Demystifying brand and long term effect measurement in Marketing Mix Modelling

Did you recently make a significant purchase, like a new phone, car, furniture or even jewellery? If yes, then you are probably not buying them again right now, nor will you for quite a while. According to research performed by the Ehrenberg- Bass Institute for Marketing Science (2021) there is a 95-5 rule in marketing, which means that in general 5% is in-market at any given time. In other words, 5% of consumers are ready to buy at this moment. Consequently, 95% of consumers are currently not in-market. This percentage is of course heavily dependent on the type of products and/or services a company provides so it is not meant as a precise measure. Rather it is used ​​as a heuristic to get the idea across that the vast majority of consumers and businesses, for a large proportion of products, are not in the market in particular time periods.

This rule suggests that marketers should not just focus on the small, immediate market but also work to build mental availability among the larger audience who may purchase in the future.

Mental availability is one of the two key concepts from the book “How Brands Grow: What Marketers Don’t Know” by Byron Sharp (2010). In this theory Sharp explains that in order to increase the chance to be chosen by consumers in competitive markets, brands need to focus on both mental and physical availability.

Mental availability refers to the likelihood of a brand being thought of in a buying situation. In other words, the more easily and frequently a brand is thought of in relevant buying situations, the higher its mental availability.

On the other hand, physical availability refers to the ease with which consumers can buy a brand. Maximising physical availability means making the brand easy to find and buy, which can involve broad distribution, prominent placement on shelves, and maintaining adequate stock levels. Ensuring that a brand is both easily thought of and easily bought can significantly enhance its market performance.

Traditionally, we can identify two different marketing strategies that generate revenue, namely performance marketing and brand building. Performance marketing encompasses online marketing and advertising programs in which advertisers pay marketing companies or advertising platforms when a specific action is completed. This action can include a click, a lead, a sale, or another predefined measurable engagement or conversion metric. Therefore, performance marketing is related to an immediate action and more to the currently in-market consumers.

On the other hand, brand building is related to mental availability and therefore about creating and enhancing a brand’s identity, awareness, and perception to establish a positive and distinct image in the minds of consumers. Unlike performance marketing, brand building is not about a short-term measurable action but about taking a longer-term perspective. Therefore, brand building is more important for the currently not in-market consumers in the sense that an enhanced awareness or image of a brand may affect future buying decisions.

To start with, the effect of marketing on immediate buying decisions or the short term effect is generally measured with Marketing Mix Modelling (MMM). Marketing Mix Modelling consists of a statistical model that credits both online and offline marketing channels and calculates their effects on many different targets like online sales, retail sales, and/or traffic. It can also consider factors like competitor actions, pricing, economic conditions, holidays, and weather. By analysing the impact of each element on business outcomes, companies use MMM to allocate budgets across various channels and activities.

Many companies have adopted MMM, moving beyond click-based attribution models, but with an effect-window of a couple of weeks/months it is still primarily focused on in-market consumers. To measure longer windows of effect and therefore buying decisions of current not in-market consumers, an MMM solution needs to include brand building effects / increases in mental availability.

This is easier said than done. Data related to brand building is limited in availability and quality. Next to that, how do you translate marketing theory as described above into a statistical solution?

To integrate long-term brand effects into Marketing Mix Modelling (MMM), several crucial steps must be followed. Firstly, acquiring the right data is essential. Ideally, this data comes from brand trackers that monitor various brand metrics on a weekly or monthly basis. Additionally or in absence of brand tracking data, social listening data can be utilised to track the reach and sentiment of your brand and product mentions in online conversations. Search data, such as Search Query Volume, can also be valuable for understanding how many people are considering buying your product or service. This data can help calculate the share-of-search or a brand’s branded search volume from paid search ads.

Secondly, it is important to investigate which of the available metrics correlate with media investment and revenue. For instance, a well-known brand with over 95% brand awareness within its target population is unlikely to drive further awareness with additional marketing investments. However, consistently communicating a specific message can influence consumers’ perceptions of the brand, potentially unlocking future sales. A new metric, which can be termed “brand equity” or “brand power,” is formed using the highest correlating brand metrics. Since the concept of a brand is multifaceted, this new metric should encompass multiple brand metrics.

Thirdly, the MMM model should be modified to calculate two brand effects. The new metric, brand equity or brand power, affects long-term growth in revenue, represented by the baseline in MMM. Additionally, a strong brand enhances short-term activation efficiency. For example, when more people are aware of and considering your brand, it becomes easier to convert more consumers in the short term.

Lastly, marketing budgets should be optimised based on total return-on-ad-spend (ROAS) rather than short-term ROAS. By implementing brand equity into MMM as described, it is possible to quantify the extent to which the brand has impacted baseline revenue over time and improved the effectiveness of short-term marketing efforts. When building different budget scenarios for the future, the model should forecast these effects alongside existing short-term sales effects.

In summary, integrating long-term brand effects into MMM involves gathering the right data, correlating metrics with investment and revenue, modifying the MMM model to account for both long-term and short-term brand effects, and optimising marketing budgets based on total ROAS. This comprehensive approach ensures that both immediate and future impacts of brand building are accurately measured and leveraged for optimal marketing strategy.