This is the first part of a blog series that introduces the concept of explorative simulations and takes you on a journey to the state-of-the-art methods that can be leveraged to gain competitive edge in the consumer marketplace with Trade Promotion Optimisation (TPO) – a process which factors in goals, promotions, supply constraints and predictive analytics to strategize continuously for improving trade promotion initiatives. 

Our key observation, given our extensive experience in the CPG space with Fortune 500 customers, has been that over one-third of consumers have shifted brands or have at least experimented with new ones in the last year, primarily due to marketplace behavioral changes triggered by the pandemic and promotional offerings.

One of the primary levers in a Strategic Revenue Management (SRM) framework, is promotions; supported with pricing, assortment and trade investments. Past analyses have shown that 28-55% of retail sales volume across CPG categories (non-food, ambient food, FMCG, alcohol) in Europe is heavily influence by promotions. This contributes directly to the topline KPIs for any retailer/manufacturer.^[1] Trade promotion is one of the key line items for a CPG company, consuming well over 20% of gross revenue.^[2] The B2B commerce is fundamentally distributed and involves lots of interaction across various functions. Companies trying to break out of flat sales should re-evaluate trade promotion from an optimisation lens. The idea behind trade promotions is simple:

• Spend money with a retailer to raise your topline KPIs directly affecting your PnL
• Invest in strategic promotions to improve household penetration or market share
• Acquire the next generation consumer base
• Improve success rates of new product launches

At the same time, it is crucial to keep in mind the vision and strategies of the partner (retailers) ecosystem. A well-designed promotional calendar would not only help in effectively planning the above-mentioned initiatives for a manufacturer but would also help in aligning well with the KPIs of a retailer for whom the promotion is being designed. For instance, if a retailer cannot shell out the assortment shelf space or banner space, the negotiation falls apart or the implementation is delayed.

Enterprises, these days, are gearing focus towards precision-SRM – a data-driven, analytics-powered, and action-oriented framework that empowers them to enhance their capabilities on their promotional performance. Over the past decade, businesses have set up many central and dedicated teams to nurture this process, thus, making the marketplace even more competitive.

The vanilla TPO framework

With data science gaining maturity within enterprises, the consumption and usage of machine learning and artificial intelligence to derive actionable insights is on the rise across business functions. Understanding various micro-factors across micro-segments are crucial to drive macro-outcomes. In the study of our TPO framework, we see that attribute specific to consumer segments, geography, categories and products drive margins for enterprises. The end consumption tool is a simulator that business leaders can use to create simulations of business processes and scenarios to evaluate the impact on their goals or targets and act accordingly to impact the macro-outcomes of topline enterprise KPIs. The simulator that we built is from a predictive model and this is called explorative simulation. The scenarios considered in the simulator are the cases that have occurred in the past.

Data enrichment for visibility and tracking

To develop a granular level understanding of the consumer marketplace, we need to ensure a neat data enrichment framework to gain insights on the following attributes –

• Volume insights – Accurate understanding of volume related metrics to depict business health and topline KPIs
• Shopper data – Precise understanding of end consumer price, price gaps, price ladders
• Promo mechanics – Impact of different attributes of an offer on promotional events
• Causal metrics – Impact of merchandising elements on promotional performance
• Financial planning – Granular records of trade spend effectiveness and ROI metrics across investments
• Geospatial – Impact of local and geographic distribution on event performance
• Correspondence & household data – Understand consumer-consumption behavior, psychographics and segments

Occasion mapping and segmentation

Often, the granularity of syndicated data segments is not in line with business definition and execution. A translation or roll-up is required to a level where it can be consumed for redefining consumer segments (persona) and a time-level roll-up is required for model building. Integrating this into various promos, weather, and events calendar delivers intelligent insights and levers for mapping external triggers into the model.

Choice model

This is a discrete model that analyzes and explains consumer choices based on a promo consumption. This allows us to dig deeper and leverage insights to understand what works and what does not. Keeping everything as-is and changing the dependent variables, we can model for both margin and market share penetration. All granular level information, underlining product definition, preference coefficients, segment drivers, and utility definition, is typically used and built at week level granularity, enabling promo decisions for a short-term (week level planning) across retailers.

Simulator for promo pre-planning

The simulator, here, shows multiple promo combinations at a category level. This exploration tool can recommend optimised promo calendars for each group of retailers. In recent years, organisations have been extending their simulator tools with vendor-view capabilities, depicting a retailer-level KPI dashboard for building promos that benefit both businesses.

Seizing the untapped

The above framework is an iterative process across multiple scenarios, since the crux of its implementation is based on negotiations driven by business representatives. As of today, AI/ML is still evolving to show promising results. The TPO framework is built on the premise of past variations and holds multiple constraints and is often unable to cater to simultaneous effects in the B2B marketplace. These explorative simulations are highly aggregate in nature reconciling top-down or bottom up or both. Even though we can use the same framework to design like-for-like products for new product launches, there is significant scope of improvement. The vanilla framework has been commoditised over the years and is squashing innovation.

Stay tuned for our next segment to get a glimpse of an innovative “what-if” simulation against one of the traditionally mature econometric perspective of the TPO world.

Footnote:
*Promotion and promo have been used interchangeably. Both imply the same idea.
[1] Tim Eales, “Price and promotion in Western economies: A pause in promotion escalation,” IRI, November 2016, iriworldwide.com.
[2] George Lawrie, “Use trade promotion management to boost return on promotions,” Forrester, September 24, 2019, go.forrester.com.