Best Practices in Forecasting & Optimization
2 likes1,042 views
The document presents a webinar by m/a/r/c research focused on best practices in forecasting and optimization, covering various methodologies and key requirements for effective volumetric forecasting. It highlights significant opportunities for optimization, critical pitfalls to avoid in forecasting, and the importance of differentiation and competitive context. The session aims to provide marketers with actionable insights to improve their forecasting practices.
Best Practices in Forecasting & Optimization
- 1. Presented by M/A/R/C Research® Sponsored by Quirk’s ® Best Practices in Forecasting and Optimization March 9, 2011 Page 1
- 2. Presenters Amy Barrentine EVP – General Manager Randy Wahl EVP ‐ Advanced Analytics Scott Waller Vice President – Business Development Page 2
- 3. Who is M/A/R/C Research? 46 years of research service and innovation Industry experience includes… Consumer Packaged Goods Pharmaceuticals and Healthcare Telecommunications and Technology Dining and Hospitality Retail and Financial Services Part of the Omnicom Group Page 3
- 4. Objectives Today we will discuss… …the range of volumetric forecasting approaches that marketers use …key requirements in a custom, buyer‐based system …things to avoid in forecasting …big opportunities to fine tune through optimization Page 4
- 5. Today’s Agenda Frequently Asked Questions Range of Methods Key Requirements Big Opportunities to Optimize Things to Avoid Forecasting Fiction Q&A Page 5
- 6. Frequently Asked Questions Can I get a copy of today’s presentation? Yes, a copy will be emailed to you Is today’s webinar being recorded? Yes, downloadable Can I ask questions during the event? A Q&A session will commence at the end of the presentation Page 6
- 7. Range of Volumetric Forecasting Approaches Marketers Use Page 7
- 8. Forecasting Approaches Primary Examples Pros Cons Methods Assessment is fast No road map Qualitative Anecdotal Easy buy in Easily understood Adjustment for offering Analog discrepancies Observable Historical Can encompasses large Backward looking Econometric/Time # of variables Unexplained variables Series disregarded Provides Marketing Mix Direction Estimate across many Predictive within test range Choice launch scenarios Calibration required Survey Response Norm Comparison Easy to understand Static context (inflexible) Based Collective history Limited to experience/ category availability Decision‐Driver, Adaptable Hurdles provided, but Self Calibrating Innovative offerings/ norms don’t apply emerging categories Page 8
- 9. Key Requirements in a Custom, Buyer‐Based System Page 9
- 10. Key Requirements Validated Methodology “Well, most of the time we’re right!” Competitive Context Included ~ Consideration of new offerings vs. other in‐market options is important… Page 10
- 11. Key Requirements Marketing Spend Levels Estimated and Incorporated Marketing Spend (MM) $12.5 $18.0 $27.0 Advertising (MM) $12.0 $17.5 $25.0 30" HH GRPs 816 1044 1400 15" HH GRPs 669 854 1200 Online (MM) $0.5 $0.5 $2.0 Banner Ads TRPs 75 75 200 Emails 2MM 2MM 2MM Facebook (75k fans) Ad Ad Promo AWARENESS 19% 23% 35% Page 11
- 12. Key Requirements Ability to Integrate Cross‐Channel Purchasing ~ Ability to account for purchasing through alternative channels within one respondent – avoiding double counting volume and keeping costs down Page 12
- 13. Key Requirements Ability to integrate multiple layers of influence and decision making ~ Kids impact moms…insurance decisions made jointly…office managers, patients influence physicians and vice‐versa Inputs % Lives Reimbursed Formulary Decisions Copay Price Prior Authorization Reimbursed Physicians Physician's Share of Share of 20% mark-up Prescriptions Prescriptions Copay Scripts Scripts Retail price Patient Patient Complex DTC Marketing Awareness Decisions Decisions Interaction Page 13
- 14. Key Requirements Ability to identify offer acceptors at the individual level for targeting and offer optimization (penalty analysis). Sample Trier Profile Profile Index Age % % 18-24 11 11 106 25-34 20 15 74 35-49 41 47 114 50-64 28 27 96 Outside Franchise 17 11 66 Light Users 24 36 153 Heavy Users 39 45 115 Super‐Heavy Users 21 8 39 Too Hot Just Right Too Cold Temperature 23 50 27 Repeat Index 96 120 72 Page 14
- 15. Key Requirements Adaptable to Accommodate Complex Launches ~ Methodology should be flexible enough to address alternative launch scenarios: staged introduction? discounting? potential for added features? UNITS 78.9 73.8 62.2 65.8 17.6 28.8 Product Y 22.2 21.5 Product X 62.22 Current 47.25 43.62 39.81 Current Current & Current & Current, Product X Product Y Product X & Y Page 15
- 16. Key Requirements: Flexibility ~ BACKGROUND: Multiple generations of an offering were under consideration – each one delivering more than the previous one and the client had a desire to price each commensurately with the added benefit. ~ OBJECTIVE: Forecast each new offering and determine the opportunity for price escalation and coexistence. ~ OUTCOME: Able to identify price thresholds for each generation, when to phase out previous offerings and where opportunities for enhanced margins resided. Page 16
- 17. Big Opportunities to Fine Tune through Optimization Page 17
- 18. Alternative Strategy Assessment Variations of: Pricing Volume Branding Forecast Features Portfolio Choice Set 1 Respondent evaluates Choice Set 2 alternatives in competitive Choice Set 3 context Choice Set 4 Page 18
- 19. Product Optimization – Best Product Offerings Retail Factory Legitimizing Competitive Sales Retail Sales Sales # Bundle Claim Claim Price Form (MM) Index to Base (MM) 1 A M A Low X $194.2 139 $119.5 2 A M B Lower X $187.3 134 $115.2 3 A N A Lower X $183.1 131 $112.7 4 A O C Lower X $178.9 128 $110.1 5 A N C Lower X $177.5 127 $109.9 6 C M A Low X $174.7 125 $107.5 7 E M D Low X $171.9 123 $105.8 8 D N B Lower Y $167.7 120 $103.2 9 B M A Low Y $167.3 120 $102.9 10 E M B Lower X $166.3 119 $102.3 Page 19
- 20. Simulating Outcomes Page 20
- 21. Things to Avoid in Forecasting Page 21
- 22. Forecasting Pitfalls Ignoring a model, believing a model ~ Forecasting is part science, part art – experience counts! Page 22
- 23. Forecasting Pitfalls Relying solely on one measure Purchase Interest to predict in‐market outcomes Likeability/Benefit ~ Decision making is complex Value Failing to incorporate a measure Uniqueness of differentiation Competitive Context ~ The offering must provide a meaningful unfulfilled benefit ~ This benefit can’t be ignored Trial Page 23
- 24. Forecasting Pitfalls: Ignoring Differentiation ~ TESTED: New line of cookies that were co‐branded with current brands of candy bars. ~ OUTCOME: Utilizing a forecasting methodology that utilized differentiation only as a diagnostic measure, the lift in volume a truly differentiated offering could deliver was lost; hence, revenue projections were way‐underestimated. Page 24
- 25. Forecasting Pitfalls Sampling: Too broad/Too narrow ~ Too broad = Waste Target (F 21-36) Non-Target 80 70 60 50 ~ Too narrow = Missed Volume 40 30 71 53 20 47 29 10 0 % of Sample % of Volume Page 25
- 26. Forecasting Pitfalls Testing non‐executable offerings ~ Products that over‐promise, are over‐communicated or have an over stimulating concept drives volume that will never be achieved Ignoring cannibalization (source of business) $1016 Sourced From $691 Cannibalized 68% Current Product Line Current Product 1 35% Incremental $325 Current Product 2 15% Current Product 3 18% Revenue Page 26
- 27. Forecasting Pitfalls: Ignoring Cannibalization ~ PLAN: NEW PIZZA offering was going to be successful because it would just capture new occasions – parties, get‐togethers ~ OUTCOME: Traded current buyers down from “2 for 1” which generated higher margins and revenues Page 27
- 28. Forecasting Fiction Page 28
- 29. Forecasting Fiction X “Homeruns capture 10% market share” In most categories 3 to 5% is more realistic…fragmented categories more like .5 to 1% Parent Line extensions typically garner 10 to 30% Child SOM of parent – cannibalizes parent at 2 to 3 times “fair share.” Page 29
- 30. Forecasting Fiction X “A restage can drive 25% growth” ~ 10% is the most yr1 growth expected ~ Primary objective should be to hold SOM or re‐capture lost share Restaging SOM Assessor: An Overview “Real” Gain 2009 2010 2011 Page 30
- 31. Forecasting Fiction X “Not to worry – they will learn to like it!” ~ Most trial occurs within first 6 months – typically peaking at month 4 Cumulative 75% 100 Trial (%) 80 Assessor: An Overview60 30 40 Monthly 20 20 Trial (%) 10 10 2 4 6 8 10 12 Page 31
- 32. Forecasting Fiction X “This product will be everywhere!” ~ Maximizing distribution is critical to success ~ Impact has an almost linear impact on volume ~ Rate of distribution build is also important ~ Disappointed potential buyers ~ Less time for repeat Page 32
- 33. So, what is important in choosing a Forecasting Methodology? Competitive Context and Differentiation incorporated Marketing spend fairly represented Source of volume considered Flexible enough to accommodate complex launches Account for multiple layers of influence, cross‐channel buying Ability to profile identified triers (targeting, penalty analysis) Page 33
- 34. Scott Waller Vice President 1660 North Westridge Circle Irving, TX 75038-2424 M/A/R/C® Research Strong brands start with tel: 972-983-0412 fax:972-983-0444 strong research Scott.Waller@MARCresearch.com www.MARCresearch.com Amy Barrentine Executive Vice President, General Manager Randy Wahl 1660 North Westridge Circle Irving, TX 75038-2424 Executive Vice President M/A/R/C® Research Strong brands start with strong research tel: 972-983-0476 fax:972-983-0444 1660 North Westridge Circle Amy.Barrentine@MARCresearch.com Irving, TX 75038-2424 www.MARCresearch.com M/A/R/C® Research Strong brands start with tel: 972-983-0469 fax:972-983-0444 strong research Randy.Wahl@MARCresearch.com www.MARCresearch.com Page 34