Clustering Considerations for Customer Segmentation
Description changed:
Welcome to the November meetup for the PyData Edinburgh group!
Delighted that we're back at SolarWinds again (to avoid ending up in the parking lot, don't use sat nav, follow the instruction on our address 😆 )
We'll have our regular update on the PyData community, lightning talks and we're delighted to have Regina Berengolts join us from TVSquared to give our main talk.
We look forward to seeing you there - doors open at 1815, make sure you've clicked attend to reserve your place!
TALK DETAILS
=============
Clustering Considerations for Customer Segmentation
---------------------------------------------------------------
Presenter: Regina Berengolts
Working as the Head of Data Science at TVSquared, Regina rarely gets a chance to work directly with python these days, but that wasn't always the case. Hailing from Canada with a quick one-year layover in Spain, Regina originally came to Scotland to work as a data scientist for a start-up that aimed to change the way retailers and marketers understood their customers. The start-up, a Tech Awards Start Up of the Year Winner, ultimately failed which, for the purposes of this talk, makes Regina absolutely giddy at the opportunity to talk freely about lessons learned, both from successes and more importantly, from the problems inherent in trying to disrupt the status quo.
Description:
It's probably a fair statement to say that most retailers have performed some form of customer segmentation in order to identify their high value customers and drive business decisions to optimise that value (much like pokemon, sometimes you need to sacrifice your less valuable customers in order to capture the more powerful ones). But what happens if "high value" is defined incorrectly? What happens when a company is built on an incomplete and ultimately misleading understanding of the customer?
We'll explore how customer segmentation models are typically built, how considering even one or two "uncommon" variables can dramatically change the segments more so than any change in the clustering algorithm itself, and the downstream effects on businesses that are built entirely around a poor definition of customer value.
LIGHTNING TALKS
==================
A little change this month...read on...
We really want to encourage more people to 'get talking' and so we'd like to get people to pre-register for the lightning talks - there's more opportunity for people to prepare & we can support any new speakers by giving advice on presenting, topics etc. As such, we'd also like to focus the talks a little more, keep them more technical and related to python & data science - but there is still a lot of scope there so please do just get in touch if you have an idea!
So if you want to give a lightning talk at this event, send us a message either here on Meetup, through Twitter @PyDataEdinburgh or any other way you know how to find us!
Still 5 minutes, still informal, no need for slides unless you want them - go on, give it a go!
LOGISTICS
===========
Please note all attendees MUST sign up to the Meetup to be admitted.
1815: Doors open, drinks & networking
1845: Talks start - welcome & community announcements followed by our main speaker, then lightning talks.
We'll end the evening with pizza and more refreshments, wrapping up by 2100.
SPONSORS & REFRESHMENTS
=============================
As always, we couldn't do this without our sponsors help to provide a venue and drink & pizzas — Cathcart Associates (https://www.cathcartassociates.com), Solarwinds (https://www.solarwinds.com) and Canon Medical (https://research.eu.medical.canon)
CODE OF CONDUCT
====================
The PyData Code of Conduct governs this meetup. ( http://pydata.org/code-of-conduct.html ) To discuss any issues or concerns relating to the code of conduct or the behavior of anyone at a PyData meetup, please contact NumFOCUS Executive Director Leah Silen (+1 512-222-5449; leah@numfocus.org) or the group organizer.
