Dale Hagemeyer of Gartner on Trade Promotion Modeling and Optimization
By Dale Hagemeyer, Research VP, Manufacturing - Life Sciences and Consumer Goods, Gartner Industry Advisory Services, Gartner Inc.
Responses from the consumer goods industry with regard to predictive capabilities in trade promotions range from "totally skeptical" to "it will cure world hunger." Here is an industry analyst's take on why being more predictive in trade promotion management (TPM) has merit.
1. It can reduce complexity. This doesn't mean it always gives the answer, but it can make forming the solution easier. Think of it like a prediction of whether or not it will rain. A forecast doesn't typically state whether or not you should take an umbrella along for the day, instead it gives a percentage chance of rain occurring and lets you determine what to do. So rather than looking out the window and trying to sort through multiple indicators of inclement weather from clouds to temperature to recent weather patterns, a predictive forecast cuts down the information and allows the consumer of the information to make an informed choice based on available information.
2. Simulation is a great way to redirect spending and increase ROI on promotional spend. In a typical negotiation between retailer and manufacturer, one of the first questions, whether posed or merely implied, by the retailer is "how much more are you going to spend during the upcoming period?" Being able to do scenarios-based predictive modeling can turn the conversation to more of "what can be generated or achieved" instead of "how much will be spent." It can also frame up the scenarios as to how both manufacturer and retailer will benefit. Naturally, the predictive capabilities have to be credible, but early results by some "first movers" in the industry are very promising.
3. It can improve operational efficiencies. So it isn't just about ascending the pinnacle of fact-based selling but actually modeling and improving outcomes. Knowing about product mix and volume beforehand can help to reconcile two opposing forces: Sales optimism and supply chain skittishness. Thus, predictive modeling can bring more harmony to the demand creation and demand fulfillment processes.
4. It can ease the analytical burden of the sales force. For starters, predictive modeling will require that the elements of post-event analysis are automated because this is how the past is used to model the future. So rather than making post event analysis a manual, drawn out process with all thoughts for the future being a matter of "guessing and hoping," the technology can create a base scenario. This can then be modeled further by tweaking the assumptions and variables with considerably less manual input and number crunching.
5. It can make sales people more effective. High-performing account managers typically have good instincts about cause and effect relationships in the market. However, weak performers don't. Both can benefit from predictive modeling because the high performers can better quantify the scenarios that make sense to them intuitively, and the less intuitive can benefit from having some insights that they couldn't have otherwise.
6. This may be the way to take sales agents to a new level. Since broker organizations often handle multiple products form multiple principals, they could benefit from good modeling tools because brokers are usually leveraged for their market expertise and relationships, not because they are more insightful than a direct sales force. Predictive tools could help them with decision making and could even be used to help them simulate their commissions instead of spending time tracking orders.
7. It will keep manufacturers relevant with retailers. As a whole, retailers jumped out early in predictive modeling and are ahead of manufacturers. Showing competence means gaining credibility with retailers. Retailers generally don't invest in IT to the degree that manufacturers do.With the right investments, manufacturers can come up to speed quickly. Using largely the same data as retailers (like point-of-sale data) increases both their credibility and indispensability with retailers.
The journey from randomness and chaos to basic analytical competence to being predictive to actually influencing outcomes can be summarized in Figure 1. See where you are in the journey and how you might be able to increase your competitive advantage by becoming more predictive in your promotions. But, most importantly, remember that it is a journey.
Responses from the consumer goods industry with regard to predictive capabilities in trade promotions range from "totally skeptical" to "it will cure world hunger." Here is an industry analyst's take on why being more predictive in trade promotion management (TPM) has merit.
1. It can reduce complexity. This doesn't mean it always gives the answer, but it can make forming the solution easier. Think of it like a prediction of whether or not it will rain. A forecast doesn't typically state whether or not you should take an umbrella along for the day, instead it gives a percentage chance of rain occurring and lets you determine what to do. So rather than looking out the window and trying to sort through multiple indicators of inclement weather from clouds to temperature to recent weather patterns, a predictive forecast cuts down the information and allows the consumer of the information to make an informed choice based on available information.
2. Simulation is a great way to redirect spending and increase ROI on promotional spend. In a typical negotiation between retailer and manufacturer, one of the first questions, whether posed or merely implied, by the retailer is "how much more are you going to spend during the upcoming period?" Being able to do scenarios-based predictive modeling can turn the conversation to more of "what can be generated or achieved" instead of "how much will be spent." It can also frame up the scenarios as to how both manufacturer and retailer will benefit. Naturally, the predictive capabilities have to be credible, but early results by some "first movers" in the industry are very promising.
3. It can improve operational efficiencies. So it isn't just about ascending the pinnacle of fact-based selling but actually modeling and improving outcomes. Knowing about product mix and volume beforehand can help to reconcile two opposing forces: Sales optimism and supply chain skittishness. Thus, predictive modeling can bring more harmony to the demand creation and demand fulfillment processes.
4. It can ease the analytical burden of the sales force. For starters, predictive modeling will require that the elements of post-event analysis are automated because this is how the past is used to model the future. So rather than making post event analysis a manual, drawn out process with all thoughts for the future being a matter of "guessing and hoping," the technology can create a base scenario. This can then be modeled further by tweaking the assumptions and variables with considerably less manual input and number crunching.
5. It can make sales people more effective. High-performing account managers typically have good instincts about cause and effect relationships in the market. However, weak performers don't. Both can benefit from predictive modeling because the high performers can better quantify the scenarios that make sense to them intuitively, and the less intuitive can benefit from having some insights that they couldn't have otherwise.
6. This may be the way to take sales agents to a new level. Since broker organizations often handle multiple products form multiple principals, they could benefit from good modeling tools because brokers are usually leveraged for their market expertise and relationships, not because they are more insightful than a direct sales force. Predictive tools could help them with decision making and could even be used to help them simulate their commissions instead of spending time tracking orders.
7. It will keep manufacturers relevant with retailers. As a whole, retailers jumped out early in predictive modeling and are ahead of manufacturers. Showing competence means gaining credibility with retailers. Retailers generally don't invest in IT to the degree that manufacturers do.With the right investments, manufacturers can come up to speed quickly. Using largely the same data as retailers (like point-of-sale data) increases both their credibility and indispensability with retailers.
The journey from randomness and chaos to basic analytical competence to being predictive to actually influencing outcomes can be summarized in Figure 1. See where you are in the journey and how you might be able to increase your competitive advantage by becoming more predictive in your promotions. But, most importantly, remember that it is a journey.