YOUR ADVANCED ATTRIBUTION FRIEND

Marketing Mix Modeling for CPGs in international markets: A Case Study with Dr. Scholl’s

  The Brand Dr. Scholl's is a renowned foot care and orthopedic footwear brand with over a century of history. Founded in 1906 by podiatrist William Mathias Scholl, the company has become a global leader in foot care products and solutions. With over $600MM in international retail sales between the Dr. Scholls brand in the US and the Scholl brand internationally   The Challenge The brand lacked a good sense of what media was driving in terms of retail sales and wanted to understand where the opportunities were to invest and drive efficient growth. The company was also embarking on a Performance Marketing initiative and needed to understand what measurement could look like in that context.  The brand operates in a multi-channel retail environment that includes brick-and-mortar retailers like pharmacies and supermarkets, Amazon and its own D2C channel. Exploratory Data Analysis As part of the exploratory analysis, marketing measurement teams have to frame outcomes that marketing can be reasonably tasked to drive, but also contemplate certain marketing investments in the context of their business relationship with retail partners. This is particularly the case with trade and promotional investments with specific retailers. The M-Squared team had several data sets to assess and synthesize. These datasets included media investment and delivery data across a diverse mix of channels and tactics going back 2 years. The data also included retailer-specific trade and promotional investment. The media channels evaluated included Google Search (brand and non-brand), Facebook, Retail Media Networks (RMN), Digital Display, TV+VOD, and Out of Home (OOH). The Approach These large and diverse datasets needed to be processed and transformed to build the Analytics plan. The analytics plan developed as part of a structured Advanced Attribution Audit was designed to address the questions inspired by the business and entailed: Data Harmonization: Collection of the historical data (2 years) on units sold, retail sales, and media spend/activity which was reviewed and processed as per modeling needs. Preliminary Analysis: Performing trend analysis, correlations, and a basic MMM to understand the fit of individual variables driving units sold and retail sales. Further reviewed the retail store categories and created a hypothesis to determine the number of MMMs to run. Media Mix Modeling: Running 1000’s of iterations and 100’s of tranches for the different segments to come up with the best-fit models explaining the drivers of retail sales - the result of which was a 6-model structure based on retail category groupings.  Triangulation: Using the MMM decomps, did a triangulation exercise to understand the impact of media, promotions, and the value they bring in driving units sold and retail sales. The Marketing Accounting Framework (MAF) was oriented around incremental retail sales (revenue) driven by media across the retail category groupings. This was distilled into incremental ROAS (iROAS) by media channel/platform. This was a 6 P&L structure - meaning 6 different models were built - one for each of the retail categories. Learn about Marketing Accounting Frameworks with this short course from M-Squared: The Results Media drives incremental impact/revenue - nearly  12% of retail sales were driven by media and promotions on an incremental basis with a iROAS range of 0.2 - 2.4. Trade and promo spend had nearly  double the investment  as National Media & RMN’s but  less than half the iROAS. An opportunity to optimize trade and promo spend based on top-performing tactics. Amazon and other retailer-specific trade and promotional spend  have a halo on sales in other retail categories. National media influences  nearly all retail categories. Amazon RMN exhibits significant  cross-retailer halo  which suggests consumers are potentially  using Amazon to comparison shop. Retailer-specific promos have an impact outside the specific retailer running them. This may also be a comparison shopping situation. Overall, the insights suggested there was growth opportunity within the existing budget based on making some investment shifts to higher-performing (based on incremental retail sales & iROAS) media tactics. There also appeared to be an opportunity to optimize spend within trade and promo spend understanding investment there is required to maintain the relationship and is something of a “cost of doing business”. Working with individual retail chains to invest in the highest performing tactics based on their incremental contribution and ROAS is a branded or presented recommended approach. The Conclusion The model outputs and Triangulation surfaced some interesting and actionable insights. The project also raised some additional hypotheses and questions. What might this look like across different regions/geography? Would different product segments perform differently based on media-influenced retail sales? These are questions that can be answered through testing and model iterations to contine to refine the insights and decision-making information. As the Scholl brand continues to diversify and modernize its media mix, M-Squared will continue to identify opportunities to leverage advanced attribution as a competitive advantage, make smart bets, and identify growth opportunities. This is just the first step in the beginning of a journey.  

Case Study: Reviving Revenue Through Optimal Contact Strategies and CRM Experimentation for RBX Active

In today’s competitive landscape, direct-to-consumer (DTC) brands face unique challenges, particularly when growth begins to plateau or decline. Our case study focuses on an established athleisure brand experiencing a troubling trend: a -12% year-over-year revenue drop. As the brand has successfully expanded its audience beyond millennials into an older demographic, it must pivot its strategies to ensure sustainable growth.   “When working with brands, we recognized that the first-party data they collect—their customer file data—is one of the most valuable assets at their disposal.”- Cara Manion   Brand Overview RBX Active is a DTC athleisure brand known for its high-quality yoga pants, hoodies, and tees. The pandemic propelled significant growth for the brand, but as the market stabilized, so too did its performance. With Black Friday and Cyber Monday (BFCM) approaching, there was an urgent need to reverse this trend and reignite momentum.   Go-to-Market Strategy The company has traditionally focused on prospecting, targeting specific audiences through Facebook media and email marketing. While these efforts successfully drove customer acquisition, our study highlights the opportunity for enhanced retention strategies. Currently, retention initiatives rely on hand-picked audience selections, primarily targeting individuals who visited the website but haven’t opened an email in the last six months. This approach presents an opportunity for improvement, as the brand is still exploring how effectively these efforts drive incremental performance. By incorporating data-driven insights into retention strategies, the company can strengthen customer loyalty and maximize long-term value. Attribution Challenges RBX Active utilized Facebook’s reported metrics and their own analyses to track performance, but there was some confusion regarding the impact of their media spending on sales. This uncertainty led to questions about the effectiveness of their audience targeting and the return on investment. We engaged with them around September, aiming to run the test at the end of October or early November. This timing was strategic, as it allowed us to activate insights for the critical Black Friday/Cyber Monday period. With these important sales periods approaching, anxiety about making the right decisions increased, highlighting the need for clearer attribution and more reliable performance metrics. Objectives The primary goals of this initiative were to: Understand the incremental impact of Facebook media on customer retention and revenue during the BFCM period. Identify a contact strategy that maximizes customer engagement and increases revenue contribution from existing customers. Optimize media spend to ensure a more efficient allocation of resources during peak sales periods. Reverse the current -12% trend in customer retention, focusing on strategies that effectively re-engage lapsed customers and drive sustained revenue growth. By addressing these goals, we aim to create a robust framework for improving customer loyalty and maximizing the effectiveness of marketing efforts during critical sales periods. Methodology: RFM Analysis and CRM Experimentation When we engaged with the client, there was a prevailing assumption that their audience selections were effective. However, our data analysis revealed otherwise, indicating significant opportunities for optimization. To address this, we deployed a two-pronged strategy. First, we focused on the RFM (Recency, Frequency, Monetary) model to gain a clearer understanding of the composition of their customer file. This foundational analysis allowed us to identify key segments and behaviors within their audience. Next, we conducted a benchmarking analysis comparing customer behavior during the 20 days leading up to Black Friday and Cyber Monday (BFCM) in 2023 versus 2022. We specifically examined the change in revenue per customer ($/Customer) and uncovered a concerning trend: the lapsed segment—customers who hadn’t purchased in over 18 months—was declining by 35%. This decline emphasized the urgent need to re-engage this group, particularly since many customers acquired during the pandemic were not returning for subsequent purchases. From a P&L perspective, this scenario is problematic, as it is costly to acquire customers only to lose them shortly after. These insights laid the groundwork for our targeted re-engagement strategies.   Testing After performing the RFM analysis, it became clear that lapsed buyers were not returning, necessitating a focused effort to re-engage them. Our hypothesis was that the brand's hand-picked, conversion-optimized audience targeting strategy on Facebook might be leading to less incrementality. This approach typically targets individuals more likely to convert, which tends to favor recent buyers and overlook those who had lapsed. To test this, we aimed to understand the impact of Facebook’s conversion-optimized campaigns compared to reach and frequency (RF) campaigns. At a high level, the conversion-optimized approach performed well, as anticipated. However, a deeper dive revealed that some segments responded more favorably to RF campaigns, while conversion optimization did not yield the same results. Based on these insights, we recommended a targeted approach: select segments that performed best with conversion optimization while employing RF campaigns for segments that included a significant number of lapsed customers. This dual strategy aimed to effectively re-engage those who hadn’t purchased in a while while leveraging the strengths of conversion optimization for high-performing segments. It's worth mentioning that the client had some initial reservations about running a fully RF campaign, which is why we proposed a balanced approach that combined conversion optimization with RF strategies. This allowed us to explore the strengths of both methods. We also considered the implications of opt-in and opt-out strategies. By integrating RFM analysis with opted-in/out data and campaign objectives, we developed a variety of targeted segmentation options. This framework provided a flexible set of contact strategies, enhancing the potential to re-engage lapsed customers while effectively optimizing media spend across different segments.   The Results RBX Active experienced a successful Black Friday and Cyber Monday (BFCM) by implementing a targeted approach. As a result, there was a 10% increase in revenue per customer ($/customer) from the lapsed segment. Overall, the brand achieved a remarkable turnaround during this period, realizing a 16% increase in total revenue, effectively reversing the previous trend of -12%. This success can be attributed to a well-crafted contact strategy tailored to micro-segments. By leveraging insights from the RFM analysis and strategically optimizing media spending, RBX Active not only re-engaged lapsed customers but also maximized overall revenue. This outcome underscores the effectiveness of a data-driven, segmented approach to customer retention and engagement.      

Getting to the Root of Media Effectiveness: An Advanced Attribution Case Study with Jefferson Dental & Orthodontics

The Brand Jefferson Dental & Orthodontics (JDO) is a large regional dental services organization with more than 60 offices throughout Texas. The company provides high-quality, affordable, and convenient dental and orthodontic care with their advanced technology and comprehensive oral health plans for each of their patients. They call this their “Smile Roadmap”.       The Challenge A new CEO and marketing leadership came on board in 2024 and had an inclination they were over-invested in certain marketing channels. They were having a hard time understanding the true unit economics of the media spend. They didn’t entirely trust what they were reading from Google Analytics or in their platform reporting. Dental services organizations are also known for their notoriously incompatible systems. Patient CRM, patient billing, marketing, and everything else tend to be siloed leading to data fragmentation and debilitating opaqueness for leaders managing these businesses. The JDO team was also in the process of onboarding a new agency partner and opined that it would be great to have new insights into media’s efficiency to get oriented on a new go-to-market strategy for the brand.     Exploratory Data Analysis As part of the exploratory analysis, marketing measurement teams have to frame outcomes that marketing can be reasonably tasked to drive, like “Appointments Booked”, but also draw a straight line to unit economics that CFOs care about, like “Appointments Completed”. The M-Squared team had several data sets to assess and synthesize. These datasets included media investment and delivery data across a diverse mix of channels and tactics going back 2 years. The media channels included Google Search (brand and non-brand), Facebook, CTV, and Direct Mail. The data analyzed also included scheduled appointments and completed appointments as the dependent variables. More than 60 Jefferson Dental offices throughout Texas. “As a marketer, the dream is to be able to predict what drives consumers”- Cale Pritchett   The Approach These large and diverse datasets needed to be processed and transformed to build the Analytics plan. The analytics plan developed as part of a structured Advanced Attribution Audit was designed to address the questions inspired by the business and entailed: Data Harmonization: Collection of the historical data (1.5 years) on appointments, revenue, and media which was reviewed and processed as per modeling needs. Preliminary Analysis:  Performing trend analysis, correlations, and a basic MMM to understand the fit of individual variables driving appointments and revenue. Media Mix Modeling: Running 1000’s of iterations and 100’s of tranches for the different segments to come up with the best-fit models explaining the drivers of appointments across new and existing patient groups. Triangulation: Using the MMM decomps, did a triangulation exercise to understand the impact of media and the value it brings in driving appointment bookings and completed appointments.   The Marketing Accounting Framework (MAF) was oriented around incremental scheduled appointments, completed appointments, and production (revenue) driven by media for new and existing patient cohorts. This was distilled into incremental ROAS (iROAS) by media channel/platform. This was a 4 P&L structure - meaning 4 different models were built - one for each of the appointment and customer types. Learn about Marketing Accounting Frameworks with this short course from M-Squared:     The Results Media drives incremental impact/revenue - nearly 5.8% of booked appointments on an incremental basis with a ROAS of almost 1.60 and 4.6% of completed appointments on an incremental basis with a ROAS of almost 1.50. GA and ad platform reporting overstate media contribution by many multiples. The JDO team’s hunch was right that they were probably overspending - in part based on platform and GA reporting not accurately capturing impact. Media contributes to both new and existing patient appointment booking - which is important in understanding the holistic value of media - and how it contributes to LTV.   Overall, the insights suggested there was a growth opportunity within the existing budget based on making some investment shifts to higher-performing (based on incremental appointments) media tactics. There also appeared to be an opportunity to tighten and validate some of the model insights through incrementality testing geospatial analysis.      The Conclusion The model outputs created a set of initial recommendations, and validated what the JDO marketing team’s “gut was telling them” but also raised a number of hypotheses that would need to be tested. These tests will serve to validate and/or further refine the models and in the shorter term can be used to update iROAS numbers to ensure a high degree of confidence before looking to scale many of the recommended investment shifts. As Jefferson Dental continues down the path of test-learn-grow M-Squared will continue to fine-tune the models to improve their fit and outputs. Along the way, the team will continue to find questions, form hypotheses, and test them in the market. Growth is always an ongoing and iterative process.  

Managing profitable growth for a portfolio of brands: How Garden’s Alive is governing their media mix with advanced attribution

The Multi-Brand Marketing Problem Statement For businesses operating in highly seasonal markets, strategic media planning is both an opportunity and a challenge. Gurney’s is a leading name in gardening and home solutions under the Gardens Alive! umbrella. Historically, Garden’s Alive has relied on Catalog to drive new customer growth and to engage existing customers. But the productivity of Catalog has waned over the years, as it has for many direct marketing brands around the country. With shifting consumer behaviors and the rise of digital channels, marketing leader  J.P Kinerk faced a crucial question, What is the most effective media mix to drive profitable growth while improving the unit economics for each of the brands?   Exploratory Data Analysis The existential measurement challenge for Garden’s Alive was that most of the brands were horticultural gardening brands. Therefore consumer demand for their products was acutely seasonal, and naturally marketing investments rose and fell with the seasons as well. This manifested itself as the big colinearity problem that measurement practitioners dread. It’s nearly impossible to tease out from historical data how much of the demand was a function of the marketing versus the natural seasonality of the business. Initial exploratory data analysis (EDA) revealed many such instances. An archetypal example was an initial correlation study that revealed a very strong relationship between existing customer catalog drops and new customer acquisition growth. Further critical reasoning established that the insight was nonsensical since existing customer catalogs are unlikely to have had such a large impact and the seasonal impact was confounding the analysis. Nonetheless, the leadership required a deeper understanding of the incremental performance of their marketing investments and a clearer read on key business KPIs like incremental customer acquisition cost, contribution margins, and LTV for each of their brands to bring fiduciary discipline to their marketing investments. Even more critically, there was a need to align these metrics across finance and marketing teams to create a unified marketing accounting framework.   Approach: A nuanced marketing measurement solution to solve for constraints faced by seasonal brands To address these challenges, JP and the team at Garden’s Alive embarked on a comprehensive marketing measurement initiative. They invested their energies in setting up an advanced attribution program across all their brands to bring academic discipline into their marketing investments.   Some of the key business questions they laid out for the program to address were: Can we meaningfully replace catalog investments with digital marketing investments?  Can we invest in upper-funnel branding to drive growth? How do we reallocate investments to drive efficient growth using causal attribution? How do we communicate marketing’s effectiveness with finance and senior stakeholders? By leveraging the geo-match market testing, marketing mix modeling, and triangulation within the advanced attribution framework, Garden’s Alive set measurements for each of its brands to address the business questions laid out while handling the various seasonal and media mix nuances for each of the brands.    1. Triangulation Using Multipliers For highly seasonal brands like Breck’s Gifts and Bits & Pieces, where 80-90% of sales occur in Q4, an attribution audit was conducted to evaluate the impact of the current media mix. By combining benchmark multipliers with historical performance data, the team gained clarity on where digital investments could be maximized. To complement this, M^2 Flywheel Analysis was introduced to assess: The true cost of acquiring a new customer. Break-even scenarios are based on varying levels of media spend. Scenario planning for different media mix strategies, including revenue maximization, return on ad spend (ROAS) optimization, and cost savings.   2. Marketing Mix Modeling for New vs. Existing Customers Gurney’s had previously run incrementality tests and established certain multipliers for measuring media effectiveness. However, to ensure accuracy and adaptability, a fresh round of MMM studies was conducted—one focused on new customer acquisition and another on existing customer retention. By taking this dual approach, the business was able to: Determine how media influenced different customer segments. Derive new multipliers to refine future marketing investments. Improve budget allocations for Q1 and beyond.   3. Geo-Match Market Testing With Google phasing out VAC campaigns on YouTube, Garden’s Alive needed to evaluate the effectiveness of its new Demand Gen video campaigns. A structured 3-cell geo test was implemented, covering: Business-as-usual (BAU) markets Hold-out markets with no spend Scaled test markets with increased investment Additionally, mid-funnel KPIs, such as engaged sessions, email sign-ups, and catalog requests, were tracked alongside direct sales. On the Meta side, Facebook demonstrated over a 10% lift in sales in previous triangulation studies. However, ASC campaign limitations on geo-targeting necessitated a 2-cell hold-out test, designed to: Validate Facebook’s contribution to incremental lift. Fine-tune audience segmentation strategies. Determine optimal budget distribution across digital channels.     Initial results & key takeaways 1. Improved Attribution & Budget Allocation Through Triangulation, the business uncovered: • Breck’s Gifts: Media drove a 45% incremental new customer rate. However, short-term break-even required a 30-percentage point increase in media impact. With a $500K digital budget, optimal reallocation could achieve a ROAS of 2.12.• Bits & Pieces: Media drove a 57% incremental new customer rate. Catalog investments, though effective, presented an opportunity for cost reduction. A more efficient digital strategy could yield a ROAS of 4.2 with a $500K spend.   2. Incrementality Testing Insights Garden’s Alive validated the following findings through Incrementality Testing: Facebook: Delivered a 13% incremental lift with a ROAS of 5.2, justifying increased investments. YouTube Demand Gen Video Ads: Showed no significant short-term impact on sales, yielding a ROAS of only 0.25. Email Sign-Ups: Increased by 12%, demonstrating the potential for first-party data capture.   3. A Roadmap for 2025 & Beyond With these insights and others, Garden’s Alive was able to establish a medium-term investment and media mix thesis to Diversify its media mix for sustained growth. Implement more robust attribution models for decision-making. Maximize seasonal revenue opportunities while reducing inefficiencies.   Final thoughts: The  power of advanced attribution JP and the team are putting in place a robust advanced attribution program to govern their marketing investments and right-size their marketing investments to deliver profitable growth. By embracing a nuanced marketing measurement framework, they have set the foundation for a scalable, data-driven future.   “Finally an attribution model that’s on OUR side, very glad we found M-Squared. If I only had one shot at making it work, M-Squared would be in my corner.”- JP Kinerk  

Geo Match Market Testing
Marketing Mix Modeling - A Modern Case Study: Beyond Meat

The Brand Beyond Meat is a leading plant-based meat company founded in 2009. The company’s products are available in over 133,000 retail and food service outlets across 65 countries. Beyond Meat went public in 2019, becoming the first plant-based meat alternative company to be listed on a U.S. stock exchange. Beyond Meat continues to focus on expanding its product range, improving its existing offerings, and increasing its global market presence in the growing plant-based protein industry.   The Challenge Like many CPG brands with brick-and-mortar retail distribution, Beyond Meat was looking for a more sophisticated attribution and marketing accounting framework to better understand and report how their media and marketing investments were driving sales. In addition to traditional price promotions, the company also had a diverse media mix that included both online and offline channels. Beyond Meat, like many others, had also been increasing its investment in the burgeoning Retail Media Network (RMN) ecosystem - though it was not entirely clear how RMN’s were helping drive sales. We wanted to verify some of the outcomes from another MMM provider. The question always begged was “What is actually working?”.   Exploratory Data Analysis The M-Squared team complex data sets to assess and synthesize. These datasets included media investment and delivery data across a broad and diverse mix of media channels and tactics. The media channels included TV, social, audio, display, video, and RMN’s. The data analyzed also included retail sales information from a variety of retail outlets ranging from Big Box (Costco) and traditional grocery (Walmart) to specialty natural and health food retailers. These large and diverse datasets needed to be processed and transformed to build the Analytics plan.   The Approach The analytics plan was developed to address the questions inspired by the challenges entailed: Data Harmonization: Collection of the historical data (2 years) on sales, media, and events which was reviewed and processed as per modeling needs. Preliminary Analysis: Performing trend analysis, correlations, and a basic MMM to understand the fit of individual variables driving sales. The M-Squared team further reviewed the retail store categories and created a hypothesis to determine the number of MMMs to run. Media Mix Modeling: Running 1000’s of iterations and 100’s of tranches for different retailer groupings to come up with the best-fit models explaining the drivers of sales across what ended up being 4 primary retail sales channels. Three of them were stand-alone, major chains and one was an aggregate of specialty retailers. Triangulation: Using the MMM decomps, did a triangulation exercise to understand the impact of media and the value it brings in driving retail sales. The Marketing Accounting Framework (MAF) was oriented around incremental retail sales volume ($$$) driven by media. This was distilled into incremental ROAS (iROAS) by media channel/platform. This was a 4 P&L structure - meaning 4 different models were built - one for each of the 4 retail store groupings. This was the best approach from a model feasibility perspective (grouping smaller speciality retailers) and the business understanding that the other 3 retailers were different/unique enough to warrant their own model and “P&L”.   The Results Media drives incremental impact/revenue - nearly  10% of sales on an incremental basis with a ROAS of almost $4.00. Different types of media impact sales differently depending on the sales channel and/or retail network specifically. Retail Media Network (RMN) performance varied quite a bit by network -            RMN A had a ROAS of 3.9           RMN B had a ROAS of 2.9 and           RMN C had a ROAS of 0.3 Upper-funnel and lower-funnel media tactics performed better than mid-funnel tactics. Upper funnel tactics had an average ROAS of 4.85 while lower funnel tactics had an average ROAS of 2.6. Overall, the insights suggested there was a growth opportunity within the existing budget based on making some investment shifts to higher-performing (based on incremental sales contribution) media tactics. There also appeared to be an opportunity to drive additional growth through increasing investment levels.    The Conclusion Using MMM reads from the initial round of modeling and forecasting based on nearly doubling the media budget there is the potential for 26% growth in retail sales driven by media. The model outputs created a set of initial recommendations, but also a number of hypotheses that would need to be tested. These tests will serve to validate and/or further refine the models and in the shorter term can be used to update iROAS numbers to ensure a high degree of confidence before looking to scale many of the recommended investment shifts. As Beyond Meat continues down the path of testing, learning, and growing, M-Squared will continue to fine-tune the models to improve their fit and outputs. Along the way, the team will continue to find questions, form hypotheses, and test them in the market. It’s important to understand that practicing data-driven decision-making is an ongoing iterative process. At no point does it become “set it and forget it” because the media, consumer, and business dynamics are constantly evolving. As long as this is the case we need to think of Advanced Attribution techniques like Marketing Mix Modeling to be evolutionary and never static.    

Data-Driven Marketing Marketing Mix Modeling Marketing Mix Modeling Best Practices ROAS
How effective is your investment in Facebook really, and how can you  optimize spend?

The Brand CustomTees is a successful online clothing brand, with annual revenue exceeding $50M. Despite their success, the company's marketing team grappled with one  critical question: How effective was their substantial investment in Facebook? Was there an opportunity to optimize Facebook spend to increase profitability while not impacting revenue?   The Challenge The brand had historically relied heavily on Facebook for growth, allocating north of 40% of their marketing budget to the platform. With quarterly Facebook spend reaching $600,000, the stakes were high. Facebook's prospecting platform reported an 2.8 ROAS (Return on Ad Spend) on a 7-day click one-day view basis, but the client’s marketing team suspected this might not tell the whole story despite the previous success of the channel.    Digging Deeper The initial analysis that was done, using a 7-day click only attribution benchmark which painted a very different picture. This method, which applies a discount to platform-reported results, suggested the brand was merely breaking even on Facebook ads. The UTM attribution calculated off of URL parameters reported an even lower ROAS. If this were true, it would imply the channel isn't profitable after considering costs of goods sold, shipping, etc. Moreover, it could be a sign that Facebook needs to be optimized and is cooling off despite its earlier success.  The brand decided to take up incrementality testing with M-Squared to uncover the ground truth behind Facebook performance.   The Approach:Incrementality & Platform Lift Testing To uncover the truth, M-Squared recommended two measurement approaches to test Facebook Prospecting:  Geo Incrementality Test Facebook Platform Lift Study Both tests were conducted over 30 days, with the geo test designed as a multi-cell experiment to assess incrementality across various channels, including Facebook Prospecting. Catch a quick primer on different types of incrementality tests from the M-Squared masterclass here:   The Results: The results were eye-opening: The geo incrementality test showed a total lift of about 10%, primarily driven by new orders. Calculating the multiplier revealed a channel multiplier of 41%. When applied, this showed that the ROAS on Facebook prospecting was actually around 0.6 - significantly lower than initially reported. The UTM attribution reported ROAS seemed to closely approximate the incremental reads coming out of the tests. Interestingly, the Facebook platform lift study corroborated these findings, showing a multiplier of 41%. Triangulating reads across multiple studies is a common best practice that marketers are adopting en masse. Catch a quick primer on this best practice from M-Squared masterclass here: Implications and Next Steps Armed with this new data outside of platform-reported results, these consistent incrementality results provided the marketing team with a much clearer and concise direction. The data strongly suggested that cutting the Facebook budget and re-allocating resources elsewhere could be a logical next step for optimizing overall marketing performance and driving increased profitability, without harming topline revenue. M-Squared recommended testing budget re-allocation before rolling it out nationally by cutting spend on FB and re-allocating to other channels.  The Conclusion: In the world of data-driven marketing, having multiple consistent data points is crucial for avoiding false positives and making informed strategic choices. For this e-commerce clothing brand, the journey of advanced attribution not only revealed the true performance of their Facebook advertising but also opened up the discussion for new growth opportunities. By reallocating budget from underperforming channels, they can now explore other channels to drive more efficient growth and profitability. This case study serves as a powerful reminder of the importance of rigorous, multi-faceted attribution in modern marketing. In an era where data drives decisions, looking beyond surface-level metrics can uncover hidden truths and unlock new paths to success.  

Geo Incrementality Test Incrementality testing
Incrementality testing unlocks upper funnel branding investments for Study.com

The Brand Study.com provides high stakes online learning solutions to more than 34 million learners and educators a month across professional test preparation, college credit and K-12 education. Recognized as one of the world’s most innovative companies by Fast Company and the GSV150, Study.com has helped save students more than $475 million in tuition costs through its College Saver program and donated some $29 million across social impact programs committed to increasing educational equity The Challenge The company's new customer acquisition efforts had been primarily centered around SEO and PPC investments, and managed like a classic performance marketing program, governed to tight CAC guardrails agreed upon between marketing and finance. The strategy served the organization well for many years helping the brand drive predictable and profitable growth, but eventually these channels reached maturation and could not continue the accelerated growth rate desired. So like many other brands in the same predicament, Study.com has been seeking diversification and growth opportunities in upper funnel channels.  As Study.com explored other channels it also aspired to build out a stronger brand, investing in video creative with a branding narrative. The brand campaign-focused on inspiring students and educators to reach their academic and career goals through Study.com's online college courses, exam preparation and classroom resources. But as always the key question is “Will branding drive new customer demand?”.  When the company rolled out smaller scale paid video campaigns on platforms like YouTube, it saw few last-click conversion although other positive indicators led to the desire to continue investment at scale and with more precise measurement. There were several back-and-forth discussions to determine if the organization should put more budget behind this and to launch in CTV, but would it be a risk worth taking? Would this bring in new students?  How can you measure it since it serves more as an upper funnel tactic? To find out, Study.com embraced geo-based incrementality testing with M-Squared to learn the efficacy of branding campaigns on CTV and YouTube.   “Really just taking the test-and-learn approach and leaning into creativity”- Emily Johnson Watch as Emily Johnson unpacks her experience with us! Exploratory Data Analysis As with any brand, the devil is in the details. Real businesses have real complexity in their business model, and in their data. Marketing measurement practitioners have to process these complexities and assess which of those nuances are meaningful to incorporate into the measurement solution architecture and which ones are irrelevant to the business questions being considered. For Study.com, there were many such dimensions of complexity:  Like many businesses, that service multiple audiences with several products, there is a mix of hero and long tail products.  Different products drive unique LTV and hence different levels of contribution margin for the business over the long term. The tactics being tested are demand generation (vs demand harvesting) and the video medium, which intuitively means the measurement has to account for a longer time-to-conversion period, and low last click attribution. Should the measurement plan consider upper funnel outcomes like Engaged Sessions or Email/Phone collection which could serve as leading indicators of demand being generated. Would the data collected so far support that measurement plan? The Approach:Geo Match Market Test - Design At M-Squared we take a structured approach to designing a geo test.    Catch the 11 minute snippet on geo testing from the M-Squared masterclass:  Study.com already had evergreen YouTube campaigns in flight, and there had been a significant push over the summer months to launch a new branding video.  Post geo testing feasibility analysis, we determined that the current spend levels on YouTube might not clear minimum detectable lift (MDL) levels and hence may not yield statistically significant reads with a holdout test. As a result, we tested YouTube with a scale cell where spend was intentionally elevated to support readability. A holdout cell was also included to measure the lift at current spend levels, with the understanding that the reads may come back inconclusive, but provide valuable insight on incremental vs marginal contributions. CTV was a brand-new channel and had no prior spend or performance history. It was not meaningful to select a holdout treatment for incrementality testing. Instead, a scale cell was designed as part of the feasibility analysis. Since it was a new channel, warming up the channel was recommended before starting measurement. So the first 2 weeks were slotted for campaign warmup and the next 4 weeks for the read, running a total of 6 weeks in test flight. Market selection algorithms were run and DMA’s were identified for the three different testing cells. Test budgets and test flight period were determined as part of feasibility analysis.  YouTube Scale cell  YouTube Holdout cell CTV Scale cell Markets: 14 DMAs Flight: 6 weeks Budget: $120K Markets: 14 DMAs Flight: 6 weeks Budget: No spend in selected markets during test period Markets: 13 DMAs Flight: 6 weeks Budget: $120K The test was flighted in Q4 2024, and the flight was monitored for execution aligning to the test design that was put in place.  The Results:Lift Reads & Interpretations for Growth As with the design, at M-Squared, we take a structured process for estimating the lift reads from the test.  Catch the 7 minute mini course on estimating lift from the M-Squared masterclass:  After carefully estimating the lift with multiple algorithmic approaches, Study.com learned: CTV drove a 6%  incremental lift to new member acquisition at a 3.8 ROAS YouTube showed good potential as well 3% incremental lift on  new member acquisition at a 2.5 ROAS Both of these reads provided meaningful insights on employing upper funnel tactics for driving customer acquisition growth for Study.com.  With reasonable assumptions on scale and diminishing returns, and annualizing the estimates for seasonality, employing YouTube and CTV in their customer acquisition plan could drive an estimated 20% growth for the brand. The Conclusion: These insights now provided Study.com a clear path to diversify its media mix into upper funnel channels using a test-learn-grow approach for risk mitigated and fiscally responsible growth. Disclaimer: The data presented is blinded to protect brand’s P&L confidentiality but preserve insights for educational purposes.

Geo-Testing Strategies Incrementality testing
Marketing Mix Modeling: An Origin Story

Marketing Mix Modeling (MMM) has been used to measure the impact of marketing and advertising for around 40 years. While the exact origins are difficult to pinpoint, MMM emerged in the 1980s as a way to analyze the effectiveness of different marketing activities on sales where there was no direct or deterministic way of doing so. Early adopters were primarily consumer packaged goods (CPG) companies who had the necessary data on sales and marketing spend and had the challenge of tracking sales dispersed across various physical retail channels. Legend has it that Coke was among the very first brands to use MMM in the 80’s. See the masterclass interaction with William (Todd) Kirk, one of industry’s OG MMM scientists discussing history of MMMs.   Here's a brief timeline: 1960s The foundation for MMM was laid with the development of econometric models.  1980s MMM gained traction as computing power increased and more companies began collecting detailed data on their marketing activities. 1990s - 2000s MMM became more sophisticated with advancements in statistical techniques and software. 2010s - Present MMM continues to evolve, incorporating new data sources (like digital advertising data) and addressing challenges like attribution in a multi-media, multi-channel world.   It's important to note that MMM is not a static concept - it is in a constant state of iteration. To be effective on an ongoing basis it needs to continuously adapt to changes in the marketing landscape, incorporating new technologies, media channels and data sources to provide more accurate and granular insights.

Data-Driven Marketing Marketing Mix Modeling Marketing Trends MMM History
Unlocking growth for a cosmetic brand

When diving into the data of any brand, there are so many factors up for consideration. As we already know - advanced attribution is not a one size fits all game and nor should it be - all data is different and all company’s needs and targets are too.  Getting to the meat of any brand - we first need to get to know them better and jump in with two feet in order to deeply understand the value of the business. When we first began our engagement with a well known cosmetics brand, the team was trying to answer a few simple questions, like ‘What is the ROI from our media investments?’ or - more pointedly, ‘How much budget should be allocated to the Top of Funnel?’. The last question, and probably one of the most important ones - is, ‘What is the Contribution Margin and Revenue per customer?’  All valid angles to approach and all important for making and the next move.  Understanding the true value Let’s begin - In order to take the first step, we need to understand the true value being driven by marketing. Our first step in the process was to understand the Contribution Margin from an observed media perspective and an advanced attribution standpoint. For us to calculate the contribution margin, the team worked closely with the brand’s marketing team to gather their underlying factors, such as Cost of Goods Sold (COGs), Promotional Spend, and Shipping Cost. Once we understand the client's contribution, we can begin the analysis of Revenue per customer.  To calculate this, we divide the newly discovered contribution margin by the total number of customers as shown in the graphic below:     Analysis of media spend  Diving even further - once we understood the contribution margin, we wanted to look at the entirety of media spend to understand the impact seen across their media portfolio. Since the client had no custom attribution methods, we used platform-driven attribution as our anchor. In order to understand the advanced attribution of their media portfolio we applied M-Squared’s  multipliers to estimate the true impact of their media.  Through this process we were able to gather some impactful insights; such as their Meta campaigns are drastically underperforming as compared to the industry average. Another insight would be that their Google Shopping campaigns are the most impactful to their overall bottom line, and we should continue to fund that platform. The final insight that caught our attention was that affiliate marketing is one of the strongest driving factors within their overall media portfolio. Have a look at the graphic below highlighting the lowest and highest returns:   Test and Growth Plan Now that we had some hard facts to play with, we could start testing different marketing routes and develop a sustainable growth plan. From our analysis of their media performance, we can instill what’s called a ‘test and grow plan’. This specialized report calls for shifts within the company to go bigger - such as budget reallocation to more robust performing media channels, conducting measurement experiments to better understand diminishing returns within specific platforms that are not performing the way we wanted them to and why.  Some examples of the recommendations would be to run a Geo Scale test within Meta and some of their display partners in order to ascertain the scaling opportunities within the market. We also recommended a Pulse Test for their affiliate program to better correlate the impact of sales periods and the affiliate program itself. In the next graphic you can see that through our analysis, we estimated that we can grow revenue by 10%, all the while cutting the budget by 80k!

advanced attribution Cosmetic Brand Marketing Media Spend Analysis
YOU CAN ALSO CATCH US HERE
GET IN TOUCH WITH US!

Whatever questions you may have or topics you want to cover, we would love to hear from you!

Fill out the form below or make a booking with us.

Thank you for submitting your form!
We are so excited to learn more about your attribution needs and will get back to you within 24 hours.