Marketing Decisions

Marketing Decisions

Extracting superior value from marketing investments is a necessary goal made difficult by scale, speed-to-market and decentralization of marketing decisions to line marketing. Historical analysis to justify marketing ROI is no indicator of future marketing success unless backed by an enterprise wide framework of predictive decision making. Planning, monitoring and evaluating effectiveness of marketing decisions on a predictive framework helps organizations extract superior ROI from each marketing dollar.

Risk Scorecard

Download a Sample Position paper on one of solutions for a flavor of our recommendations. Request a position paper on a particular solution by writing to solutions [at] kiesquare.com

marketing mix models
link advertising and promotion with business goals

picture

"Each element of the marketing activity provides different yield due to varying levels of reach, effectiveness, exposure weight and interactions with other activities. Most statistical models often overestimate the impact of the mass media elements like TV advertising or conversely fail altogether in their sensitivity to determine the impact of sharp spurts in media activities or point activities. Advanced residual modeling techniques are required to separate the impact of point variables like promotions, unconventional media, seasonality etc. from base variables like distribution depth and width, competitior activities and pricing. The entire process of impact estimation is data vintage and data variety driven and hence only data committed firms are advised to pursue this exercise for its full benefit.
Alternatives to POS data originating from scanner source has been retail audit information or even shipment data depending upon the geography while in case of marketing activities, variables vary from detailed activity level allocations to marketing spend based allocations. A full rank fixed effects model requires as many as 1000 to 3000 variables to be created while it eventually supports up to 200 variables across activities depending upon available data points and number of panels in a pooled regression.
For best optimization of Marketing Mix Models and real time integration into day to day marketing decisions, request advanced intervention team's input by writing to solutions [at] kiesquare.com

advertising effectiveness
develop superior advertising campaigns

picture

"Advertising has its short term as well as long term impact on sales, but most of the effectiveness determination methods are able to estimate only one of these impacts and often overestimate the impact for short term. There are advanced residual techniques that help marketers make realistic estimate of the advertising impact after removing the impact of and interactions with other activities. While most estimations focus on the weights of TV advertising, it is required to segregate the impact of message effectiveness and copy quality from the copy weight at each campaign level. An optimized advertising plan would not only indicate the total annual and weekly GRPs and its likely sales impact, but also layout the GRP exposure plan given saturation, decay and adstock parameters, TARP:GRP ratio, seasonality factors, day part splits and suggest complementary activities to maximize the media interaction impact. Basis these, the marketers can decide on number of interspersed ad copies, versions and messages on air at a time and over time to achieve the category goals.

Request "KiE Square Advertising Optimizer Demo" by writing to solutions [at] kiesquare.com

promotions effectiveness
build superior consumer promotion plans

picture

"Retailers and Categories have harmed themselves more by reading the promotion effectiveness right, but optimizing it wrong. Most methods of promotion effectiveness determination account for the lift that individual promotional event creates, but do not adjust it for the the pre and post promotion cannibalization, which often leaves the promotions net negative. The effectiveness determination is also confounded by simultaneous occurence of other marketing activities, proportion of stores in which the promotion was driven, proportion of stocks/SKUs on which promotion was run and coinciding factors like seasonality or competitor promotions. Often the base price determination is itself a challenge as it continually dips for highly promoted categories and a real measure for lift is difficult to arrive at. The knowledge of effectiveness of promotion is a part of the finding as the more elaborate exercise is to optimize it vis-a-vis implementation challenges like
Which promotion type for which store given fixed and variable costs of the event and the offer, how to bundle it with other activities, which SKUs to promote, what type of communication to support it with, whether catalogue promotions with display bring adequate incremental customers vis-a-vis a non- catalogue campaign,
the depth of discount, the duration of promotion, the frequency, what proportion of stores, what proportion of shelf space, which aisle, On shelf/ Off shelf, FGE/BGE or wings?
Request KiE Square's Promotion Optimization Expert Panel for ready solutions by writing to solutions [at] kiesquare.com

cross-sell / up-sell
improve revenue per customer

picture

"All marketing driven businesses face a classic tradeoff between harnessing value from New Users vis-a-vis that from New Uses or Usages. The organizational focus shifts in phases, but creating an incremental customer base is always a more cost and time intensive exercise than deriving incremental value from existing customers by fulfilling their new related or unrelated needs (X-sell) or increasing current consumption levels (Up-sell). Regular response models deployed to ascertain the response rates and next likely products at customer segment level is not enough to obtain the Maximum Value Threshold (MVT) of the portfolio. Most response models operate at meagre 20 to 30% of MVT, when there are analytical methods to enhance performance.
Request KiE Square's Advanced interventions in Cross Sell/ Up sell Optimization by writing to solutions [at] kiesquare.com

loyalty & rewards
reward loyal customers

picture

"A suitable CRM based BI application is required to analyze the customer's recent or past history and to deploy real time behavioral models. Such models not only recommend appropriate reward mechanism or tier movement for the customers but also track loyalty indices over time and integrate with Customer LIfetime Value (CLTV) measure or Cross sell / Upsell framework. However, the best Loyalty models amalgamate these findings with the primary research surveys and Household panels to not only corroborate model outputs but also to enable root cause detection for Loyalty swings.
Request KiE Square's Sample Loyalty and Reward Program on the basis of real time analytics by writing to solutions [at] kiesquare.com

consumer behavior models
obvious and not-so-obivious behavior of consumers...

picture

"The intended behavior of the customers is mainly influenced by communication, perception and experience through interplay of two key components of attitude - Affect and Cognition, that indicate a ‘feel’ based evaluation and a ‘rational’ evaluation of a choice object respectively. When the familiarity with a choice object is low, its evaluation tends to be rational and as the familiarity increases the evaluation basis becomes more global in nature. A marketer needs to analyze the attitudinal composition of its customers from time to time and align the marketing cues through in-situ and mass communication. Affective cues are often stronger than the cognitive cues and are absorbed more easily. There are loyalty implications through early transition of customers from the cognitive mode to the affective mode, as the latter state of attitude is more stable and difficult to be influenced by the competition cues.

Request details of KIE Square's proprietary models on Harnessing combined effects of Cognition and Affect for Long term Market Governance by writing to solutions [at] kiesquare.com