introducing data analytics 15062015
TRANSCRIPT
Introducing Data AnalyticsBEN RYMER
PRESENTATION FOR RESEARCHERS IN FUNDRAISING
16 T H JUNE 2015
Why prospect researchers need analytics
To increase the value of donations in the UK; no overall rise for decades
Building stronger relationships and increasing donor retention
To improve high-value strategy and product offerings (‘missing’ middle donors and philanthropists)
Maximum HNWI coverage of 50% in screenings, many of the best prospects not Rich List-ers
It need not be rocket science
The robots are coming to get us: automation is real
Picasso, ‘The First Communion’, 1895-6
accessed at www.artexpertswebsite.com/
pages/artists/picasso-gallery.php
Picasso, Self-Portrait with Pallette, 1906, http://www.artexpertswebsite.com/pages/artists/picasso-gallery.php
Jacqueline with Crossed Hands, 1954, http://www.artexpertswebsite.com/pages/artists/picasso-gallery.php
A Lie That Reveals the Truth “It took me four years to paint like Raphael, but a lifetime to paint like a child”: abstraction reveals more than realistic representation
Analytics and modelling seek leverage through stylised, simplifications of things ‘in the world’
However: ◦ Remember models and analytics are useful but not ‘true’◦ Epistemology/assumptions of knowledge◦ Keep the overall objective in mind
Building a Team and CultureA culture of data analytics can be built from the bottom up or middle out
Key qualities are “curiosity, communication and common sense”
Clara Avery, Head of Insight at Macmillan: “we probably called ourselves an evidence-based organisation for two years before we really were one” available at http://insightsig.org/wp-content/uploads/2013/11/6a-Macmillan-love-Insight.-pdf1.pdf
MacDonnell & Wylie: “In our experience, [improved]
analytics has not come from the top. It’s come from
staffers who attend conferences, read books
and blogs on their weekends and conduct side-
projects on their own initiative”
Drivers of Giving (McCoy, 2013) Most predictive models require you to first look at a group of supporters who have already been seen to ‘do the thing’ that you are modelling, eg: make a major gift, and build a picture of who they are
For your own organisation, you need determine what are the significant, defining characteristics of giving across the base, ie: the key drivers?
These characteristics might be demographic, behavioural or attitudinal
The elements that make a good supporter may differ from charity to charity. Certainly the importance, or ‘weighting’, that you put on each element, or variable, will differ greatly
Accessed at http://insightsig.org/networking-events-20122013/
Potential Drivers of Engagement & Value (McCoy, 2013)
Active Committed Giver? Number non-CG GiftsTenure of giving Gender, Marital Status recorded?Current Lifetime Value Number Active RelationshipsFirst & Last Gift non-CG Amount Recruitment SourceGift Aid sign-up Response RatiosQuestionnaire Response ACORN, Mosaic, CameoMaximum Gift Amount Age (capture Date of Birth not Age!)Proximity to Cause Flags from Wealth ScreeningLegacy supporter? ‘Miss’ aged 55+ (for Legacies)Recruitment Date First gift amountEmail opens & click-through Professional titleEvent participation Questionnaire responderRFV LifestageOpt-ins & opt-outs MembershipProperty value Velocity of GivingAverage non-CG Gift AmountDid they inform you of a change of address without being prompted?
Accessed at http://insightsig.org/networking-events-20122013/
MacDonnell & Wylie: “You have
to start by properly framing the question. The
rest is just technique”
Suggested starters
Tenure/continuity of givingGiving velocityRecruitment DateResponse RatiosUnprompted communicationsWealth flags: private bank, property value, equity salesFirst gift amountCurrent Lifetime ValueQuestionnaire responderEvent participant/volunteer
First gift date-presentThis years cash total/av previous three
Date addedAppeals/responsesSum total no. comms
NO MATHS REQUIRED
NO MATHS REQUIREDSum total givingNO MATHS REQUIREDNO MATHS REQUIRED
A Basic Affinity Model Principle of analytical thinking more important than specifics of the method
“Go where the money is, and go there often"
More advanced analysis Regressions
Text Analytics Algorithms
Automated scoring and screening Machine Learning
Resources Kevin MacDonnell’s blog: Cooldata
His and Peter Wylie’s 2014 book ‘Score!’ (ISBN 0899644457) Josh Birkholz: Fundraising Analytics
See list at: https://www.worldcat.org/profiles/BenRymer/lists/3257763
Join Prospect-DMM: scary but well worth it https://mailman.mit.edu/mailman/listinfo/prospect-dmm
Twitter: @joshbirkholz @iofinsight @n_ashutosh
Pitfalls Causality; correlation does not equal causation
MacDonnell and Wylie on roadblocks: “conservative nature of our institutions, a natural preference for intuition and narrative over data and analysis, a skills shortage, a fear of disruptive change, scepticism over the claims made for algorithms and a lack of time and resources”
RFV: only part of the picture
Just because you get the result you want might not mean it is accurate!
It all comes back to data quality: ‘garbage in, garbage out’
Complex maths ≠ better results! Judgement, expertise are key
Summary Analytics offers powerful insight using sometimes simple methods
Huge potential to identify wealth and understand affinity, much of which is already in our supporter base, and a great career move for prospect researchers
And the final word to MacDonnell and Wylie:“Data analysis is a rewarding, challenging, and above
all fun line of work that will provide much value to your employer and a stepping stone in your career in
fundraising to you”
Thanks & Q&Ahttps://fundraisingvoices.wordpress.com/
@benrymer