Digital marketers are drowning in data – more data than they have time to absorb and leverage. Marketers and their analyst counterparts typically use analytics solutions to distill data into “reports” used quarterly, weekly, or via self-serve access. Reports are a mainstay of digital marketing but are merely a rear-view mirror into what has already happened.
Bridging the Chasm Between Reports and Actions
Reports often fail to bring about decisions and actions because their consumers don’t pay them enough attention. However, the failures can also be due to missing skills on the part of report consumers. They may lack the analytical skills to interpret what’s happening and/or the domain knowledge to know what actions to take. In this article, we’ll explore how prescriptive analytics bridges the chasm between reports and actions by eliminating the human effort and prerequisite skills required to reach insights and decide what actions to take.
Prescriptive analytics is an advanced form of analytics that goes beyond reporting to deliver explicit decisions and recommended actions that can greatly improve a marketing team’s efficiency and revenue outcomes. Prescriptive analytics is an evolutionary step beyond predictive analytics that answers not only the questions, “What happened?”, “Why did that happen?”, and “What will happen next?” - but also “What should I do about it?”
Reports generally rely on descriptive and possibly predictive analytics but lack explicit directives on what to do next. And reports have other limitations, starting with lack of transparency and cognitive bias. According to a June 2022 Gartner survey exploring the role of marketing analytics in decision making, “One-third of respondents reported that decision makers cherry-pick data to try to tell a story that aligns with their preconceived decision or opinion.” The Gartner survey went on to cite data integration issues, data access restrictions, and process/workflow issues as top challenges. Prescriptive analytics eliminates or mitigates all those challenges by delivering explicit, prioritized directives on what actions to take next.
Why Digital Marketing is Ideal for Prescriptive Analytics
Digital marketing encompasses a continuous cycle of goal setting, strategy, tactics, execution, and measurement across multiple teams and platforms. There are many “handoffs” involved - from managers to their teams, from paid media managers to their agencies, to the individual practitioners “living” in the ad platforms. (And let’s face it, not every client gets the agency’s superstar assigned to their account.)
In 2022, digital ad campaigns running on paid media platforms (Google Ads, Microsoft Ads, Amazon, Facebook, LinkedIn, and TikTok) consumed half a trillion dollars of advertisers’ money, with an estimated 30% or more of that wasted. According to the February 2022 edition of The CMO Survey, 57% of marketing budgets were allocated to digital marketing. Yet 30% of marketers surveyed said they experience average-to-no returns on their digital marketing investments.
Bottom line: digital marketing is costly, tedious, and too often unaccountable to business goals. The reality is, managing the ongoing execution of digital ads can be overwhelming. There are hundreds or thousands of optimization opportunities hiding deep within the different ad platforms, with conditions changing dynamically. That’s far too many for humans to process, so they simply do what they can with the time they have available … then spin a nice “report” to justify their (lackluster) results.
What Prescriptive Analytics Bring to the Party
A prescriptive analytics solution for digital marketing optimization, like Condati’s Quant Marketer, goes far beyond integrating and reporting data from disparate sources (though it does all that). It provides continuous intelligence and decision support to the digital marketing practitioner. Prescriptive analytics reveals which specific optimizations will have the highest impact on desired business outcomes, and what actions need to be taken to maximize returns.
Prescriptive analytics is not about replacing humans with automation. A misguided approach to dealing with the volume of digital marketing decisions is to leave all the decision-making to the ad platforms’ artificial intelligence (AI) algorithms. Ad platform algorithms like Google’s Smart Bidding can certainly add value, but any skilled marketer knows you can’t set them and forget them. The algorithms may be proficient at bid optimizations, but it’s humans who must do the work of improving ad copy and landing pages. And how about those ad platform recommendations? A discerning look reveals that while they may (or may not) increase awareness or conversions, they will undoubtedly increase your ad spend! Those recommendations probably weren’t designed to root out and redeploy wasted spend. 😉
Prescriptive analytics alleviates the shortage of skilled talent in digital marketing. Looking at ad platform data, it’s hard for a beginner to know which assets and individual metrics to spend time on. The challenge is exacerbated by the fact that each channel / ad platform has different (biased?) metrics and control levers. Prescriptive analytics encapsulates both logical decision sequences and ad platform expertise to guide a novice user to success and/or offload work from skilled practitioners. The following Paid Search example illustrates the analytical and domain expertise necessary to achieve optimal results:
| Expert Paid Search Optimization Example |
|
Consider a keyword with a low Click-through Rate (CTR). Improving the ad copy associated with that keyword would likely improve its CTR, right? Right, but not so fast!…The expert would first examine the return-on-ad-spend (ROAS) or cost-per-action (CPA) associated with the keyword. Improving CTR too early would be like opening the floodgates to unprofitable business! The expert would initially work to improve the keyword’s Conversion Rate by improving its landing page, adding negative keywords, or adjusting bid settings until an acceptable ROAS or CPA is achieved. Only then would the expert work on improving CTR to turn up the volume. An expert would likely use benchmarking to determine attainable CTR and Conversion Rate goals for the keyword. He or she might run statistical analysis on the keyword’s “cohorts” - other keywords with the same purpose and objective – and use the 70th percentiles as the CTR and Conversion Rate goals to work towards. And how about the bigger question, which keywords would the expert spend time on? He or she would identify the keywords with the potential to deliver the highest, most-profitable growth. To find those high growth opportunities, the expert would likely follow this analytical sequence (assuming the objective is ecommerce transactions):
|
The above example illustrates the type of expert domain knowledge encapsulated in Condati’s Quant Marketer, a prescriptive analytics system for digital marketing optimization.
So, prescriptive analytics sounds like a great vehicle to get your job done faster and better, right? But the thought of analytics making the calls should also raise some rational questions (and associated requirements):
- Are the results transparent and trustworthy?
A great prescriptive analytics solution explains decisions and actions based on the original data, in a way that users can understand. - Can directives be validated before making changes on the ad platforms?
A prescriptive analytics solution engenders trust by allowing users to review and understand proposed changes before executing them. - Which actions are the most important?
Directives must be prioritized by positive financial impact, so that time spent by users delivers maximum results.
Conclusion
A prescriptive analytics system:
- enforces consistent methodology
- provides expert guidance to those who need it
- increases efficiency and effectiveness of existing experts
- prioritizes the human workload
…all at the same time!
