CPG executives are painfully aware that they’re investing billions of dollars in trade promotions each year, but as many as 72% fail to break even¹. It’s clear that promotions have become more complex and harder to manage as CPGs must respond to changing consumer behavior, increasing demands from retailers and blurring of physical and online channels. Traditional forecasting and promotion-planning systems are unable to provide real-time, accurate insights to help managers understand the big picture. Below, we’ll explore seven ways in which AI can help CPG companies more effectively plan promotional events, measure outcomes and make adjustments. You can read more in the companion paper on how AI transforms promotional trade funds management.
1. Driving global objectives through locally tailored strategies
As CPG account teams are busy managing trade promotions in several categories for hundreds of retailers and multiple channels, it’s impossible to align global objectives and local strategies using legacy processes. Now, with AI, local tactics for a retailer, category, region and brand can be aligned with global strategies as models can be autonomously orchestrated by algorithms. The algorithms define distinct promotion postures based on goals and priorities, and assign the correct posture for each segment, adjusting over time as needed. By leveraging AI-enabled insights from current POS data, promotions can be used as guided weapons to surgically strike where and when needed.
2. Discerning patterns across promotions
If a promotion tactic has been used for a product group in a similar market before, machine learning models can predict performance with precision superior to traditional statistical techniques. For promotions without a history of usage, deep learning capabilities in AI can be leveraged to find similarities in hidden patterns that go beyond what’s considered “look-alike” promotions. Additionally, by using patterns and influences determined by deep learning models, product and promotion combinations emerge that have never actually been run before, empowering account managers to attempt bold tactics for new brands.
3. Promotion management optimization
By using last year’s promotional data as a starting point, AI-driven recommendations and what-if analysis can allow users to view forecasts and tweak promotion mechanics in real time. AI can also help CPGs become more strategic about the investments they make in feature placement and displays by automating the optimization of promotions and enabling systematic test and learn cycles to maximize lift and store traffic.
4. Win-win deals and mutually beneficial promotions events
Effective management of trade promotions requires end-to-end collaboration between CPGs and their retailer customers. AI capabilities can provide a central hub for creating and deploying mutually beneficial and impactful promotional events, resulting in tactics that enable CPGs to maximize revenue and maintain margins within their desired thresholds, while also maximizing sales lift for the retailer’s category. What’s more, CPGs can then analyze the effects of switching between brands at a retailer as well as switching between retailers within a market.
5. Compliance management
How a promotion is executed has a direct impact on its performance, and retailer execution can vary in several ways, such as if a promotion is not activated on time, or if graphics are placed incorrectly in a circular. Considering CPGs pay retailers trade incentives based on expectations about how well the promotions will be executed, compliance is essential. Historically, compliance is measured through inaccurate and costly manual processes, adding additional workload for busy store employees. Connecting AI-enabled mobile apps to the supply chain, merchandising and planning system of record gives users immediate access to real-time compliance reports across their entire geography, ensuring pricing, stock and promotional compliance data are accurate at an individual store level.
6. Continuous analysis and learning
Only with artificial intelligence can the three stages of promotion – forecasting, planning and evaluation – be connected at scale to gain the efficiency and accuracy needed to achieve the greatest impact. Many technology solutions in the market today help retailers and CPGs streamline execution, but do not effectively connect planning and forecasting stages. What’s more, it’s not humanly possible to forecast what-if scenarios, as supermarkets tweak promotions every week, with millions of possible combinations of SKUs, variables and constraints to consider. With the ability to continuously learn from data, AI and machine learning provide the perfect opportunity to analyze patterns to help understand consumer motivations and actions, which provide account managers informed insights on the most profitable promotions they should run.
7. Self-learning models
Most retailers use a forecast model with two or three years of historical data to predict sales, volume and margins. However, these models can quickly become stale and accuracy can degrade if not maintained. Machine learning can not only continuously tune thousands of models, but also recognize and incorporate trends and new factors as they start to become significant. Such agility at scale is simply impossible with traditional forecasting systems, which require highly skilled staff to tune and maintain each individual model.
Using AI-enabled solutions to drive revenue growth
Artificial intelligence can help CPGs promote products with far greater precision through much more accurate data analysis and create promotions that benefit both retailers and CPGs. The insights and guidance AI provides enables CPGs to allocate trade fund dollars more wisely and alleviate margin erosion, leading to an increase in conversions and sales.
To learn more, read Making Promotions Profitable: 7 ways AI is transforming promotions.