The marketing landscape is saturated with digital channels, retail media networks, trade publications and other traditional channels. In other words…it is complex and it makes the role of a marketer difficult. There are thousands of decisions that need to be made when activating any omni-channel marketing plan.
In this blog we will discuss why marketing mix modeling utilizing machine learning is superior to traditional methods. When marketers are armed with the data required to make future marketing investment decisions, they can confidently defend that investment, secure their budget and job, and help their organization achieve it’s growth goals.
Introduction to Modern Marketing Mix Modeling
A Brief History of Marketing Mix Modeling
For more than 30 years, marketing mix modeling has been the standard among large brands; keep in mind this solution predates the internet and digital marketing. MMM uses regression analysis to calculate a short-term ROI (12-16 weeks) for programs that have been fully implemented. Using this data, marketers make future investment decisions with a goal of meeting or exceeding that past return.
Traditional marketing mix modeling is a heavy lift, in terms of time and cost, so typically it’s generated annually to support the budget planning cycle. It’s not designed for decision support; it’s up to the marketer to assess the data for decision-making value. Also the short-term ROI fails to address:
- The longer return window on brand-equity programs
- The interaction effect between and among channels
- The impact of market and environmental factors
But perhaps the biggest shortfall of traditional marketing mix modeling is that it fails to consider the full market opportunity, instead assuming that what has been done is the best mix. There is no “what-if” scenario capability, and as a result, marketers are limited to a low-risk, but also low-return view of how to drive incremental improvements in the status quo.
The Limitations of Traditional MMM
There are several limitations with traditional marketing mix modeling (MMM) Including::
- High cost, slow speed, limited flexibility, while relying nearly 100% on historical data and offering only outdated reporting.
- Typically a standard regression analysis is used, which involves fitting a linear model to the data. These models can be limited in their ability to capture the complexity of real-world marketing environments, and may struggle to account for factors like seasonality, non-linear relationships between variables, and interaction effects.
- Results are received too late and in a summary format that is not actionable.
The Evolution of MMM: From Traditional to Machine Learning
The Integration of Machine Learning into MMM
Mixed media modeling that utilizes advanced modeling techniques, such as machine learning and causal inference, provide a more thorough understanding of the relationship between different touchpoints and conversions. This results in deeper insights into customer behavior, enabling businesses to make more informed and effective marketing decisions. Additionally, the use of these techniques allows businesses to simulate and test various marketing strategies and scenarios in a virtual environment, allowing for evaluation of potential impact before real-world implementation, thus minimizing the risk of costly mistakes.
Why Machine Learning is a Game Changer
Marketing mix models like Keen’s that utilize machine learning are more agile, adaptable, and responsive than traditional MMM. The approach is focused on the current and the future, providing marketers with the insights and recommendations they need to make informed decisions in real-time.
Bayesian Methods: The Heart of Adaptive MMM
Understanding Bayesian Methods in MMM
Bayesian models use probability theory to estimate the likelihood of different outcomes and the degree of uncertainty associated with those estimates. This framework enables the model to incorporate prior knowledge about the relationships between marketing variables and business outcomes, which can improve the accuracy of the model. Additionally, by addressing uncertainty with a monte carlo simulation, Bayesian MMM provides a range of possible outcomes with associated probabilities, rather than a single point estimate. For example, if a company is deciding between two marketing tactics, a Bayesian marketing mix model can estimate the probability of success for each tactic and offer several potential scenarios, which leads to more informed decisions about how to allocate marketing resources.
The Role of Tactic Elasticity in Marketing Decisions
Keen brings value with our patent-pending Marketing Elasticity Engine as our knowledge estate. This is informed by meta-analysis of 40 years of academic models, 10 years of client metadata and data from recent marketplace dynamics, and thousands of economic elasticity models which acts as our custom priors for your brand. These priors can be influenced through the regression model when we load in your data (time series spend, clicks, impressions, COGS, etc.)
Revolutionizing Decision-Making in Uncertain Times
Scenario-Based Planning: The Key to Navigating Uncertainty
There’s no shortage of uncertainty when building a marketing plan. Considering a pandemic, social unrest, wars and inflation which all play into how marketing investment decisions are made. Scenario-based planning in marketing emerges as a strategic approach designed to help brands navigate through uncertainties with confidence. This methodology involves identifying and analyzing possible future events or conditions—ranging from economic shifts, technological advancements, competitive movements, to consumer behavior changes—and developing strategies to address each potential scenario.
By employing scenario-based marketing planning, brands are not only able to mitigate risks but also seize opportunities that may arise from future market dynamics. This approach requires a thorough understanding of the current market landscape, predictive analysis, and creative thinking to envisage various future states. Scenario planning equips marketers with a proactive blueprint, enabling them to adapt quickly and maintain a competitive edge in ever-evolving markets.
Asking the Right Questions with Keen’s MMM Software
When uncertainty arises, there are still decisions that need to be made. Figuring out what those are while demonstrating results and earning buy-in from the leadership team is what Keen enables marketers to do. Our system empowers you to answer questions like these:
- How high can we profitably spend across our media mix?
- What is all this brand spending doing for our bottom line?
- How can we start to un-silo our media measurement?
- ROAS seems high but where’s the tangible uptick in sales?
- What will happen to sales if we shift or change channel spend?
- How much would we have to spend to hit this revenue target?
- Are we squeezing as much as we can out of this budget?
Every dollar you spend in one channel has an impact on every dollar you spend in another. So, when you are creating a plan or even making a quick adjustment, understanding all of the intended and unintended consequences is a challenge. As you evaluate channel and timing options, our system quickly accounts for how your marketing decisions contribute to or distract from your targeted financial outcomes.
Unlike traditional marketing mix modeling tools, the Keen Platform provides brand leaders with a continuous decision loop. Every week, every month, every quarter, new data becomes available. This new data is appended to our system to inform the most recent coefficient. So, at any point in time, you can re-run your plan and proactively make adjustments based on the latest information.
Optimizing Your Marketing Spend with Machine Learning
Machine learning technology can synthesize vast amounts of data in minutes, giving marketers a view into how a campaign is performing in real-time. This helps marketers adjust their campaign strategies if necessary, allowing them to optimize its performance throughout its life cycle.
Marketers want to feel like they’re working on the cutting edge of the industry by using the latest technology and data analytics tools. In a recent study conducted by Keen, we surveyed 600+ full-time marketers at mid-size to enterprise size companies and discovered that almost half of marketers not using AI technology would prefer to work for a company that does. As such, it’s important for companies to invest in tools like AI and machine learning to maintain a competitive edge and ensure that they’re at the forefront of their industry. It will also free up their employees’ time for other revenue-generating activities, which can help boost the company bottom line.
Seasonal Spend Optimization for Maximum Impact
For Dramamine, the Keen Platform unlocked marketing opportunities that extended beyond their traditional season. Historically, the marketing team focused solely on marketing during peak seasonal periods and was hesitant to invest in other times throughout the year.
By leveraging the Keen Platform, the brand was able to evaluate the marginal return on their investments. This highlighted that profitable revenue could be unlocked in the off-season period. Through planning simulations, they quantified the value they could create for the business by incrementally investing during this time.
Dramamine saw a profitable 41.8% in incremental revenue and secured an additional $2.6M for 2H marketing support and a 9.5% increase in marketing ROI.
Case Studies: Real-World Success with Adaptive MMM
Enhancing ROI and Business KPIs with Keen
A leading challenger brand needed to maintain their growth from marketing efforts while contributing positive net profits. To sustain their growth, the brand needed to strategically optimize for profit.
The team used Keen’s platform to drive this pivot and find the right mix that would achieve their growth goals while doing so in the most profitable way possible.
Keen’s platform and service helped drive both revenue and profit with less investment. The brand saw a 52% net profit increase by spending 15% less. The brand has learned that their investments must continue to balance growth with profitability to be sustainable.
Long-Term Benefits of Top-of-Funnel Media Activity
For our next example, let’s take a look at how Keen’s recommendation to move to an always-on marketing strategy in online video (OLV) benefited one coffee brand substantially. This growing coffee brand wanted to figure out a way to substantially increase their marketing investment without sacrificing their profitability as a result. After running several plans to determine which channels had room to grow investment and remain profitable, OLV was identified as showing opportunities for optimization.
Per Keen’s recommendations, moving to an always-on strategy in this tactic led to a huge increase in effectiveness on a 113% spend increase. The team saw an increase in same year profit ROI from $1.03 to $1.83 and an even larger increase in all years profit ROI from $1.83 to $3.19.
The Future of Marketing Strategy: Data-Driven and Adaptive Marketing Strategies
This is where we tell you that Keen can be your next-generation marketing mix modeling partner. Our platform’s machine learning and AI insights are a trustworthy source that are forward-focused, decision-oriented and accessible, all in service of your financial goal. The Keen Platform is a SaaS based tool that is considered a unified marketing measurement and optimization solution.
We deliver:
- The data necessary to help marketers make the best data-driven decisions possible.
- The accuracy and confidence of precise response curves to help you decide the best marketing mix for your future marketing investments.
- Insight as to how your programs are performing in-flight.
- The speed, agility and accessibility you need to continually optimize your marketing investments and react to external forces as distribution and competitor activity.
See for yourself how Keen’s marketing mix modeling with machine learning is changing the way marketers are making data-driven decisions by taking a tour of the platform.
