By Admin, Dec 06, 2019
As a marketer, you feel the constant pressure to show how you are delivering value for the organization. The rise of marketing technology should be making it easier. While multi-touch attribution helps to understand how digital activities contribute to sales, it isn’t designed to account for offline activities. Therefore, companies that sell through both online and brick-and-mortar stores still heavily rely on marketing mix modeling as their primary means of demonstrating ROI and optimizing their marketing strategy.
Marketing mix modeling dates back to the early 90s before the rise of digital. It was designed to identify correlations between in-store sales and time-specific media channels like TV and couponing. Today, many analytics consultants use a blend of marketing mix modeling with multi-touch attribution. The theory is that by blending these two approaches one can provide a unified marketing analysis.
However, many question whether this is effective. At an Attribution showcase event held by the Advertising Research Foundation, there were many questions about the trustworthiness of blending these methods because data isn’t uniform and there are holes due to collection methods. Many attendees also questioned validation approaches, as there was no standard method of validation across providers.
Personally, I’ve sat in many a meeting where marketing mix modeling reports are presented to only hear the distrust in the room. After the meeting, everyone would go back to their desks to keep working on what they were already doing because the report had come in too late to make any changes anyway.
I heard one CMO even say, “The only reason we do marketing mix modeling is for air cover.”
Two years ago, I set out to find a better way. In a world where Waze can tell you how to avoid traffic, Netflix knows what you might like to watch next and Facebook knows just which ad to show you to get you to buy, I thought there had to be a do-it-yourself technology that could address this problem and the business need for faster, trustworthy answers about ROI for both digital and non-digital activities.
That is when I discovered market simulation. Since then I’ve started working for Concentric – a market simulation software company - and I’ve heard this diatribe from many of our customers looking to build a better way to meet the needs of marketers. Here is why many analytics leaders are deciding to make the switch to market simulation:
1. Unified attribution results on a weekly basis. In today’s fast-moving digital world, marketers need access to real-time analytics to inform strategic decisions. Market simulation provides a single modeling approach that unifies both online and offline activities, including events, PR and even in-store activities. With automated data feeds set up to constantly refresh models maintained by an internal data science team, business leaders can access insights in real-time for a unified view of a market.
2. Recommendation engine for marketing mix optimization. Just like Netflix can give you recommendations on what you might like to watch next, market simulation helps you figure out how to allocate or even reduce spend to maximize results. There are two options available for doing this. You can auto-optimize your media mix using machine learning or you can run your own what-if scenarios based on a variety of strategies you are considering. Optimization through market simulation is flexible in its scope and is easy to configure and use. For example, maybe you are considering increasing your digital budget while constraining your total marketing spend to a certain level. Or maybe you are interested in putting more investment in in-store displays to respond to a competitor threat. By running what-if scenarios, you can easily compare how likely each of these strategies is to create a lift in business outcomes.
3. Forward-looking perspective leads to more actionable insights and trust. Market simulation doesn’t stop at attribution. That is only part of the story. In order to make analytics actionable and trustworthy, you need a way to generate predictive insights. With market simulation, marketers have a way to forecast based on the investment choices they’ve made in their marketing plan to see whether they are likely to hit their sales targets. With an attribution plus forecasting capability, it is possible to build greater trust in the modeling approach over time by experiencing whether the forecast was in fact accurate.
4. Access multiple perspectives of optimization from one model. Marketing mix models are typically constructed through the lens of the CMO. While this is helpful for setting budgets across marketing managers, it doesn’t necessarily help those marketing managers decide what creative messaging to use or which specific tactics to employ. Market simulation gives you the flexibility to run attribution analyses at different levels of granularity without needing to rebuild the model.
5. Trusted results because it is managed in-house through a collaborative process. Market simulation software is designed to enable internal teams to build and maintain their own models without relying on outside consultants to generate reports. Internal cross-functional teams work together to build their own models using a 5-step workflow that doesn’t require any coding. The final report doesn’t feel like it’s based on a black box of mathematics because everyone is aware of the data that was used and the assumptions made. It is this collaborative process that helps to build trust in the findings and the team starts to gain a deeper understanding of what strategic levers need to be pulled to influence outcomes.
6. Scalable process across multiple brands and metrics. Oftentimes, companies will have a portfolio of brands, yet only some can afford to invest in marketing mix modeling. With market simulation, analytics teams can create a standard process that provides all brands with access to the same insights helping to drive better business performance across business units. Marketers are also often interested in the impact that marketing has on metrics other than sales, such as brand awareness or consideration. A single market simulation can provide attribution results for sales, brand health metrics, and word-of-mouth. This enables CMOs to have a holistic view into the effectiveness of their marketing.
Do you think market simulation might be worth exploring?