Data-mining: getting to know your customers and donors to increase their commitment to your organisation

Transcription

1 Data-mining: getting to know your customers and donors to increase their commitment to your organisation Lucy Sinclair Director of Media and Audiences Royal Opera House, roh.org.uk Royal Opera House, London. Lucy Sinclair. opera europa October 2017

2 Revenue from customers and donors is ¾ of our income other Enterprise Public Funding Fund-raising and membership Box Office Royal Opera House, London. Lucy Sinclair. opera europa October 2017

3 Our objectives for audience data 1. Improve how we price our operas and zone our auditorium 2. Improve our understanding of how we use our marketing budget 3. Improve the response we get from our marketing by understanding who to target with our messages 4. Improve our understanding of how we schedule and plan the opera season, so we can maximize audience revenue potential (whilst retaining our artistic integrity) Royal Opera House, London. Lucy Sinclair. opera europa October 2017

4 A whole organisation vision for audience data-mining Royal Opera House, London. Lucy Sinclair. opera europa October 2017

5 We want to understand our digital audience as well MEMBERS FREQUENT ATTENDERS OCCASIONAL OR ONE OFF ATTENDERS CINEMA ATTENDERS PHYSICAL DAYTIME VISITORS LIVE STREAM VIEWERS DIGITAL SHORT-FORM CONTENT CONSUMERS SOCIAL FOLLOWERS INCREASING REVENUE & ENGAGEMENT 2015/16 figures derive from different sources using different methodologies, and are unduplicated. They give a snapshot indication of how, how many and from where our audience consume our branded content and productions. Royal Opera House, London. Lucy Sinclair. opera europa October 2017

6 In 2016, we realised that 82% of our known audience on our database had stopped coming Royal Opera House, London. Lucy Sinclair. opera europa October 2017

7 Things we needed to think about, in order to get better at using data PEOPLE Skills Culture Structure Resource PROCESSES Data strategy Asset management Legal compliance Business rules Media buying Marketing PRODUCT / TECH Database Flexible Pricing Analytics tools Digital marketing tools Royal Opera House, London. Lucy Sinclair. opera europa October 2017

8 PEOPLE Skills Culture Structure Resource PROCESSES Data strategy Asset management Legal compliance Business rules Media buying Marketing PRODUCT / TECH Database Flexible Pricing Analytics tools Digital marketing tools Royal Opera House, London. Lucy Sinclair. opera europa October 2017

9 Our team structure wasn t focused on our customers needs. We had silos, overlaps, and big skills gaps REWARD LOYALTY CREATE OPPORTUNITY CREATE DEMAND ON-BOARD & DELIGHT GROW VALUE REACTIVATE Media L&P Comms D M L&P C Dev M marketing Media DDT m Dev Mktng Dev M Royal Opera House, London. Lucy Sinclair. opera europa October 2017

10 Royal Opera House, London. Lucy Sinclair. opera europa October 2017

11 We now focus on serving the different stages of our audience and customers experience REWARD LOYALTY CREATE OPPORTUNITY CREATE DEMAND ON-BOARD & DELIGHT GROW VALUE Cinemas Television & Radio Big Screens Streaming & VOD R&D New Platforms R&D New Formats Always on dialogue Influencer outreach Publicity & press Omni-channel campaigns Social media Engaging content Consistent brand voice Programmes & materials Web & App exp Ticketing process Purchase path Feedback Dynamic pricing Personalized CRM Up-sales Cross-sales Packages Friend Recruitment Friend On-boarding Friend Benefits offer s REACTIVATE Member mgmnt Friend dialogue Reengage campaigns Film & Broadcast Audiensc e Lab Marcoms Creative Studios Digital products Strategic Analytics & CRM Royal Opera House, London. Lucy Sinclair. opera europa October 2017

12 We added data skills through a mix of training & hiring REWARD LOYALTY CREATE OPPORTUNITY CREATE DEMAND ON-BOARD & DELIGHT GROW VALUE Cinemas Television & Radio Big Screens Streaming & VOD R&D New Platforms R&D New Formats Always on dialogue Influencer outreach Publicity & press Omni-channel campaigns Social media Engaging content Consistent brand voice Programmes & materials Web & App exp Ticketing process Purchase path Feedback Dynamic pricing Personalized CRM Up-sales Cross-sales Packages Friend Recruitment Friend On-boarding Friend Benefits offer s Reengage campaigns REACTIVATE Member mgmnt Friend dialogue Film & Broadcast Audiensc e Lab Marcoms Creative Studios Digital products Strategic Analytics & CRM Royal Opera House, London. Lucy Sinclair. opera europa October 2017

13 Our new in-house data team supports both marketing and fund-raising Head of Strategic Analytics and CRM Senior CRM Manager Digital Analytics Manager Audience Analytics Manager Data Scientist Jan 2018 CRM Officer Audience Analyst Royal Opera House, London. Lucy Sinclair. opera europa October 2017

14 PEOPLE Skills Culture Structure Resource PROCESSES Data strategy Asset management Legal compliance Business rules Media buying Marketing PRODUCT / TECH Database Flexible Pricing Analytics tools Digital marketing tools Royal Opera House, London. Lucy Sinclair. opera europa October 2017

15 Data audit: Where were we starting from? And where do we want to be? Royal Opera House, London. Lucy Sinclair. opera europa October 2017

16 Becoming more audience-focused has required lots of process change! Royal Opera House, London. Lucy Sinclair. opera europa October 2017

17 Royal Opera House, London. Lucy Sinclair. opera europa October 2017

18 Early results are very encouraging 2017 vs 2015/16 Revenue from marketing +55% Revenue due to much more detailed pricing and seat management Summer 2017 delivered our highest revenue and the least discounting since 2012 Royal Opera House, London. Lucy Sinclair. opera europa October 2017

19 Opera Europa 2017 Autumn Conference Data-mining: getting to know your clients and donors to increase their commitment to your company Francesca Di Nuzzo, Senior Research & Consultant Tim Baker, Director and Principal Parma, 12 th October 2017 Baker Richards +44 (0)

20 Our role at the Royal Opera House Bookers and Tickets by Frequency Bookers Churn Sales Index Sales Over Time Price Type Uptake Holds Timeline Price Tracker Hotseat TM Index 20

21 Our role at the Royal Opera House Bookers and Tickets by Frequency Bookers Churn Sales Index Sales Over Time Price Type Uptake Holds Timeline Price Tracker Hotseat TM Index 21

22 Outcomes Understand customer value to increase engagement Festivals, rehearsals, etc. Insights Ancillary sales Opera tickets Recency Frequency Value Ballet tickets Membership Donations Visiting productions External venues 22

23 Outcomes Understand customer value to increase engagement RFV score by Price Schedule RFV score by Area Area by RFV Score Orchestra Stalls CircleGrand Tier Balcony Amphi Slips Stalls 23

24 Outcomes Respond to demand Customer Value Hotseat TM Index 24

25 Outcome Beyond ticket purchase High-Level Members Hotseat TM Index Low-Level Members Hotseat TM Index 25

26 audience insights data business case 26

27 Getting to know your clients to increase their commitment to your company 1. Understanding Customer Behaviour Case Study: Royal Danish Theatre 2. Segmentation: Analysis to Action Case Study: English National Opera Case Study: Metropolitan Opera 3. Beyond data-mining: using primary research to enrich understanding Case Study: Lyric Opera of Chicago 27

28 Understanding customer behaviour 28

29 Scope of Analysis Included in this report Excluded from this report Productions Costs Marketing Activity Culture Segments Catering Sales Customer Satisfaction Survey Repertoire Score Mapping and Catchment Analysis KGL Guides 29

30 In any given year, more than 70% of bookers attend only once, generating around 40% of ticket sales. The majority of tickets are sold to bookers attending more than once. 100% All Bookers and Tickets by Season Year Frequency Band 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Bookers Tickets Once this Year 52,111 39,784 55,399 51,325 53,066 53,107 48, , , , , , , ,750 Twice this Year 8,712 6,374 8,560 8,473 9,162 8,503 8,607 48,399 35,401 52,669 47,980 51,474 48,983 45, times this Year 4,979 3,675 4,536 4,711 5,361 4,398 4,609 49,653 35,872 59,069 44,520 51,205 44,795 43, times this Year 4,162 4,195 3,668 3,842 3,784 3,616 3,615 71,193 67,508 61,116 62,615 64,149 64,112 59, times this Year 1,166 1,093 1,159 1,279 1,434 1,451 1,339 48,632 42,607 51,682 55,337 52,684 60,701 49,563 30

31 Annual churn is high, at just below 70%, including a significant number of customers who had been before. Bookers Churn 100,000 % Churn: 73% 65% 70% 66% 70% 69% % New Churn: 73% 79% 84% 83% 86% 87% % Retained Churn: - 39% 46% 45% 54% 54% 80,000 Bookers 60,000 40,000 20,000 0 Everyone appears new here -20,000-40,000-60,000 Lost - previously new -80, "Lost" New 0-51,764-28,261-39,511-32,131-30,955-28,648 "Lost" Been Before 0 0-7,562-12,149-13,820-19,728-20,513 New 71,130 35,755 46,835 38,829 36,014 33,104 29,039 Been Before 0 19,366 26,487 30,801 36,793 37,971 37,461 Lost - previously been before NB: Churn is the % of last year bookers who didn t come back this year (it is calculated by dividing total lost this year by total bookers last year) 31

32 Churn is strongly correlated with frequency of attendance: those who come more often are also more likely to come back! 100% : Percentage of Bookers and Churn by frequency of visits 100% 90% 90% 80% 80% 80% 70% 70% % of bookers 60% 50% 40% 30% 75% 56% 30% 60% 50% 40% 30% % Churn 20% 20% 10% 0% 9% 12% 5% 6% 5% 1 time 2 times 3 to 5 times 6 to 9 times 10+ times Frequency in % 0% 32

33 Using RFV to identify and understand your most valuable customers RFV Scores Cohorted 26 to to to to 14 3 to 9 Total Total Bookers 5,455 7,477 12,006 31, , ,992 % of Bookers 2% 3% 4% 11% 81% 100% Total Subscribers 4,555 1,656 1, ,031 Subscribers as % of Bookers 84% 22% 14% 2% 0% 3% Total Members 437 1,607 1,411 2,073 3,753 9,281 Members as % of Bookers 8% 21% 12% 7% 2% 3% Total Tickets 687, , , , ,495 2,604,760 % of Total Tickets 26% 15% 11% 14% 34% 100% Average Tickets Total Ticket Value 284,296, ,678, ,101, ,954, ,270, ,302,505 % of Total Ticket value 29% 14% 11% 13% 33% 100% Average Ticket Value 52,117 18,012 8,588 4,036 1,357 3,324 Yield/Ticket Total Membership Value 404, , , , ,683 2,891,574 33

34 A clearly defined Affiliation Programme helps to realise the value of core customers Fixed series: 15% discount: Choose Your Own (6+) series: 10% discount Theatre card (245 kr. Per year): 10% discount Four free drinks per person per season ticket holder 15% discounts on extra tickets 10% discounts on cafés, bars and restaurants 20% discount on all tours Free programmes No-charge exchange policy right up to the day of performance Pre-sale invitations 50% discount on parking Four free drinks per person per season ticket holder 10% discounts on extra tickets 10% discounts on cafés, bars and restaurants 20% discount on all tours Free programmes No-charge exchange policy right up to the day of performance Pre-sale invitations 50% discount on parking Free drinks during intermission four free beverages are already included in the card, which can be enjoyed in all of our cafés and bars. Discount in cafés and restaurants get a 10 % discount in The Royal Danish Theatre s cafés and bars as well as Café Ofelia at The Playhouse for up to 6 persons. Discount on parking get a 50 % discount on parking in Q-Park s car park. 34

35 Segmentation: Analysis to Action (1/2) 35

36 Segmentation: using clustering to identify key customer segments Segment Grand Total Description Infrequent Low yield Mid discount Recent Late booking Infrequent Very high yield Very low discount Lapsing Mid booking Frequent Low Yield Mid discount Recent Late booking Frequent High yield Low discount Recent Early booking Infrequent Mid yield Mid discount Lapsed Very early booking Super frequent Very low yield Very high discount Recent Early booking Infrequent Mid yield Mid discount Lapsing Late booking Infrequent Very low yield Very high discount Lapsing Very late booking Bookers 11,240 12,353 11,476 8,376 4,080 5,245 12,333 21,341 86,444 Average Bookings Average Seasons Attended Average Season Frequency Face Value Discount % 13% 1% 14% 9% 11% 31% 14% 33% 17% Average Value , , Average Season Value Average Yield Max Yield Min Yield Average Weeks in Advance Average Days between Attendances 1,

37 37

38 Segmentation: Analysis to Action (2/2). How to sell 900,000 tickets each year? 38

39 There are only five ways to increase sales! Acquisition Get new customers Reactivation Get lapsed customers to come back Retention Get more of last year s customers to come back 4 5 Frequency Get customers to buy more shows Party Size Get customers to bring more people 39

40 Segmentation to Action (2/2): how to sell 900,000 tickets per year? 40

41 The Met uses 24 Behavioural Segments Clustering Analysis used to identify segments based on a combination of: Ticket purchase, Membership Donations A total of 24 segments was agreed Segments are re-created in the Segmentation Engine 41

42 Group 4: Multi-booking non-subscribers ID Segment # Description / Analysis Objectives / implications 19 Top price singles 17, Last minute bargain hunters 6,496 Made multiple bookings with average Would yield added value / service encourage retention, maximize frequency and generate higher donations? Are they just interested in a good night Mix of booking time mainly early, but out? Identify them in advance to offer significant minority booking late. add-on packages etc. Significant minority choose weekday Key eve focus group. slot. Possible target group for Membership / Donations? Made multiple bookings with an average yield >$150 and booked since They're the 2nd most valuable after Super-frequents and only 30% of them are >$150 donors. Not Only necessarily 30% long tenure of them >50% two are years donors. or less. prices! MixHigh of booking prices! time mainly early, but significant minority booking late. Significant minority choose weekday eve slot. Less than 50% with strong orchestra preference. Made multiple bookings with an average yield <$75 and discount of >50% and booked since Very little donor value. Mainly short tenure. Majority attending 2-3 times p.a. Very low price because buying Rush and other discounted tickets. Mainly booking week of performance. Significant minority choose weekday eve slot. Strong orchestra preference (Rush)! Price promotions n/a. Is there anything that can be done in the way of added value / service that will encourage retention, maximize frequency and generate higher donations? Opportunity here around membership? But how engaged are this group with supporting the Met? Are they just interested in a good night out and willing to pay for it? In which case biggest opportunity is to get more money from them whilst they are attending. Identify them in advance to offer add-on packages etc. Key focus group Does this group have anything in common with high price subscribers? Possible target group for Membership / Donations? How to a) tie them in and b) increase frequency? Could they be encouraged to book any earlier? Combine with 21? But how to get them to take spare inventory? 21 Low price singles 13,569 Made multiple bookings with an average yield of <$75 but without discounts and booked since Very little donor value Mainly short tenure. Mostly attending 2-3 times p.a. Very low price. Large minority choose weekday eve. c. 40% FC preference How to a) tie them in and b) increase frequency? If low price is driving, then possibly elastic around frequency super saver CYO offer? Saver subs? There are probably enough of these kind of bookers to think about introducing some kind of Family Circle club. Can access to the club be driven by frequency, e.g. use appeal of low price FC and access to best seats in that area as incentive to sign up and engage, but would it be better to drive them into the Orchestra and free up capacity in the FC? 42

43 Setting Realistic Objectives for Segments Less than 20 small changes created practical strategies to increase ticket sales by 97,000 in total Retaining 200 more Super Frequent Singles generates 1,969 more ticket sales. Increasing average frequency of attendance of Semi - frequent Singles generates nearly 8,000 more tickets 43

44 Using primary research to enhance data mining 44

45 Using Conjoint Analysis to understand decision making factors 3 separate conjoint analysis studies were undertaken to test the following: 1. Price elasticity for Single tickets in 5 areas of the house for 5 operas 2. Discount elasticity for CYO subscriptions in 3 areas of the house 3. Price elasticity for Fixed Series subscriptions in 3 areas of the house 45

46 Conjoint Analysis: a completely different approach to research Respondents were given the option of choosing between 5 tickets on offer or none at all in a total of 12 different scenarios Title Time Slot Seat Location Price Other attributes tested: CYO: mainly discounting and priority booking Fixed subscription: number of titles 46

47 For all titles, a lower price would generate more sales and for Bohème a significantly lower price might also generate more income. But the same does not apply to less popular titles % Demand 70% 60% 50% 40% 30% 20% 10% Current Start Price $230 to $245 % Demand La Boheme King Roger 0% $125 $150 $175 $200 $225 $250 $275 $300 $325 $350 $375 $400 $425 Revenue Increasing price for all titles beyond current level reduces demand, but beyond the $325 point further increases make little difference, suggesting a small price insensitive segment. Revenue For less popular titles, taking price below $200 may increase demand sufficient to generate greater income but a law of diminishing returns kicks in after that prices are not sufficiently motivating to overcome effect of less popular repertoire $125 $150 $175 $200 $225 $250 $275 $300 $325 $350 $375 $400 $425 47

48 Selected Other Findings Implications for Programming: Demand for La bohème is twice as strong as Falstaff and Queen of Spades suggesting that if everything was priced the same you could offer twice as many performances of La bohème. Balancing performances against this demand results in, e.g. 12 performances of La bohème 6 performances of Falstaff 6 performances of Queen of Spades 4-5 performances of Grimes 3-4 performances of King Roger Everything is Relative: changing one variable affects all of the others: There is consistently greater demand for La bohème and while reducing price for this opera increases demand it has the adverse effect of making the other operas on offer less attractive even if you were to apply a similar reduction in price. Same applies to seating area: one finding was that increasing the price of poor quality seats would make the better seats more attractive This means there are also trade-offs to be made 48

49 Trading off volume for income Modelled Tickets by Scenario* 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1, $500,000 $450,000 $400,000 $350,000 $300,000 $250,000 $200,000 $150,000 $100,000 $50,000 Number of Operas Base 10% 20% 30% Scenario 1 5% 10% 20% Scenario 2 15% 20% 40% Scenario 3 5% 5% 10% Scenario 4 5% 5% 40% Modelled Revenue ($) by Scenario 0 Base Scenario 1 Scenario 2 Scenario 3 Scenario 4 No Operas 8 Operas 1,392 1,232 1,480 1,096 1,580 6 Operas 1, , Operas Operas Operas Opera $0 Base Scenario 1 Scenario 2 Scenario 3 Scenario 4 No Operas 8 Operas $138,930 $145,393 $127,366 $144,221 $136,008 6 Operas $125,500 $117,067 $121,612 $107,209 $99,272 4 Operas $64,985 $70,580 $68,601 $74,203 $56,717 3 Operas $63,524 $68,875 $61,554 $72,378 $65,646 2 Operas $43,895 $47,587 $42,526 $49,985 $45,350 1 Opera $21,951 $23,793 $21,275 $25,003 $22,684 *Based on party size of 1 49

50 Helping cultural organisations achieve their commercial potential Software & Consulting Admissions Pricing Affiliation Segmentation