It’s an interesting time to be a marketer in the world of offline attribution. Reporting that was once limited to end-of-campaign validation has been transformed by better data collection techniques and artificial intelligence (AI) modeling. By embracing up-to-the day campaign monitoring and the concepts of predictive and prescriptive analytics, offline attribution is now an actionable science, complete with optimization recommendations to increase return on ad spend.
Through descriptive, predictive, and prescriptive analytics, you can now think about campaigns in new ways, allowing you to do the following:
- Anticipate outcomes by evaluating cross-channel KPIs on a daily basis
- Optimize media in real time to improve results before it’s too late
- Decrease media costs while getting better results (aka improve ROAS)
Let’s take a deeper look at how each type of analytics works and how marketers can leverage the insights to strengthen their strategy.
Descriptive Analytics: What Has Happened?
Descriptive analytics looks at data to answer, “What has happened?” So for attribution, it means monitoring a campaign’s delivery metrics and quantifying campaign performance in driving consumers to store.
Consumer behavior changes each day, so when measuring attribution, it’s important that you use data that was collected daily for real-time results. Also, as more and more marketers run integrated campaigns, you’ll need a solution that looks at each channel individually while also measuring the effectiveness of cross-channel buys.
In Cuebiq’s platform, Clara, descriptive metrics include Uplift, Visit Rate, Projected Visits, and Cost per Visit to answer questions such as:
- Is my campaign driving to store?
- How many users went to store?
- Which channels are most effective?
- What did it cost to drive a visit?
For example, if an automotive group runs a regional campaign to support a weekend sale, they can use a simple uplift report to measure whether the marketed dealerships saw an increase in traffic week-over-week and whether their increase was higher than that of regions without an active campaign. Furthermore, Cuebiq can report on which dealership saw the greatest traffic per DMA, suggesting the greatest responsiveness to the campaign long before sales data is available.
Predictive Analytics: What Could Happen?
Predictive analytics uses data, statistical algorithms, and machine learning techniques to identify the likelihood of future outcomes based on historical data.
As incrementality becomes a priority for marketers, predictive analytics has been getting more buzz as a way to determine if and how much your campaign is changing consumer behaviors.
Incrementality using AI is generally calculated by first looking at a user’s past behavior to predict future behavior in the absence of exposure to advertising. The prediction also considers a control group to take into account variations like seasonality. The next step is to measure the actual visits that took place after campaign exposure. The difference between the predicted and the observed visits is the number of incremental visits. Performing these calculations at the consumer level is the best way for a marketer to get actionable insights.
Predictive analytics is the mechanism by which Cuebiq reveals Incremental Visits, Cost per Incremental Visit (CPIV), and Customer Acquisition Cost. Leverage these metrics to answer the following questions:
- Is my campaign changing behaviors?
- What is the incremental impact for each consumer/segment?
- What did it cost to drive additional visits?
- Do I spend more attracting new vs. returning customers?
For example, if a quick-service restaurant brand has just introduced a coffee menu, their Brand Managers would want to know how well the menu launch campaign is performing. Is it driving visits among new customers? Or are loyal consumers visiting more frequently? Furthermore, what is it costing them to attract each new visitor? Marketers can leverage consumer-level incrementality to quantify these questions and more.
Prescriptive Analytics: What Should We Do?
Prescriptive analytics is all about providing advice using AI-based models to predict the possible outcomes of various courses of action, then scientifically determining which solution will likely yield the best results. By quantifying the decision-making process, marketers can trust they’re making the best choice before putting their precious budget behind it.
Prescriptive analytics is versatile and can be applied to all levels of media planning — from your targeting strategies to the channels and publishers you select, even down to the creative types you use. Likewise, recommendations can be used for in-flight optimization or to inform future campaign strategy, depending on when a marketer reviews results and how flexible their media buys are. There are multiple “levers to pull” when optimizing your strategy, including mid-flight adjustments, so it’s important to clarify your campaign goals (and limitations) from the outset.
In Clara, prescriptive analytics is the foundation of two tools. The first is Budget Allocation, which makes recommendations for the optimal media mix. Second is Behavioral Effect, which looks at the impact your campaign is having on consumer behavior to validate targeting strategies or recommend audience segments that will yield a higher impact for your marketing dollars. Specifically, these tools can help you answer the below questions:
- How do I maximize ROAS?
- Which channel combination works best?
- Who should I target for my next campaign?
For example, a Retail Brand Marketer evaluating a digital + linear TV campaign wants to check the cross-channel effect on store traffic as well as each channel’s individual impact. Budget Allocation does the hard work of comparing spend to performance and making recommendations for how to reallocate budget to the channels and even creative types driving the most in-store activity.
Cuebiq believes that the best type of campaign reporting is actionable and delivers insights that you can use. Read more about how artificial intelligence fuels the most sophisticated measurement in the advertising ecosystem through predictive and prescriptive analytics.