepped food; overprocessing - similar packaging for 46 items may force extra checking; transportation - items fetched from kitchen to window. The data point to waiting as the largest customer-visible waste, but waiting is likely a symptom of overproduction (batch fries), layout/motion (fetching) and defects (remakes).
Prioritize by impact on the 7.5-minute TAT and by evidence: build a Pareto of the 240 orders' delay reasons (waiting for fries, waiting for other food, remake, payment, handoff). If waiting-for-food dominates and fries are a large share, investigate the fry batch policy and demand signal first. The 6% remakes are smaller but costly and may contribute to congestion. Do not start with a new fryer; that is a capital solution that may not address batching or layout. The first step is to quantify delay minutes by reason and by station, then select the waste with the largest verified time contribution.
2. The team's 5-why ends with 'staff are not following the procedure.' Explain why this is a weak root cause and rewrite the chain to reach testable process causes.
Read the worked answer
Ending at 'staff not following procedure' is a human-error stop, not a root cause. It does not explain why the procedure is not followed, so it invites training or discipline rather than process change. It also ignores the possibility that the procedure is unclear, impossible at peak, or not the actual standard. A good 5-why should reach causes that can be changed and verified.
Rewrite: Why did orders wait after paying? Because food was not ready at handoff. Why? Because fries were cooked in batches every 8 minutes. Why? Because the kitchen used a fixed timer rather than demand. Why? Because no signal told the fry station when an order entered. Why? Because order entry and kitchen are not linked and the standard work does not define a replenishment trigger. That is testable: add an order signal or cook-to-demand trigger and measure wait time.
For the window person leaving: Why? Items were stored away from the window. Why? No labeled home and the layout puts them in the kitchen. Test by relocating high-run items. Verify each cause with observation before changing.
3. The shift manager proposes customer-service training and a reminder sign; the regional manager asks about a new fryer. How would you evaluate both proposals and what would you pilot?
Read the worked answer
Evaluate against the verified causes and the data, not against how familiar the solution is. Training and a sign address 'people' and may improve greetings, but they do not change the 8-minute batch cycle, the window layout or the missing replenishment signal, so they are low-leverage for TAT. A new fryer adds capacity, but if fries are batched and demand is not signaled, it may simply produce more waste; it is also capital-intensive and slow. Neither proposal is proven by the current data.
Pilot a process change first: a visible order signal to the fry station and a small-batch, cook-to-demand trigger for the top-selling items; relocate the most-fetched items within reach of the window; define peak roles so the window person does not leave. Run for two weeks at the same peak hours, measure median and 90th-percentile TAT, % waiting after payment, remakes and fries discarded. Compare against the 240-order baseline, stratified by hour. If capacity is still the constraint after batching and layout are fixed, revisit equipment with data. This sequence tests the cheaper, higher-leverage causes first.