Running an email marketing program requires making a lot of decisions. Making all those decisions be overwhelming, lead to burnout and keep you from tackling important strategic issues.
Delegation is a great way to take some decision-making responsibility off your plate. Having a person (team member, contractor or assistant) take items off your to-do list can be helpful. Alternatively, you can use prescriptive analytics to delegate some of your decisions to software.
What is prescriptive analytics?
Prescriptive analytics lets you tap into your treasure trove of data to make decisions auto-magic-ly.
In email marketing, send time optimization (using data to decide which time to send a message) and dynamic content (using data to decide what content different audience segments should receive) are two examples of how prescriptive methods make decisions on your behalf about your campaigns.
When using prescriptive analytics features like send time optimization and dynamic content, you're saying, "Okay, computer, I trust you to use my data to match my subscribers up with my content and send my campaigns without asking for my approval."
That's a powerful timesaving move that can lighten your workload! Prescriptive analytics also can reduce the risk of human error and bias.
At the same time, however, using prescriptive analytics can be scary if you are used to making all the decisions. Moreover, it can introduce other risks.
3 risks to watch out for with prescriptive analytics
1. Getting crappy results
You've heard it before: Junk data in = junk data out. You must give a prescriptive algorithm good data if you want to get good results. These programs run automatically and are designed to run without human intervention. Because you can't double-check, verify or even stop messages from going out automatically unless you shut the entire process down, using these methods with unreliable or inaccurate data will undermine the benefits.
2. Assuming you can set it and forget it
Even with basic automation there's no such thing as "set it and forget it." The same is true – maybe even more so – when using prescriptive algorithms in your email marketing program. If you skip regular reviews and updates, you risk sending out campaigns with major mistakes and errors, especially if you change your email templates or strategic direction.
3. Using the wrong predictive algorithm
To use prescriptive analytics you'll either use a feature built-in to your ESP or work with a developer, data scientist or consultant to add-on these features. No matter which approach you use it is essential that algorithms used match your business structure.
Before you begin to use prescriptive tools make sure you understand how it works so you can make sure it is a fit with your program and brand.
For example, I was recently surprised to read that one ESPs send time optimization method defaults to a random send time if there isn't enough data available (and you aren't notified when this happens). That may work for you, or it may not (for example, it may fit better with your business and brand to send messages during business hours instead of randomly throughout the day).
4 action items to avoid potential pitfalls
You can reduce your risk exposure with these four steps:
1. Get clear on whether or how you want to use prescriptive analytics.
Start by considering the question, "What decisions about my campaigns am I comfortable outsourcing to software?" If you can answer it, prescriptive analytics will be a fit for you.
If not, no worries. You can call on many other ways to improve your email marketing with data and analytics. Learn about these in this earlier Only Influencers post, A Quick and Easy Roadmap for Using Analytics in Email Marketing.
2. Keep your database up to date and clean.
This is good advice for every aspect of your email program, beginning with acquisition. But it's essential when you rely on data to drive decision-making. You must keep your database of subscribers and products or services clean and accurate to make the most of these tools.
3. Ensure the prescriptive method(s) you use fit your organization -- and if you aren't sure, ask.
Before you start using prescriptive analytics in your program, make sure you understand how the methods will use your data to drive decisions. This will ensure that you use the right prescriptive tools for your program.
Also, note that these algorithms are often proprietary and can be different from one platform to another -- even if the feature has the same name.
Don't be afraid to ask questions! Your ESP, developer or agency has a vested interest in keeping you happy. Ask about resources like live or on-demand webinars and training sessions, FAQs, online manuals or user communities to get familiar with prescriptive analytics features offered to you. If all else fails, hire a consultant to interpret the jargon for you.
Above all, be persistent! You deserve to understand how the prescriptive tools you use work and how they affect your program.
4. Schedule regular audits to maximize the use of your tools.
Have you checked lately to see what prescriptive analytics features your platforms offer? An informal rule of thumb among technology vendors is that their customers use only 20% or so of the features a platform offers.
You probably keep tabs on day-to-day trends in your email program, tracking engagement, revenue, retention and other benchmarks. A quick audit of your needs and technology can help you identify opportunities. Investing a little time upfront to maximize your use of technology can save you hours of work later.
Wrapping up
It's natural to feel overwhelmed from time to time with the demands of operating a digital marketing channel. But if you feel as if you're running from one task to another without the time to reflect on what you're doing or where your program is going, consider delegating some of your decisions to software using prescriptive analytics!
Photo by Andrew Neel on Unsplash