Correct -
Think of the Historical Data Import tool as importing data for specific route paths. So it is up to the customer to provide that data in the exact format that GC needs for a successful import. (Which may require your assistance as a partner, depending on the customer's data and technical expertise.)
For example, at Avtex, we have a query that we can run for legacy PureConnect customers on their iWrkgrpQueueStats view. This query does several things, but the big ticket items are:
--selects only the data necessary for the GC import.
--realigns the headings to match that of the export template file.
--breaks up the 30 minute intervals to 15 minutes.
--splits up any metric data to the 15 minute intervals.
--realigns the new media type names for GC
--breaks up realigns the PureConnect skills used for languages as ACTUAL Genesys Cloud languages
We still have to take that export and
manually align any new queue names, skills, and languages as they were built in GC. But for that, the client is required to provide us a mapping of the old PureConnect queues to the new Genesys Cloud Queues. And the same for PC -> GC Skills, etc. This can be an arduous process, but having actual accurate forecasts and schedules for Day01 for clients, and allowing our consultants to build out actual route paths during implementation (as opposed to waiting until after go live) is paramount.
And to Jay's point, what you are importing with the Historical Data Import tool isn't seen by GC as an "aggregate of everything that happened in that interval". GC sees it as "ACTUAL data for that interval's specific route path(s)."
------------------------------
Trent Vance
Avtex Solutions, LLC
------------------------------
Original Message:
Sent: 09-08-2022 09:49
From: Aleksandar Narancic
Subject: Historical Data Import - Forecasting duplicates or not?
Hi Community,
Just wanted to do a sense check on something bothering me for some time now. Namely, we are working on client implementations of GC and WEM and one of the typical requests our clients have is to import historical data into WEM in order to get the possibility to forecast with high accuracy from day one. Not all of these clients have historicals that match the Genesys setup 100% in terms of Queue-Skill-Language. Many clients have only Queue level data thus their data import csv will probably look something like this:
Historical Data |
Queue |
Media |
Skill |
Language |
Offered |
Queue 1 |
Voice |
None |
None |
30 |
Now let's imagine that their Genesys setup contains 2 skills (Billing and Claims) and no Language for the sake of simplicity. When we import the data into the system, their first month's forecast should look something like this:
Forecast month 1 |
Queue |
Media |
Skill |
Language |
Offered |
Queue 1 |
Voice |
Billing |
None |
0 |
Queue 1 |
Voice |
Claims |
None |
0 |
Queue 1 |
Voice |
None |
None |
30 |
Nonetheless during their first month of production, their actuals will differ from the historical data import since their current flows always assign a skill to every inbound interaction thus their actuals will look as follows:
Production month 1 |
Queue |
Media |
Skill |
Language |
Offered |
Queue 1 |
Voice |
Billing |
None |
15 |
Queue 1 |
Voice |
Claims |
None |
15 |
Queue 1 |
Voice |
None |
None |
0 |
If we look at the totals, the Forecast achieved 100% accuracy and the client seems to be set and ready to continue using the platform. On the other hand, if we keep the historical data import in the system, next month's forecast will be as follows:
Forecast month 2 |
Queue |
Media |
Skill |
Language |
Offered |
Queue 1 |
Voice |
Billing |
None |
15 |
Queue 1 |
Voice |
Claims |
None |
15 |
Queue 1 |
Voice |
None |
None |
30 |
Right now the system has actual live data to forecast on and can't understand that the historical data import was an aggregate of the two rout paths that are live thus there is a possibility that it will forecast duplicate values if we don't purge the data import before we forecast for month 2 (or any other interval of choice).
This is a very specific situation, but many Call Centers being onboarded on GC simply did not have the exact same architecture as they want/need and as they will have in GC, thus my question is if there is any way to mitigate this issue other than purging the data import since the clients most definitely need as much data imported into the system as possible so that WEM can get a hold on the seasonality.
Looking forward your thoughts on this.
Best
Aleksandar
#Genesys Cloud CX
#Forecasting
------------------------------
Aleksandar Narancic
Star Telecom
------------------------------