Case Study: Data Gathering and Implementation
This case study highlights the successful execution of a data-driven marketing strategy, focusing on the objectives, challenges faced, solutions implemented, and the remarkable results achieved. The innovative use of data collection, analysis and predictive modeling played a pivotal role in guiding strategic decision-making and driving marketing success for the client.
OBJECTIVE: TO CREATE A DATA-DRIVEN STRATEGY AND METHODOLOGY FOR SMART AND RAPID OPTIMIZATION OF CLIENT’S DIGITAL MEDIA CAMPAIGN.
1.1 Background: A news publishing company sought to increase market share and brand visibility for their flagship product in a highly competitive market. 1.2 Client Goals:
- Meet or exceed yearly acquisition goals (subscriber volume and CPA)
- Enhance brand recognition and increase market share
- Improve customer engagement and loyalty
- Optimize marketing campaigns through data-driven decision-making and predictive intelligence
- Maximize ROAS by targeting the most relevant customer segments
Section 2: Challenges
2.1 Unclear Measurement Goals There was a blanket yearly Cost per Acquisition Goal that was established based on internal revenue models that the client had to hit in terms of keeping the business profitable. There was no further delineation in terms of how that relates to the longevity of audience retention, to the nuances of marketing channels and fluctuations of performance therein, and to various aspects of the sales funnel like visits, article viewership and seasonal news cycles. 2.2 “Always on” Multiple Marketing Channels & single DMA Due to the publishing firm being a popular brand and a product that is relevant year-round, the marketing mix needs to be comprehensive, to stay top of mind for the consumer. This meant that the media buy stayed active across multiple digital channels, all year round, limiting opportunities to test individual channels separately and gain insights that are channel specific. This point needs to be highlighted because the need for sophisticated data analytics arises if there is a problem that can’t be solved by controlling external parameters. Without the “always on” and “single DMA” constraints, we could run a test and know whether one channel is profitable or not, run longitudinal tests and make scientific observations for the same. But we did not and do not have those options in this case. Additionally, even if we could run those tests to determine individual channel profitability, we still wouldn’t be able to predict what would happen if they were in market on all channels together. The constraints disallowed us to easily see how channels interact with each other. 2.3 Data Fragmentation: The client faced challenges due to fragmented customer data across various internal systems, making it difficult to gain a comprehensive view of the overall impact of KPIs on yearly measurement goals. The one single goal, mentioned in 2.1, which is a yearly acquisition goal, translated into yearly cost per acquisition but KPIs collected in a marketing sales funnel are numerous – Impressions, Clicks, Visits, Google Trends Index and platform-attributed acquisitions to client reported growth in acquisitions. There was no clear insight into how these different KPIs were related to each other. Qualitatively they did know the sales funnel but pragmatically they did not know how each of the KPIs affected the bottom line.
Section 3: Solutions
3.1 KPI Delineation & Consolidation To reduce the clutter in understanding created by the data fragmentation, we created and had the client agree on a system of KPIs that included important “signals” such as Google Trends data; audience survey data, via a cadence of weekly tracking via Harris Poll, to measure branded impact; and a few others. These KPIs were then consolidated into a singular consumer journey that showed how certain tactics affected awareness and consideration engagement (as reflected in, for instance, Harris Poll data) as well as the final conversion sales KPI. 3.2 Predictive Mixed Media Modelling Hawthorne Advertising created a predictive performance model that looked at individual digital-platform-reported cost per acquisition (CPA) performance, enabling us to project sales for the remainder of the year with a high degree of accuracy, and unified the platform-level insights into a singular CPA metric. This approach gave us the ability to link between the yearly Sales Goals KPIs and individual digital-platform-reported CPAs. Thus, the modelling enabled us to predict how varying spend levels amongst the digital platforms would affect sales in those platforms as well as overall sales. 3.3 Consolidation and Integration: To overcome the challenge of fragmented data, a data integration platform was implemented, facilitating the consolidation of customer journey data from various sources, enabling a single, unified view of each customer. 3.4 Automation and Optimization: Marketing automation tools were leveraged to streamline campaign execution and monitor performance. Automated workflows allowed for timely and relevant communication with customers, optimizing engagement and conversion rates.
Section 4: Results
The implementation of the data-driven rapid optimization system and the high confidence delivered by the predictive performance model led to a significant increase in market share, with a YoY subscription growth of 8% in a fiercely competitive vertical.
Conclusion:
This case study demonstrates the significant impact of predictive performance intelligence on achieving business objectives. By overcoming the news publishing company’s unique set of challenges via implementation of highly actionable and transformative predictive modeling, followed by data integration, automation and optimization, Hawthorne was able to drive impressive results.
brand success? Look to the hawthorne team
