RestOps
KPIs
Products
Stakeholders
Vendor Fail Rates
Offline Hours
Trends & reasons by country (which fail rates are top for each)
Consistently low
Increasing
Decreasing
Consistently high
Vendor Monitor
Vendor Score / CRC
GoDroid
Restaurant Portal
Things to consider
Constraints
Like rider score – find pain points for different countries and incorporate into vendor score
Countries resistant to implement as it increases offline hours, which impacts growth. COVID helped push out VM, however have high OH now which is a huge pain point
Constraints
Global
Local Commercial
Regional Commercial
Purpose
Reduce / Maintain low fail rates
Offline Reason Definitions [RANKING TBD - depends on dashboard]
Vendor Offline Themselves
Increase in OH impacts growth; commercial does not like high OH
Relationship to Fail Rate
indirect KPI of sorts: low offline hour >> helps implement monitor >> keeps our fail rate down
Context
One of JJ's top 3 priorities for 2020
Have to keep VM rolled out at 100%, because countries are pushing back due to increased offline hours (countries are more incentivized by this than fail rate).
Monitor
Frank: POC
PM: Azucena (previously Alex)
Features
Unreachable Offlining
Varied unreachable period thresholds for countries before they are offlined
Long-term aim is to bring this down. Countries have pushed-back, hence we compromised by increasing the thresholds
Key Definitions of Vendor Fails (ranked by APAC levels)
Product OOS Feature
Purpose
Reduce "Product Unavailable" fail rate
Purpose
Reduce cancellations i.e. fails
Vendor Back End
High Fails
VENDOR UNREACHABLE: if the order was sent by fp but didnt arrive on the tablet because of connectivity issues
VENDOR NOT RESPONDING: 'Restaurant cannot be reached via phone to accept an order after it goes to 53'. After 5 minutes, goes to 53. OPT will call the vendor to accept the order or cancel. We send the order to the vendor device, and give some time for the order to be accepted, (e.g. 10 minutes), after which the order will fail
Low Fails (~0.0%)
WRONG MENU: Order contains item from an old/outdated/unavailable menu. E.g. item from a lunch menu at dinner time.
PRODUCT UNAVAILABLE: single or multiple items unavailable
Medium Fails
TOO BUSY: Restaurant too busy to process this specific order
TECHNICAL PROBLEM: Restaurant cannot process the order because of technical issues, e.g. with the tablet. Can occur before and after accepting the order (status 42).
(PU ONLY) DELAY CUSTOMER LEFT: Pick-up only - customer arrived at the restaurant on time but left again, because the restaurant took too long to prepare the pick-up order
(VD ONLY) WEATHER PROBLEM: restaurant cannot deliver because of the weather situation
(VD ONLY) NO DELIVERY TO THAT ADDRESS: restaurant cannot deliver because the customer address is outside of the delivery area of the restaurant
(VD ONLY) DELIVERY RIDER UNAVAILABLE: restaurant cannot deliver because has no riders available.
(VD ONLY) INCOMPLETE ADDRESS: restaurant could not find customer'`s address.
(VD ONLY) UNABLE TO FIND CUSTOMER: restaurant could not find the customer.
(VD ONLY) RIDER ACCIDENT: Restaurant rider had an accident.
VENDOR CLOSED: not busy, but out of operating hours
DEAL PROBLEMS: Restaurant does not wish to work with us any more and wants to be taken offline.
OTHER: Outdated - not used in backend dropdown: But can still be send by legacy vendor devices (Vendor app, printer, POS)
Monitor Offlining
Contact Center
NA
VENDOR DEVICE: Closed hours due to vendors setting themselves as temporarily offline on device
ORDER DECLINED: Closed hours due to order cancellations by vendor
DEVICE OFFLINE: Closed hours due to vendor device not being properly connected or logged in
VENDOR COMPLIANCE: Closed hours due to compliance by vendor
VENDOR COURIER DELAY: Closed hours due to vendor delaying couriers
OTHER: Closed hours due to other Vendor Monitor actions
CC Special Day: Closed hours due to CC agents setting vendors as special days
CC Temporarily Closed: Closed hours due to CC agents setting vendors as temporarily closed
UNKNOWN: Closed Hours due to unknown sources
Dashboards
CANCELLATIONS DASHBOARD: All countries, distribution of fail rate by reason
[has errors - being fixed] LOGISTICS WEEKLY KPIS: Overall vendor fail rate, by country WoW
Dashboards
[superseded] LOGISTICS WEEKLY KPIS: Vendor offline %, by country, WoW
RESTAURANT OFFLINE HOURS %: Vendor offline % by reason, country https://tableau.deliveryhero.net/#/site/Foodora/views/RestaurantOfflineHours/EvolutionoverTime?:iid=1
Source of truth - does not match with logistics KPI tableau, or tableau below
Unreachable Hours
Definition: defined as when vendors device fail to provide a ping to the RPS heartbeat service (Monitor) for 3 consecutive minutes
Dashboards
Context: aim is to reduce unreachable hours to avoid vendor unreachable fails
Trends by Country
[superseded??] Offline Hours %: by vertical, day, country
Zombie Vendors: 0 orders
Dashboards
VENDOR WEEKLY PERFORMANCE: vendors with 0 successful orders in last week https://tableau.deliveryhero.net/#/site/Foodora/views/APVendorPerformanceWeekly/VendorSummary?:iid=2
VENDORS DEVICES DASHBOARD FOR NON-ZOMBIE VENDORS: https://datastudio.google.com/u/2/reporting/1tOTA4eUYbo3OIKSucTJsgKCQnlGGA6t4/page/I4bDB
[superseded] COMMERCIAL VENDOR PERFORMANCE: % vendor fails + % closed hours
DATA STUDIO - OFFLINE HOURS VS FAIL RATES: by country, daily, vendor fails vs overall fail
TOTAL OFFLINE HOURS %: https://datastudio.google.com/u/2/reporting/1d03icptIOTuA1ZXJsACUv0q8a0D5J01T/page/5MkMB
MONITOR UNREACHABLE OFFLINING HOURS %: https://datastudio.google.com/u/2/reporting/1d03icptIOTuA1ZXJsACUv0q8a0D5J01T/page/s0ONB
[TO BE EDITED AFTER CHECKING WITH HJ] OFFLINE HOURS % BY REASON: https://datastudio.google.com/u/2/reporting/1d03icptIOTuA1ZXJsACUv0q8a0D5J01T/page/hRmMB
% UNREACHABLE DURATION AND % OFFLINE DURATION: %, hours, number of events, number of vendors. by day, country. https://datastudio.google.com/u/2/reporting/1fp3Yw_qKk5SPvCASUL2L7xaLQUDfPr9_/page/MggEB
% UNREACHABLE DURATION AND % OFFLINE DURATION: %, hours, number of events, number of vendors. by day, country. https://datastudio.google.com/u/2/reporting/1fp3Yw_qKk5SPvCASUL2L7xaLQUDfPr9_/page/MggEB