- HW: e-scooter startup and estimation focused - service has large fluctuation, cannot have more than two jobs, - low-cost service jobs (> 14 milion retail), median salary $10/hr, less insured leading to - are poor jobs inevitable in low-tech service sectors (forced to adopt strategy) - social saftegy net, different companies casestudy - quiktrip having twice avg salary, low turnover rates (stay 8 years avg), price competitive w. walmart - marcadona (3.8% turnover, childcare and training program, 85% full time compared to 40% industry avg), costco, trader joes - theory 1: coincidental (successful *despite* offering ) - doing as they VS theory 2: *integral* to successful strartegy other firms unable/unwilling to follow - boundary conditions for profitability of incorporating good jobs in service (parking attendance) - contagious high quality employees - policy: empower employees to cost reduction cost, expericen impovement (disseminate improvements), cross train - endogenous mechanism: no layoff from fluctuation (productive during low demand), turnover reduction keep high quality and processes - service quality, service capacity (employees * avg (skill * contribution opportunity)), job attractiveness - task richness (engage, cross train) X compensation; performance landscape. CMIN: high turnover, low labor quality, low capability VS IMAX: low turnover, high labor quality, high capability ### fixed demand - Main features affecting whther CMIN VS IMAX are: employee quality, service quality, task richness, learning rate - Peak location sensitivity to context - scenario 1: less synergy (optimal employee quality decrease) - scenario 2: no learning or training (instead of training - need to pay more) - scenario 3: less benefit from quality - CMIN has easier template, locally dominant, - For IMAX: need two nontrival things (high task richness, cross training, buffer, stable schedule, opportunity for layors - motivation, compensation peak); imitating IMAX is difficult ### variable demand: with buffers (no JIT), we can reclaim IMAX - just in time scheduling combined with variability (jobs become hectic, cannot keep high quality) - slow down labor adjustment with reliable schedule (buffer creation): JIT sucks!! ## transition dynamics - worse before better dynamics (should pay more before get better quality - couple of years for effect to kick in) - temporal complexity misleads managers towards CMIN and complicates incentive toward IMAX (delay) ## experiment - 96 time series decision for experiment subject (Hazhir and S. Gary; delay impair learning and can drive convergence to inefficient strategy) - noisy outcome -> more ppl end up in the middle - longer delay -> CMIN >> IMAX - hard to persuade there are high quality employees out there (rejecting applicants), public question, attribution - trust between manager and employees - task richness - employees doesn't wait for worse than better dynamics (narrow band) - store manager VS HQ. - accepting periods of loss (when low demand) - different direction of task standardization and richness - letting local stores have veto power to HQ - herd behavior in an uncertain environment