It’s A Marathon Not a Sprint: Following the Data

||||New Program - CPL Over Time|CPL Over Time by Individual Program

Wonder what business results and insights you could uncover with a stronger POV on data-driven marketing? Let’s chat and explore how data can boost your marketing performance.

Congratulations, your creative is approved, media buy is scheduled, and marketing analytics and tracking is in place. Your new campaign is officially live. So you can expect campaign insights and analysis in week one, right?

While there’s a strong temptation to review data quickly, there’s several reasons why jumping to analysis can negatively impact the long-term effectiveness of your marketing. Data-driven digital marketing follows a process to identify, gather and analyze data over time. This process begins with planning, continues through each stage of campaign execution, and ultimately becomes a feedback loop to iterate on live campaigns and inform future campaigns.

Planning for Success: Expectation Setting Before Launch

When thinking about analytics and data-driven marketing, it’s important to remember analytics is not a one-time input or output. During planning, data-driven marketing works in tandem with other business inputs to inform campaign goals, strategies and tactics.

As the expression goes, “Those who cannot remember the past are condemned to repeat it.” Here’s where analytics excels in the planning process—reviewing past performance to improve future marketing ROI.

Effective marketers evaluate past campaigns and ask questions like: 

  • Which channels drove the most leads? The lowest cost per lead? 
  • Which tactics supported top of the funnel vs. bottom? 
  • How did targets respond to different messages? Different creative?

By analyzing past performance and tracking trends in data, businesses can set realistic benchmarks for marketing KPIs and metrics for future campaigns based on real-world results. Plus, understanding how potential customers reacted to different strategies in the past can inform future tactics.

Data in Action: With demand for online education growing, our client the University of Cincinnati Online often engages us to market new programs. Over the past year and a half, we’ve launched campaigns for 14 different programs with the goal of driving relevant leads. But individual program performance hasn’t been the only priority. Understanding the bigger picture that UC Online will continue to add new programs,, we set in place an analytics plan that would gather and leverage months of performance data from all new program campaigns. This process would help us to 1) set benchmarks for new program KPIs and 2) understand how their lead generation progresses over time.

Evaluating Data on Campaign Performance

Once a campaign is live, data flows in like a firehose. With a high volume of information, marketers need to understand what’s valid, what’s actionable and what’s representative.

That’s why we identified 5 key metrics to guide brands in evaluating data:

  1. Accuracy: Ensure that the data is as close to real-world results as possible. A few ways to check accuracy include reviewing data in context of sample size, and comparing against past performance and industry benchmarks. 
  2. Relevance: Focus on data points that are most important to the campaign’s goals. While many things can be measured, the KPIs you aligned on in the planning process should guide you in tracking your progress to goal—whether that is awareness, engagement or conversion.
  3. Time frame: Look at both short-term and long-term trends in order to identify meaningful patterns and understand their impacts. The ability to zoom in and zoom out of data protects you from knee-jerk reactions and can help you make more informed decisions on where, when and how to optimize your campaigns.
  4. Consistency: Analyze and test the quality of data over time, in order to ensure accuracy and reliability in all analysis performed with this data set. Teams that have established practices for data hygiene are more credible and trusted with their reporting.
  5. Interpretability: Make sure that the data can be understood by various stakeholders in a clear, concise fashion. Before presenting campaign results, consider what information is most relevant and actionable for different audiences. A CEO or executive team will have a different lens than Customer Service or Sales. Visualizations can be a helpful tool to share insights into key trends and patterns in customer behavior, and opportunities for the broader organization to engage them.

Data in Action: Working with University of Cincinnati Online we gathered data from the first 11 months for all new programs to answer the question: on average, how long does a new program take to ramp up and reach an optimal Cost per Lead? By plotting the combined program data over time, we were able to visualize insights in a more compelling way than a spreadsheet and see where we should dig in to the data further. In this case, we wanted to explore more what was driving the sharper peaks and valleys.

 

New Program - CPL Over Time

 

When we visualized the data by department over time, we now could easily see where specific points were not in line with the overarching trends. Identifying and evaluating these outliers sparked discussions about why they may have occurred, as well as solutions for future campaigns.

 

CPL Over Time by Individual Program

 

Best Practices for Applying Data to Decision-Making

Data-driven digital marketing requires critical thinking at each stage. This mindset is especially crucial when it comes to making decisions based on campaign performance.

In working on hundreds of digital campaigns, we’ve identified a few common mistakes to avoid when using data for decision-making:

  • Relying on incomplete or inaccurate data. Regularly reviewing and optimizing campaigns is a key benefit to digital marketing. But if the data you are collecting is not complete and accurate, it can skew the results of your analysis. Before the campaign starts, set expectations for total conversions and other KPIs so everyone is aligned on what volume of data is needed before you make decisions.
  • Not accounting for biases in data collection and analysis. Biased data can lead to false assumptions and incorrect conclusions. While AI and predictive algorithms are tools for marketing analysis, they also can be potential sources of bias when gathering and analyzing data.
  • Assuming correlation implies causation. Every professor’s favorite mantra in statistics and research methodology is true. Just because two variables appear to be related does not necessarily mean that one causes the other. Consider what other explanations could account for the trends you see.

 

Building a Data-Driven Story

From planning to execution to decision-making, data offers multiple benefits to make your marketing work harder and smarter when you use it over time. When there is a process that guides how data is identified, gathered and analyzed, you can achieve business objectives and spot new opportunities.

Like every great story, a data-driven story follows a simple framework:

  • Context: Set a clear purpose and goals during planning to understand what you are trying to achieve and how you will measure success.
  • Data: While data flows in from day one, building a data story takes time to ensure inputs are accurate, consistent and relevant. 
  • Visualization: When you have reached statistical significance and can show what you learned in a simple yet compelling format, raw numbers become actionable insights and talking points that can be used throughout your organization.
  • Narrative: With data as a throughline, you can guide your audience through the story from planning (what you were trying to achieve) to performance (what happened, shown in visualizations and charts) to analysis (what you learned from the data and will apply moving forward).

For the University of Cincinnati Online, its data-driven story was understanding how digital marketing connected with its customer journey. Analyzing data across new program campaigns helped us identify 3 phases in digital media: learning phase (months 1-4), optimization (month 5-8) and peak efficiency (months 9-onward). Now when planning campaigns, we can focus our spend and our tactics to each unique phase of the digital campaign—and ensure consistency with other marketing channels that potential customers experience in that same time period.

Wonder what business results and insights you could uncover with a stronger POV on data-driven marketing? Let’s chat and explore how data can boost your marketing performance.