conducting an effective return-on-marketing-investment analysis

26

Upload: isaiah-adams

Post on 31-May-2015

114 views

Category:

Marketing


0 download

DESCRIPTION

Jeff Ewald, Founder/CEO of Optimization Group, speaking on the topic of measuring marketing ROI. This presentation was given at the Michigan Marketing Minds meeting presented by SPARK Ann Arbor on 5/13/14

TRANSCRIPT

Page 1: Conducting an Effective Return-on-Marketing-Investment Analysis
Page 2: Conducting an Effective Return-on-Marketing-Investment Analysis

Conducting an Effective Return-on-Marketing Investment Analysis

Jeff EwaldOwner/Founder

Page 3: Conducting an Effective Return-on-Marketing-Investment Analysis

“I know that half of my advertising budget is wasted.

I just don’t know which half.”John Wannamaker

Page 4: Conducting an Effective Return-on-Marketing-Investment Analysis

Why is Marketing ROI so important?

• ROI provides immediate and long term benefits

• Focus on results – increased growth and profitability

• Compare marketing investments to other opportunities

• Compare results to expected results (variance analysis)

• Extraordinary focus on every dollar spent during downturn

• Multi-channel marketing makes interrelationships more critical (and difficult) to measure

Page 5: Conducting an Effective Return-on-Marketing-Investment Analysis

Why is Marketing ROI so difficult?Complex environment to

analyzeInternal Variables

Product/ Service

AttributesPrice

Positioning

Marketing Activities

DistributionSales force spend

AdvertisingDirect Marketing

Social MediaEvents

PR Promotion

External Variables

Competitive ActivityEconomic Variables

Political StatusSeasonality

General mediaWord of mouth

Consumer tastes/behaviorsConsumer cycles

Trigger points

Page 6: Conducting an Effective Return-on-Marketing-Investment Analysis

Situation ofRelative Certainty

Informs range of choices

Situation ofRelative Uncertainty

Informs “best choice”

Test market

Regression

STMsStructural equation models

NeuroNet Models

SimulationGenetic modeling/

Tool Comparison

Page 7: Conducting an Effective Return-on-Marketing-Investment Analysis

Putting together a ROMI analysis

1. Crystalize goals

2. Develop hypotheses

3. Determine required data

4. Determine appropriate analysis methodologies

5. Gather data – ETL

6. Conduct analyses to test hypotheses

7. Construct/evaluate/select predictive models

8. Implement … Simulators

Page 8: Conducting an Effective Return-on-Marketing-Investment Analysis

My Goal Today:Emphasize the critical importance of clear analysis goals and articulation of specific hypotheses

Page 9: Conducting an Effective Return-on-Marketing-Investment Analysis

Goals & Hypotheses• Goal: Determine the key factors, and their

relative weights, which are contributing to increases in sales

• Specific factors hypothesized as possible contributors include: • Weather patterns• Overall economic health• Agricultural and construction industry

employment levels• Changes in cost of critical production input

(i.e. cotton)• Advertising spending

Page 10: Conducting an Effective Return-on-Marketing-Investment Analysis

Determining data and analysis requirements

Page 11: Conducting an Effective Return-on-Marketing-Investment Analysis

Average Precipitation

S&P 500

Marketing Spend

Average Temperature

Consumer Sentiment

Employment

0 20 40 60 80 100 120

37

38

48

51

61

100

Relative Variable ImportancePredictors of $ Sales• Jobs and consumer

attitudes about the economy

• Weather • Overall marketing spend• Important to note that the

specific industry sector data did NOT emerge – generalized measures of the economy outweigh the ag and construction verticals

Page 12: Conducting an Effective Return-on-Marketing-Investment Analysis

• In each market, sales (demand) increases when average monthly temp drops below a specific value

• Influence of weather is a local market phenomenon

Albany Denver

MinneapolisSt. Louis

510

300

650

Impact of temperature varies dramatically by market

Page 13: Conducting an Effective Return-on-Marketing-Investment Analysis

.67

Variable importance is similar across DMAs

Live Cattle Futures

Agriculture Employment

Ave Monthly Temperature

Marketing Spend

Wheat Futures

S&P 500

Feeder Cattle Futures

0 20 40 60 80 100 120

34

35

37

50

53

98

100

Minneapolis

Ave Monthly Temperature

Consumer Sentiment Index

Marketing Spend

0 20 40 60 80 100 120

19

86

100

Albany

Marketing Spend

Ave Monthly Temperature

Employment

Labor Force

Consumer Sentiment Index

0 20 40 60 80 100 120

56

72

76

88

100

Denver

CPI

Precipitation

Consumer Sentiment Index

S&P 500

Ave Monthly Temperature

Marketing Spend

0 20 40 60 80 100 120

32

46

83

89

90

100

St. Louis

Page 14: Conducting an Effective Return-on-Marketing-Investment Analysis

Goals & Hypotheses• Goals: determine the most

effective marketing efforts and a strategy to improve Leads and Enrollment numbers

• Key Hypothesis: different tools impact different stages of “the funnel”

LEADS

STARTS

ENROLLS

7.5%

65.6%

Page 15: Conducting an Effective Return-on-Marketing-Investment Analysis

Leads Coeff Sig Leads per $1000 Cost per LeadTV 0.007 0.981 * 7 142.86$ Website 0.007 0.574 7 142.86$ Web Based Initiatives 0.02 0.988 * 20 50.00$ Outdoor 0.164 0.95 * 164 6.10$ Newspaper -0.027 0.961 * -27Yellow Pages 0.019 0.658 19 52.63$ Radio 0.039 0.953 * 39 25.64$ Direct Mail -0.052 0.935 * -52Event Marketing 0.009 0.159 9 111.11$

• Separate analyses for Leads, Enrollments and Starts

• Example: Leads

Preliminary Results

Page 16: Conducting an Effective Return-on-Marketing-Investment Analysis

“Dual Path” Model

16

TVWebsite WBIRadio

LEADS

Out door

Newspaper

Direct Mail

Enrolls

.171.053 .053.376

.369

.268

.983

.059

.908

.-.324

(Structural Equation Modeling)

Page 17: Conducting an Effective Return-on-Marketing-Investment Analysis

Goals & Hypotheses

• Goals: determine the optimal levels of key tactics to:

• Drive higher margins/mix-up

• Drive more product sales/volume

• Across a wide range of marketing tactics

Page 18: Conducting an Effective Return-on-Marketing-Investment Analysis

Marketing Drivers Ad spend & mix

Tire line promoted Rebate levels

Credit card rebates SPIFFS

External Drivers Gas prices

Miles driven Car park

New vehicle sales Consumer confidence Weather

Outcome Measures Brand $ Sales

Brand Unit sales Brand mix (% of Corp)

Outcome Measures Margin Contribution

Ave Line Margins

Analytical Plan Overview

Page 19: Conducting an Effective Return-on-Marketing-Investment Analysis

Goals & Hypotheses (Revised): • Do rebates work to drive volume and/or margin?

• Do some tire lines demonstrate better response to rebates than others?

• How does the rebate AMOUNT impact response rate and margin?

• Do rebate responses change with changes in the external environment?

Original scope too broad … revised project focused on understanding consumer rebates

Page 20: Conducting an Effective Return-on-Marketing-Investment Analysis

Stimulus Response• Independent Variables • Dependent Variables

Rebate at a specific $-level

On a specific tire line

Tires Per Day

Ave Line Margin

Page 21: Conducting an Effective Return-on-Marketing-Investment Analysis

Almost all rebates impact TPDRelative Impact Index

Non-rebate factors stronger than specific tire line rebate

$ REBATE LEVEL

Tires Per Day

Consumer Sentiment 100Consumer Confidence 98Ad Spend 97Car Park 92Gas Price 91Circulation 90Eagle F1 Asym A/S (new) 20 51Wrangler SR-A 20 49SP Sport Signature 20 47Other brand 20 46Eagle EMT/ROF 80 45Other brand 80 45Assurance ComforTred Touring 80 44Eagle F1 Asym A/S (new) 60 43SP Sport Family - no SP Sport Signature 20 43Graspic DS-3 40 42Wrangler SA 40 41Wrangler Duratrac 20 41Graspic DS-3 20 41SP Sport Family - no SP Sport Signature 60 41SP Sport Winter 3D 80 41Assurance CS Fuel Max 40 40Other Dunlop tire 60 40Assurance CS Fuel Max 80 39Rover HT 40 39Rover HT 80 38

$ REBATE LEVEL

Tires Per Day

SP Sport Winter 3D 40 38Assurance Fuel Max 20 37Wrangler SR-A 40 37Wrangler Duratrac 40 37Assurance CS Fuel Max 60 37Eagle F1 Asym A/S (new) 40 37SP Sport Signature 60 37Assurance Fuel Max 80 36Assurance TT AS 40 36Ultra Grip Winter 40 36Other brand 40 36Assurance ComforTred Touring 40 35Wrangler SA 80 35SP Sport Signature 80 35Wrangler SR-A 80 34Assurance TT AS 80 33Wrangler Duratrac 80 33Wrangler SR-A 60 32SP Sport Family - no SP Sport Signature 40 32Ultra Grip Ice WRT 60 31Graspic DS-3 80 31Assurance Fuel Max 40 30Wrangler SA 20 30Other Goodyear tire 20 30Graspic DS-3 60 30SP Sport Family - no SP Sport Signature 80 30Wrangler Duratrac 60 28

$ REBATE LEVEL

Tires Per Day

Other Dunlop tire 20 28Other brand 60 28Assurance Fuel Max 60 27Signature CS 40 27Other Goodyear tire 60 26Assurance CS Fuel Max 20 23Rover HT 20 23Other Dunlop tire 40 23SP Sport Winter 3D 20 20Other Goodyear tire 80 19SP Sport Winter 3D 60 19Rover HT 60 17SP Sport Signature 40 16Other Dunlop tire 80 16Assurance TT AS 60 13Assurance TT AS 20 12Signature CS 80 8Assurance ComforTred Touring 60 6Eagle F1 Asym A/S (new) 80 5Assurance ComforTred Touring 20 3Wrangler SA 60 2Eagle GT 0Assurance CS TTAS 0Ultra Grip Ice WRT 40 0All Kelly tires 40 0

Page 22: Conducting an Effective Return-on-Marketing-Investment Analysis

Fewer rebate offers impact average tire margins …Ranked by Relative Importance Index relative to margin contribution(Random Forest)

Tire Line REBATE LEVELMargin Per Tire

Assurance TT AS 80 97Assurance ComforTred Touring 80 91SP Sport Family - no SP Sport Signature 40 58Eagle F1 Asym A/S (new) 80 43SP Sport Signature 40 42Other Goodyear tire 60 36Assurance Fuel Max 40 34Eagle F1 Asym A/S (new) 60 31Signature CS 40 30Wrangler SR-A 60 29SP Sport Signature 80 26Other Goodyear tire 20 21Wrangler SR-A 40 18Other Dunlop tire 80 18Assurance Fuel Max 80 16Assurance Fuel Max 20 14Wrangler SA 40 14Wrangler SR-A 80 13Wrangler Duratrac 60 12Assurance CS Fuel Max 40 11Other brand 80 11SP Sport Winter 3D 60 10

Page 23: Conducting an Effective Return-on-Marketing-Investment Analysis
Page 24: Conducting an Effective Return-on-Marketing-Investment Analysis

ROMI Recommendations

• Clear and concise ROMI objectives, coupled with specific hypotheses frames the required data and analysis methods

• Keep each analysis as simple as possible

• More actionable

• Easily repeatable – iterative learning and updating

• Implement the learning – feed it into a process change

Page 25: Conducting an Effective Return-on-Marketing-Investment Analysis

Jeff Ewald

[email protected]/optimizationgrp

facebook.com/optimizationgroup

Page 26: Conducting an Effective Return-on-Marketing-Investment Analysis