Marketing mix modeling ( MMM ) is a statistical analysis such as a multivariate regression on sales and time chain marketing data to estimate the impact of marketing tactics (sales mix) on sales and then foresee the impact of a series of tactics of the future. This is often used to optimize the advertising mix and promotional tactics with respect to sales revenue or profits.
These techniques were developed by econometric experts and were first applied to consumer packaged goods, because the manufacturers of these goods had access to good data on sales and marketing support. The first company dedicated to the commercial development of MMM was the MMA (then Media Marketing Assessment) started in 1990 and the Hudson River Group was established in 1989. The other early econometric modeling pioneers were the ATG group in the JWT advertising agency in the 1990s. and then incorporated into MindShare ATG, BrandScience at Omnicom, and the OHAL specialist modeling agency since the late 1980s. These institutions take MMM from a slightly used academic discipline and become a common and widespread marketing tool. Increased data availability, enormous computing power, and the pressure to measure and optimize marketing spending have driven the popularity boom as a marketing tool. Recently MMM has found acceptance as a trustworthy marketing tool among major consumer marketing companies. Often in the context of digital media, MMM is referred to as attribution modeling.
Video Marketing mix modeling
Histori
The term marketing mix was developed by Neil Borden who first started using the phrase in 1949. "An executive is a material mixer, sometimes following a recipe while walking, sometimes adjusting recipes to available materials immediately, and sometimes experimenting with or creating ingredients that have not never tried anyone else. "(Culliton, J. 1948)
According to Borden, "When building a marketing program to fit his company's needs, the marketing manager must consider the power of behavior and then juggle the marketing elements in his mix with a keen eye on the resources he has to work on." (Borden, N. 1964 pg 365).
E. Jerome McCarthy (McCarthy, J. 1960), was the first to suggest four marketing P - prices, promotions, products and places (distribution) - which are the most commonly used variables in building a marketing mix. According to McCarthy, marketers basically have four variables they can use when devising a marketing strategy and writing a marketing plan. In the long run, the four mixed variables can be changed, but in the short term it is difficult to modify the product or distribution channel.
Another set of marketing mix variables was developed by Albert Frey (Frey, A. 1961) who classifies marketing variables into two categories: supply, and process variables. "Bid" consists of product, service, packaging, brand, and price. The "process" or "method" variables include advertising, promotion, sales promotion, personal selling, publicity, distribution channels, marketing research, strategy formation, and new product development.
Recently, Bernard Boom and Mary Bitner built a model of seven P's (Boom, B. and Bitner, M. 1981). They add "People" to the list of available variables, to recognize the importance of the human element in all aspects of marketing. They add a "process" to reflect the fact that services, unlike physical products, are experienced as processes when they are purchased. Desktop modeling tools such as Micro TSP have made statistical analysis of this kind as part of the mainstream now. Most advertising agencies and strategy consulting firms offer MMM services to their clients.
Maps Marketing mix modeling
Marketing mix model
Marketing mix modeling is an analytical approach that uses historical information, such as syndicated sales point data and internal company data, to measure the impact of sales on various marketing activities. Mathematically, this is done by building a simultaneous relationship of various marketing activities with sales, in the form of linear or non-linear equations, through statistical regression techniques. MMM defines the effectiveness of each marketing element in terms of its contribution to sales volume, effectiveness (volume generated by each business unit), efficiency (resulting sales volume divided by cost) and ROI. This lesson is then adopted to adapt marketing tactics and strategies, optimize marketing plans and also to forecast sales while simulating various scenarios.
This is achieved by modeling the volume/value of sales as the dependent variable and the independent variable created from various marketing efforts. The creation of variables for Marketing Mix Modeling is a complicated affair and is the same art with science. The balance between an automated modeling tool rattling large data sets versus econometric builders is an ongoing debate at MMM, with various agencies and consultants taking positions at certain points in the spectrum. Once a variable is created, several iterations are performed to create a model that explains the trend of volume/value well. Further validation is performed, either by using validation data, or with the consistency of business results.
The output can be used to analyze the impact of marketing elements on various dimensions. The contribution of each element as a percentage of the total plotted year to year is a good indicator of how the effectiveness of various elements changed over the years. Annual changes in contributions are also measured by analysis because they indicate what percentage changes in total sales are caused by each element. For activities such as television commercials and trade promotions, more sophisticated analyzes such as effectiveness can be made. This analysis tells the marketing manager the profit increase in sales that can be gained by improving each marketing element by one unit. If detailed spending information per activity is available, then it is possible to calculate Return on Investment from marketing activities. This is not only useful for reporting the historical effectiveness of the activities, but also helping to optimize the marketing budget by identifying the most efficient and least efficient marketing activities.
Once the final model is ready, the results from it can be used to simulate a marketing scenario for the 'What-if' analysis. The marketing manager can reallocate this marketing budget in different proportions and see a direct impact on sales/value. They can optimize the budget by allocating expenditures for activities that provide the highest return on investment.
Some MMM approaches want to include some belligerent products or brands in an industry or category model - where cross-price relationships and share of voice advertising are important for wargaming.
Components
The mixed-marketing model outlines total sales into two components:
Basic Sales : This is a natural demand for products driven by economic factors such as price, long-term trends, seasons, and also qualitative factors such as brand awareness and brand loyalty.
Additional Sales : Incremental sales are sales components that are driven by marketing and promotional activities. This component may be further broken down into sales as any marketing component such as television commercials or Radio advertising, Print Advertising (magazines, newspapers etc.), Coupon, Direct Mail, Internet, Features, or Temporary Price Display and Deduction Promotions. Some of these activities have short-term returns (Coupons, Promotions), while others have long-term returns (TV, Radio, Magazine/Print).
Marketing-Mixed Analysis is usually done using Linear Regression Modeling. Nonlinear and lagged effects are included using techniques such as Adstock Ad transformation. Typical outputs of the analysis include decomposition of total annual sales into contributions from each of the marketing components, a.k.a. Contribution pie chart.
Another standard output is the decomposition of the year-on-year sales growth/decline, a.k.a. 'Graph-fall'. The
element is measured in MMM
Additional bases and volumes
Sales volume breakdown becomes the basis (volume that will be generated if there is no marketing activity) and the addition (volume generated by marketing activities in the short run) at all times provides tremendous insight. Bases grow or decrease over a longer period of time while activity generates additional volume in the short run also impacts the base volume in the long run. The basic volume variation is a good indicator of brand strength and loyalty ordered from its users.
Media and ads
The modeling of the market mix can determine the impact of sales generated by individual media such as television, magazines, and online display ads. In some cases, it can be used to determine the impact of individual advertising campaigns or even the execution of advertising on sales. For example, for TV advertising activity, it is possible to check how each ad execution has been done in the market in terms of its impact on sales volume. MMM may also provide information on the correlation of TVs on different media weights, as measured by the Gross Rating Points (GRP) in relation to the sales volume response within a period of time, be it a week or a month. Information can also be obtained at a minimum level of GRP (threshold) in a week that needs to be aired to make an impact, and vice versa, the GRP level where the impact on the maximum volume (saturation limit) and that subsequent activity has no reward. Although not all MMMs will be able to generate definite answers for all questions, some additional areas where insight can sometimes be obtained include: 1) the effectiveness of 15 seconds vis-ÃÆ'-vis execution of 30 seconds; 2) comparison in ad performance when executed during prime-time vis-ÃÆ'-vis off-prime-time dayparts; 3) comparison with the immediate and halo effects of TV activity on various products or sub-brands. The role of new product-based TV activities and equity-based TV activities in growing brands can also be compared. GRP converted to reach (ie GRP divided by average frequency to get the percentage of people actually watching ads). This is a better size for TV modeling.
Trade promotion
Trade promotion is a key activity in any marketing plan. It aims to increase sales in the short term by using promotional schemes that effectively increase customer awareness of the business and its products. The consumer's response to trade promotion is not straight forward and is a matter of debate. Non-linear models exist to simulate the response. By using MMM, we can understand the impact of trade promotions in generating additional volumes. It is possible to get the estimated volume generated per promotional event at each of the different retail outlets by region. In this way we can identify the most effective and least effective trade channels. If detailed expense information is available, we can compare Return on Investment from various trading activities such as Every Day Low Price, Off-Shelf Display. We may use this information to optimize our trading plans by choosing the most effective trading channels and targeting the most effective promotional activities.
Pricing
The price increase of the brand has a negative impact on sales. This effect can be captured through price modeling in MMM. This model provides price elasticity of the brand that tells us the percentage change in sales for each price percentage change. By using this, the marketing manager can evaluate the impact of the price change decision.
Distribution
For the distribution element, we can know how the volume will move by changing the distribution effort or, in other words, by each percentage of the shift in the width or depth of the distribution. These can be specifically identified for each channel and even for each type of outlet for off-take sales. Given this insight, distribution efforts can be prioritized for each channel or store type to get the most out of the same. A recent study on laundry brands showed that volume increase through 1% more attendance at Kirana neighborhood stores was 180% greater than that through the presence of 1% more in supermarkets. Based on the cost of these efforts, managers identify the right channel to invest more for distribution.
Launch
When a new product is launched, the associated publicity and promotion usually results in a higher-than-expected generation of volumes. This extra volume can not be completely retrieved in the model using an existing variable. Often special variables to capture the incremental effect of this launch are used. The combined contribution of these variables and the marketing efforts associated with the launch will contribute to the total rollout. Different launches can be compared by calculating their effectiveness and ROI.
Competition
The impact of competition on brand sales is captured by creating appropriate competition variables. Variables are made from competition marketing activities such as television commercials, trade promotions, product launches, etc. The results of the model can be used to identify the greatest threat to have a brand sale out of competition. Cross-price elasticity and cross-promotional elasticity can be used to design an appropriate response to competitive tactics. Successful competitive campaigns can be analyzed to learn valuable lessons for their own brands.
Study at MMM
A typical MMM study provides the following insights
- Contributed by marketing activity
- ROI with marketing activity
- The effectiveness of marketing activities
- Optimal distribution of expenditure
- Lessons on how to run each activity are better eg. Optimum GRP per week, optimal distribution between 15 and 30, which promo to run, what SKUS for promotion etc.
Industrial MMM adoption
Over the past 20 years many large companies, particularly consumer packaged goods companies, have adopted MMM. Many Fortune 500 companies like P & amp; G, AT & amp; T, Kraft, Coca-Cola and Pepsi have made MMM an integral part of their marketing planning. This has also been possible due to the availability of specialist companies that now provide MMM services.
The marketing mix model is more popular initially in the CPG industry and is rapidly spreading to Retail and Pharmaceutical industries due to availability of Syndicated Data in this industry (mainly from Nielsen Company and IRI and to lower levels of NPD Group and Bottom Line Analytics and Gain Theory). The availability of Time-series data is critical to the powerful modeling of mixed-marketing effects and with the systematic management of customer data through CRM systems in other industries such as Telecommunications, Financial Services, Automotive and Hospitality industries helping its dissemination to these industries. In addition, the availability of competitive and industry data through third-party sources such as the Forrester Research Consumer Panel, the Insight of Polk (Automotive) and Smith's Travel Research (Hospitality), further enhances the application of the marketing mix model for these industries. The application of mixed-marketing modeling for these industries is still in the nascent stage and many standardization needs to be done mainly in these areas:
- Interpretation of cross-industry promotional activities for eg promotion in CPG has no lagging effect as it takes place inside stores, but automotive and hospitality promotions are typically implemented over the internet or through dealer marketing and can have a longer impact. CPG promotion is usually an absolute discount price, whereas Automotive promotions can be either cash back or loan incentives, and the promotion of Financial Services is usually a discounted interest rate.
- The hotel marketing industry has a very heavy seasonal pattern and most mixed marketing models will tend to disrupt marketing effectiveness with the seasons, thus exaggerating or underestimating marketing ROI. Series-Sectional Time-series models such as Pooled Regression need to be utilized, which increase sample size and variation and thus create a strong separation of the pure marketing effects of the season.
- Automotive manufacturers spend large amounts of their marketing budget on agency advertising, which may not be accurately measured if not modeled at the right aggregate level. If modeled at national or even market level or DMA level, this effect may be lost in aggregation bias. On the other hand, going to the dealer level might overestimate marketing effectiveness as it will ignore the consumer movement between dealers in the same area. The right and appropriate approach is to determine what traders engage in 'accessible' general groups based on overlapping 'trading areas' determined by consumer postal codes and cross-shopping information. At least the 'General Dealer Area' can be determined by grouping dealers based on the geographic distance between the dealer and the county sales department. A mixed-marketing model built by 'collecting' monthly sales for these dealer clusters will be used effectively to measure the impact of dealer advertising effectively.
The development of mixed-marketing modeling is also accelerated because of the focus of Sarbanes-Oxley Section 404 which requires internal controls for financial reporting of significant expenditures and expenditures. Marketing for consumer goods can be more than 10 total incomes and until the emergence of a mixed-marketing model, depending on the qualitative or 'soft' approach to evaluating this spending. Marketing-mixed modeling is presented a rigorous and consistent approach to evaluating mixed marketing investments as the CPG industry has already been demonstrated. A study by the American Marketing Association shows that top management is more likely to emphasize the importance of marketing accountability than middle management, showing a top-down push toward greater accountability.
Limitations
While the marketing mix model provides a lot of useful information, there are two key areas where these models have limitations that should be taken into account by everyone who uses this model for decision making purposes. These limitations, discussed more fully below, include:
1) focus on short-term sales can significantly underestimate the importance of long-term equity development activities; and
2) when used for mix media optimization, this model has a clear bias that supports time-specific media (such as TV commercials) compared to less time-specific media (such as ads appearing in monthly magazines); bias can also occur when comparing broad-based media rather than to regional or demographic targeted media.
With respect to a bias toward equity development activities, optimized marketing budgets using the marketing mix model may tend to be too much on efficiency because mixed marketing models measure only the short-term effects of marketing. Long-term marketing effects are reflected in its brand equity. The impact of marketing spending on [brand equity] is usually not captured by mixed-marketing models. One reason is that the longer duration that marketing takes to influence brand perception extends beyond the simultaneous or, at best, the weeks-of-marketing impact on sales as measured by these models. Another reason is temporary sales fluctuations because economic and social conditions do not necessarily mean that marketing is not effective in building brand equity. On the contrary, it is quite possible that in short-term sales and market share may get worse, but brand equity may actually be higher. This higher equity should in the long run help the brand recover sales and market share.
Since a mixed-marketing model shows marketing tactics having a positive impact on sales does not necessarily mean it has a positive impact on long-term brand equity. Different marketing measures have an effect on short-term and long-term brand sales differently and adjusting the marketing portfolio to maximize either short-term or longer term will be less than optimal. For example, the short-term positive effects of promotion on consumer utilities encourage consumers to switch to promoted brands, but the adverse effects of promotion on brand equity occur from period to period. Therefore, the net effect of promotion on brand market share and profitability can be negative because of its negative impact on brands. Determining a marketing ROI based on a marketing mix model alone can lead to misleading results. This is because of the marketing mix effort to optimize the marketing mix to increase additional contributions, but the marketing mix also drives brand equity, which is not part of the additional sections as measured by the marketing mix model - this is part of the baseline. True 'Return on Marketing Investment' is the amount of short-term and long-term ROI. The fact that most companies use a mixed-marketing model just to measure short-run ROI can be inferred from an article by Booz Allen Hamilton, which suggests that there is a significant shift from traditional media to 'below-line' spending, driven by the fact that more promotional spending easily measured. But academic studies have shown that promotional activities are actually detrimental to long-term marketing ROI (Ataman et al., 2006). Short-run marketing mix models can be combined with brand equity models using brand tracking data to measure 'brand ROI', both in the short and long run. Finally, the modeling process itself should be no more expensive than the profit generated in profitability; it should have a positive Modeling Effort (ROME).
The second limitation of the marketing mix model comes into effect when advertisers try to use this model to determine the best media allocation across different types of media. The traditional use of MMM to compare money spent on TV versus money spent on coupons is relatively applicable in terms of TV commercials and coupon appearances (for example, in FSIs run in newspapers) both are quite time-specific. However, since the use of these models has been extended to comparisons across a broader range of media types, extreme caution must be used.
Even with traditional media such as magazine ads, the use of MMM to compare results in the media can be a problem; while the modelers coat the typical monthly magazine 'display' curve model, this lack of precision, and thus introduce additional variability into the equation. Thus, the comparison of the effectiveness of running TV ads versus the effectiveness of running magazine ads will bias support TV, with greater precision of measurement. As new forms of media multiply, this limitation becomes more important to consider if MMM is used in an effort to measure its effectiveness. For example, Sponsor Marketing, Affinity Marketing Sports, Viral Marketing, Blog Marketing, and Mobile Marketing vary in terms of specificity of exposure time.
Furthermore, most approaches to the mixed-marketing model try to include all aggregate marketing activities at the national or regional level, but as far as tactics are targeted for different consumer groups of demographics, the impact may be lost. For example, Mountain Dew sponsorship from NASCAR can be targeted to NASCAR fans, which may include some age groups, but Mountain Dew ads in game blogs may be targeted to Gen Y population. These tactics may be very effective in the appropriate demographic groups but, when included in aggregate in the national or regional marketing mix model, may appear as ineffective.
Aggregation aggregation, along with issues related to variations in the nature of the particular time of different media, poses a serious problem when these models are used in ways beyond their original design. As the media becomes more fragmented, it is imperative that these issues are taken into account if marketing mix models are used to assess the relative effectiveness of different media and tactics.
The marketing mix model uses historical performance to evaluate marketing performance and is therefore not an effective tool for managing marketing investments for new products. This is because the relatively short history of new products makes the marketing-mixed results unstable. Also the relationship between marketing and sales may differ greatly in launch and stable periods. For example, Coke Zero's initial performance is really bad and shows low ad elasticity. Although Coke is increasing its media expenditure, with an improved strategy and radically improving its performance resulting in ad effectiveness that may be several times effective during the launch period. Typical marketing mix models would recommend cutting media spending and instead using expensive price discounts.
See also
- Marketing
- Marketing Strategy
- Marketing Management
- Marketing Plan
- Strategic Management
- Strategic Planning
- Marketing Effectiveness
- Return on Marketing Investment
- Marketing performance measurement and management
- Marketing ROI
- shopper marketing
- The request chain
References
External links
- Beraman Ataman, Harald J. van Heerde, Carl F. Mela, (2006) "The Long-Term Effect of Marketing Strategies on Brand Performance", Journal of Marketing Research.
- Booz Allen Hamilton's article
- How to Know What Media Works?
Source of the article : Wikipedia
0 Comments