How can analytics help you make better marketing decisions in times of uncertainty?

Use analytics to bring data sources together, to identify new trends and leading indicators as they emerge and make confident business decisions.
28 April 2020
Analytics
Andy Hawkins
Andy
Hawkins

COO, Analytics Practice

Get in touch

Businesses need to make big, bold decisions against a back-drop of unparalleled economic and social uncertainty. The businesses that are making these decisions with most confidence are those with access to a wide variety of consumer data sources that are refreshed and updated regularly to keep pace with change. These businesses also have the data curation skills to sift and prioritise what’s important, along with the data engineering skills to harmonise and unify data sets. Lastly, they possess the analytics capabilities to use this data to predict consumers’ needs in the coming months and quarters.

Analytics is uniquely placed to aggregate and harmonise a wide variety of structured and unstructured data sources and to build constantly-learning models to identify new trends and leading indicators as they emerge. It’s this capability that will create the predictions to support confident decision-making.

The immediate questions that CMOs should be addressing that use analytics are:

  1. How do I forecast demand in the current environment?
  2. How do I make the right decisions on marketing spend – what to cut and where to invest?
  3. What are the messages my brand should be focusing on?

These questions all have an immediate dimension: How do I plan the next 4-12 weeks? And a medium-term dimension: How do I plan for recovery in a new and uncertain reality for the next 6-12 months? Trying to solve both dimensions at once is a significant challenge. We recommend splitting each question into different time horizons to support the immediate decisions that must be decided now, but to allow time and consideration for structured planning for the medium-term.

Forecasting needs both bottom-up and top-down thinking

The consumer brands that are forecasting demand with the greatest confidence are those that are combining bottom-up and top-down approaches.

When forecasting for immediate decisions in the next 4-12 weeks, it makes sense to extrapolate from detailed weekly sales data if this is available. For FMCG brands, for example, with access to this granular data from a reliable panel source, working bottom-up from this is the most pragmatic approach. Analytics to integrate real-time social media insights into consumer thinking is a vital complement to this approach to add some real-time human and cultural insights.

For forecasting further into a very uncertain future recovery, a scenario-based, top-down approach is probably more useful. Kantar has created a framework of scenarios that considers options for how the disease will behave, how institutions (critically government) will respond, and how consumers react to these dimensions. Predicting which countries are most likely to be in which recovery scenario is a first step. This may benefit from the judgement of skilled futurologists, but it’s predictive analytics, tracking available specialist data sources and forecasts, that can quantify these predictions. This can be followed by building category-level forecasts to understand how consumers are most likely to react in different areas of their lives, within each of the macro-economic recovery scenarios.

This exercise won’t be one-off, so businesses need to detect the most important lead indicators of change for their businesses and keep these regularly monitored to refresh the scenario forecasts and demand predictions. Understanding the dynamics of consumer demand will be key to medium-term planning – new occasions, motivations and segments – but sizing overall trends is the place where businesses need to start.

Marketing spend decisions

For businesses unable to trade at all, the short-term decision has been taken out of their hands. But for brands with choices to make, the most effective are approaching this in two distinct stages.

The first is to make some immediate marketing investment decisions, using data that is readily available to model the impacts of different spend levels and media allocations. Ideally, this should use real data relevant to your market and category as well as taking your competitors into consideration. You will not be alone in changing your plans. Benchmarks and norms from pre-COVID periods and previous recessions, factored into a simple simulator approach, will clearly not be a perfect predictor, but it can be realised at speed and so can guide the immediate decisions that need to be made. Whilst it’s tempting for short-term sales impacts to drive your immediate decisions, the long-term brand implications of the decisions you make today are also critical. There is no point making short-term savings that will jeopardise your recovery tomorrow.

The second stage is to plan properly for the medium-term. This should link back to the scenarios upon which your business is forecasting customer demand. Even if your overall demand levels are expected to recover, our early experience is suggesting that your pre-COVID marketing allocation decisions are unlikely to be reliable. Consumer brand preferences, media elasticities and media prices may have shifted and might continue to change, along with difference in the macro recovery patterns between markets (at this stage its simply too early to tell). It is therefore critical for brands to review their marketing decisions, but to do so as an ‘always-on’ exercise, using new post-COVID data as it becomes available and constantly adapting and learning. In many ways, this is traditional marketing ROI work, but it’s now critical for this to be a ‘live’ monitoring exercise, not a project that is run every 12 months.

Brand messaging

Our experience into this period of uncertainty has so far suggested that brands would benefit from re-considering their immediate messaging. Initial evidence indicates that leading consumer brands in many categories are gaining in market share as consumers seek the brands they trust, which will be of comfort to many brand managers. However, we also know that consumers are expecting more from the brands they trust at the moment and will be judging brands by how they respond and adapt. This relates to meeting consumer’s own needs, whether this be for reassurance or safety, or in meeting the needs of society by showing a genuine sense of social responsibility and authenticity by delivering tangible benefits back to society.

As brands emerge into recovery it’s therefore important to constantly review the messages that are resonating with consumers , that will re-build your brand. It’s the same approach for marketing ROI, requiring constant refreshing of new data to respond as markets and consumers change.

Get in touch
Related solutions
A comprehensive assessment of marketing ROI, measuring short and long-term impact of marketing on sales and brand equity.
Analytics of customer-level data to show the importance of different customer touchpoints to brand or sales outcomes.
Networked transactional, behavioural and attitudinal data to help you understand motivations and become more meaningful.