Please enable JavaScript.
Coggle requires JavaScript to display documents.
2019 Project Plan (Quarterly Goal) (Comp Rater (Tableau / R Shiny…
2019 Project Plan
(Quarterly Goal)
Comp Rater
Tableau / R Shiny
(DONE)
Automation: Automate comp rater report release
(DONE)
Enrich report content: win / bridge / close tie to the rating segment
Enhanced version of sales dashboard for Product Mgmt
Option1: Tableau Reader
1) Create a Product location to save the latest version of exe.
2) install through WESS: "Miscellaneous Software"
Option2: Embed Tableau in a company webpage:
https://community.tableau.com/thread/149800
1) Create Product Folder in Tableau Server (PREPROD, PROD)
2) Enable Guest Access (only available for core license, not available for named user license (it's likely we have the named user license)
(DONE)End of Feb
Upload 1st Production Version of Marketing Sales Dashboard
Create Product Version of Agency Level Sales Dashboard
"Pave the Way to Grow with Profitability"
(may be complex to narrow down the targeted scenarios)
Content1: Create a Full Vision of
1) Rate Competitiveness
2) Agency Engagement
3) Close Rate
4) Loss Ratio
5) UVM Flag
Content2: Quantify the growth impact by action / Map out
1) daily trend
2) easily add / plot event
3)
R Shiny UI: can be used to collect data as well
Ad Hoc
Terminated Agents
Confirm Terminated Agents with Eric Chang / Chris Park; After Dec, if there are new quotes generated (QLC)
(DONE) By end of Feb
Data / Methodology Got!
Agent Visit Feedback - Continue to process
GA win rate assessment - mid-Feb, 2019 new territory factor
(DONE) By Early Feb
Territory Analysis, competitiveness changes
Enhanced Rehab Report - Chris Cook's request, add new attributes, filter application for app review
(HOLD) By End of April
Revisit with Chris (review the attributes), revisit the timeline
Sales Visit Change pre vs. post action
Text Analysis
Process Rater Data & Internal Data
PL Rater
Transaction Data Issues (2018/09 +) Identified (checking with vendor)
By 2/20
(Done, no transaction data issue, error when reading raw data)1. Work with Wendy Haney resolve the transaction data issue
(Done) By End of July
CA HO from PL Rater
Ezlynx
HO
(DONE) By End of Mar
Finish Research Mapping HO data (excl NY) to internal database
(DONE) By End of Jully
CA and xCA HO from PL Rater & EZLynx
PPA
(Done) By End of June
win rate & close if win CA --> integrate to report
AccuAuto
UW Message
By End of April
UW message analysis
High level win rate analysis vs. MR database comparison
Commission by Quote Profile
H2H analysis - vs. PGR
Bridge & Close rate
"Create certain kind of Alert"
Process to capture dramatic changes; automatically remind the users / products
EIS, TPR process: iinternal handling data
IP locator
Daniel Cosentage?
Dave Gram?
Agency Level
PPA
(Done) By early June:
xCA states: analyze agency engagement : when avg agency rank <=3, suggest more engagement?
MR Interface Test
(DONE) From Shaguna's team: create quotes in test environment PL, EZ, AA
Agency Assessment Model
Profitability
Monitoring Process
Quantify Agency Performance Improvement
(HOLD) End of April
Every month refresh the assessment model, deliver agent list with score 2 & their past 12 month history to state PM & PA
Agenda Item Update (Target to enhance after Jan Month End)
for score 4, 5, do not display message (nothing
1/15: (Done) updates made to push different follow up items than same message (i.e. if all 0, push 3 diff items org with 0 sig)
"Nearest Neighbor"
Model for Homeowner
Enhance model setup
weight for segment with high projected loss ratio
provide more information (display: ie. excluded spouse show 20% higher LR
Growth
Ideas:
Bridge & Close Activity (*internal data only)
Demographic information (Population Density, Household, Nearest Neighbor)
Bundled information
Agents appointed with other carriers (win rate data), not-appointed agent's quote volume - market share
Monitoring Process
Quantify Agency Performance Improvement
Model for other LOB (Homeowner?)
New Data Source
SNL - all available agents data even not with MCY is available (obj data, all available agent size, est employees number, dollar amount, address etc.)
End of May
Read the downloaded data and figure out the approach to clean / connect the data with existing ones.
Agency Behavior Monitor
Our Visit Type, Days since last visit, Rep's communication Item
Enhance / Create Report
Comp Rater Report
PPA
(done) End of June
Integrate CA (win, close if win) with AQPR
HO
(done) End of July
Create HO report (PL Rater & EZLynx, CA & xCA)
Enhanced Rehab Report
2019/02/05: remove tab 4 (app review) applications that do not exist in tab 1 (usually are flat cancel with multiple issue date in the system) --> check in NG if there is any cancellation indicator for flat cancel.
(done) End of Feb
Schedule automated run, effective starting from Mar 4 Monday.
(hold) Test Email Notification: prepare for potential future auto email send out
CA: stopped appointing new agents for a period & solid communication with agent about the profiles we will / will not like to write. utilize enhanced rehab report approach, same day / next day email notification if identify profiles submitted directly to rep.
(hold) End of Mar
Discuss with Hao, similar approach as the FNOL. Integrate policy / agent data & create new processes
Other
digital marketing research & win rate info & lifetime value
RPA: robotic ... ... automation
Frequency model for FLV??
shi fou shi yi ge wu qu --> yong fancy de fang fa zuo report?
BA: shopping season?
Ad Hoc Request
By 4/1 (one day)
UW Memo Data - append Agency code and name to UW data (excel)
UW
UVM 2.5
(DONE) Rater Template Creation
(DONE) UW 2.5 Research Data Refresh - CC attributes
Ad Hoc
possible behaviors working around on UVM flags: quote transactions flagged at certain time and change certain attributes let the quote pass through
Fireline
(DONE) Map out fireline scores (returned from vendor); Scores are largely biased based on wild fires in the past 2 years.
Special Hazard Zone
(DONE) Map out SHZ (from verisk); Verisk used to provide SHZ = Yes / No, this is a new attributes further categorize zone 4/3/2/1 & no in zone:
Results look reasonable but a little artificial. The band of area on the edge of montain / hill / fuel location are categorized as zone 4; another band (with solid width) next to zone 4 is zone 3 etc. (75%+ are zone 4 ~ zone1)
Vendor provided "structure density" as well, however, 65%+ are high density <- may used CW as baseline (does not make sense).
Suspicious Agency Quoting Behavior
:red_flag: FL Agency 09F556 automated program quoting every minute
PPA
Customer
same customer name, quoted with too many agents
same customer name, blocked too many times
Agency
Gaming rating systems, dramatic rate volume increase
Gaming rating system / growth opportunity: dramatic customer volume increase
same customer name, quoted too many times by the agent
same customer names, blocked by UW process, continuous testing our system
HO
Customer
Gaming rating system, same customer name, keeps coming back with different address
Address
Same address, keeps coming back with different agents / customer name
HO UVM
research / modeling
Data / Database
ITSR 884267 Populate Code Description Value in MSTR