Guide Dogs Maximising the impact of Raffle appeals Kate Parker Senior Insight Analyst.

Guide Dogs

Maximising the impact of Raffle appeals

Kate ParkerSenior Insight Analyst

Guide Dogs Raffles - Overview One of our longest running appeals

Quarterly appeals

Sent over 3m warm mailings in 2011

To 880k warm supporters in 2011

Raised £5m in 2011 (warm and cold)

The Creative

The Problem…

Mass communication

Not targeted

Great way to keep in touch with supporters

Potential income slipping through

The Proposal…

Identify max profit line and mail minimum

Send more books to supporters who are most likely to buy more tickets

Use data to personalise the letter to make the supporter feel more engaged with the cause

Part 1- Modelling

Analyse past campaigns : – Who we mailed– Who responded– What differentiates them from the base?– RFV variables & demographics– Identify max profit line

Use the model to select lower volumes to increase RR% and ROI.

Modelling – Profiling responders

Modelling – Decision Tree

afd

Modelling – Gains Chart

Modelling - Results

0

50,000

100,000

150,000

200,000

250,000

300,000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Model node

Maila

ble

volu

me

0.00%

5.00%

10.00%

15.00%

20.00%

25.00%

Resp

onse

rate

%

Mailable volume Typical RR%

Part 2 – Multibook Strategy

Increase average gift by sending more tickets

Which supporters have the propensity to buy more?

How many more tickets do we send them?

Who are the ‘stubborn’ supporters?

Multibook Strategy - Planning Which audience?

– Non-rafflers were tested with multiple books in Summer & Winter 2010

– Those sent multiple books underperformed

Are we potentially putting off non-rafflers by mailing them too many books?

Focus on best responders to increase average gift

Multibook Strategy – The strategy

Multibook Strategy – The strategy

Multibook Strategy – The results

Spring 2011

Summer 2011

Autumn 2011

Winter 2011

New to 2 NO TEST -9.35% -8.24% -5.52%

1 to 2 -6.42% -10.98% -14.04% -18.81%

2 to 3 2.02% -2.48% -0.73% -12.81%

Response rate decreased for almost all test data

However, average gift was much higher across the board.Spring

2011Summer 2011

Autumn 2011

Winter 2011

New to 2 NO TEST 27.41% 25.64% 25.72%

1 to 2 22.08% 26.60% 19.18% 17.37%

2 to 3 7.66% 6.94% 5.58% 12.45%

Multibook Strategy – The results Need to examine ROI to determine effectiveness Test segments performed with mixed results Consistently positive results for New Rafflers – increased ROI for

those sent more books Increasing books for existing supporters looked positive at first,

but then decreased ROI

Spring 2011

Summer 2011

Autumn 2011

Winter 2011

New to 2 NO TEST 4.75% 4.85% 8.44%

1 to 2 3.65% 2.22% -6.82% -13.01%

2 to 3 4.96% -0.38% -3.90% -9.81%

Multibook Strategy – The results Moving poor responders down meant response

rate for other cells dramatically increased

Autumn 2010 Winter 2010 Spring 2011 Summer 2011 Autumn 2011

Resp

onse

rate

2 books

3 books

Multibook Strategy - Conclusions Existing raffle supporters are happy with the

level of gift they have given – maintain & monitor

New raffle supporters can be encouraged to buy more without negatively affecting ROI

Moving supporters down in the strategy strengthened results

Maximising Raffles - Conclusions Multi-faceted approach is best

Data and analysis needs to be robust to accurately predict behaviour

Success is measured in various ways – identifying this from the outset helps your strategy

Any questions?