Continuous Improvement: Driving Performance through Personalization

Think about the last time you really were captivated by an email? What captured you? The CTA? The brand? The design? The contest? The prices? After spending my entire career in digital marketing, I have admittedly become numb to most messaging. You may have as well. But maybe you can relate, there are things that keep coming back, like when the message is tailored specifically to me.

Why Personalize?

There’s no doubt about it, personalization drives results. Every day are overwhelmed as consumers with blasted messaging. Television. Radio. Billboards. Phones. Social media. Sadly, the type of messaging hasn’t effectively changed since the beginning of print—messaging when it was intended to convert as many people as possible. But what works for some, isn’t necessarily translatable to all.

There are many reasons why a person is loyal to a brand, but the common theme is simple: customers connect with brands that speak to them. Companies like Amazon have the thousands of products, but their infinite aisle makes it difficult to find what you’re looking for, if you don’t already know the specifics of your needs. The clutter can be overwhelming. A clear, concise message tailored just for you is likely to keep you from being overwhelmed and offer exactly what you’re looking for.

A brand is a sum of everything identifying about a company–the logo, the ads, all personal interactions, every communication, every interaction and every touchpoint. Your brand is the culmination of everything top to bottom. And a brand that speaks directly to your customers? Priceless.

Ok… But How?

The most important thing is to continually improve your interaction with every single customer! It’s about progress, not perfection. There are a few common concepts that you should familiarize yourself with when it comes to personalization. They are:

  • Personas
  • Common Messaging Types including:
    • Cart Abandonment / Retargeting
    • New or Updated Content / Product Announcements
    • Newsletter
    • Event Invitation
    • Transactional
    • Lead Nurture
  • Multivariate (A/B, A/B/x) testing like:
    • Statistical sampling size
    • Statistical significance
    • Null Hypothesis
  • Performance Metrics such as:
    • Email / Message Deliverability
    • Open Rates
    • Click Thru Rates
    • Conversion Rates
    • CTAs (Call to Actions)

Further more, we’ll want to follow this using as measurable and repeatable methodology as possible. Do that by considering the following example: 

  1. Make Observations. There are underlying trends about your customers that you likely already have in your data. Take the time to gather it up, analyze it, and draw some conclusions. By diving in, you’ll find areas of interest. Ask yourself these questions:
    1. I see correlations in my customer base. Some customer behaviors are repeated over and over again.
    2. I have handful of high value customers
    3. My open rate across all email is about 5%, and my click thru rate is <0.5%.
    4. Low value customers tend to purchase once and only once.
  2. Think of Interesting Questions. What is your data telling you? Here are some common questions
    1. What types of customers do I have? Do I see my Personas? Is it time to add new Personas?
    2. Are my purchasers high value, repeat purchasers?
    3. Why are my open and click thru rates so low?
    4. Why are the best customers converting? Is it because of free shipping? Is it because they like the brands? Is it because of sales?
  3. Formulate Hypotheses. Now that you have your questions, what are the ideas you want to test? Remember, personalized, intimate messaging leads to better conversations, and better conversations lead to more conversions!
    1. I have sixteen different personas, including new the following new ones: Fashionistas, Futurists, Winners.
    2. Sending high value, repeat purchasers more messages will cause them to convert more often.
    3. My Subject Lines are soft and my CTAs are weak.
    4. My Best Customers are converting because of the Daily Deals.
  4. Develop Testable Predictions. How do you test these hypotheses? Don’t forget to use Multivariant Testing to ensure that your predictions are performing as expected!
    1. Start segmenting the messaging to your personas. Using multivariant testing, test the communique with these new personas with a controlled baseline.
    2. Create a multivariant test that delivers more messaging to one subsection of your high value customers than your control group.
    3. Create new copy and CTAs, and test it using a multivariant test that delivers the new messaging to some–and the old messaging to others.
    4. Create a new Daily Deal campaign, and test it using a multivariant split.
  5. Gather Data to Test Predictions. Run those tests until you have enough data to draw reasonable conclusions
  6. Refine, Alter, and Expand Hypothesis. Some things will work, some things won’t! Use the data that you’ve collected to refine your messaging.
    1. My Fashionistas responded well to our new messaging, my Futurists had a Null Hypothesis, and my Winners performed more poorly! It’s time to rethink my strategy for Winners and Futurists. Maybe it’s time for new messaging? A new recommendations algorithm?
    2. My High Value customers are showing a 10% increase in conversions! Great! How about their customer satisfaction? How about their returns? Positive results do not necessarily mean that I’m done with refining my hypotheses!
    3. How’d my new messaging work? Does the new messaging work? Does it work for every audience? Perhaps the subject line is polarizing–disenfranchising Fashionistas but engaging Futurists. Is it time to try new messaging based on the persona?
    4. Are my Best Customers converting better? How are margins on their purchases? Is it time to employ a new strategy aimed at increasing margins, relying less on the daily deals?
  7. Develop General Ascertainments. As you finish your tests, and refine your hypotheses until you feel they are solid, you find that they are indeed performing far better than they were before! Maybe site revenue is up, returns are down, and customer satisfaction is climbing to new highs!
  8. Repeat. Repeat. Repeat. Improvement always needs to be continual! Remember, progress not perfection.

While the process above is scientific in nature, we’re striving for personalization, and that’s never cut and paste. As you work through this–over and over again–you should be refining your messaging. It should become organically more personal. You don’t have 1 person engaging with your brand. You don’t have 16 personas receiving your messaging. You have Jamal and Jane and Jill and a whole lot more. Each of them is unique, and so should your messaging.

“Your plan of action is to make sure that you add that personal touch that our world is currently devoid of. The outcome will be true and will foster long lasting relationships that are free from excuses but abundant in blessings.” — Farshad Asl, The “No Excuses” Mindset: A Life of Purpose, Passion, and Clarity

What should I watch out for?

  • Prioritize the most important. The lowest hanging fruit will always reap the easiest rewards. You have a finite number of resources, focus on your big wins and execute them with precision.
  • Don’t stop. Continually refining your digital marketing strategy is the only way to continue to have the intimate one-on-one conversations with your customers. They’ll love you for it, and you’ll continue to reap the benefits.
  • Make your observations using as much of your dataset as possible and test that agains your most important objectives. Are you really consolidating all applicable data? What about your customer’s warranty claims? Support claims? Lending rate? Rebates? Most recent purchase? Most sought after products?
  • Personas are a great way to address 80% of your audience, but the 20% at the top marketers generate the most profit. Using a personalization driven data is driving up their results, and that goes a beyond personas.
  • Don’t take the science out of the scientific method. Measure, analyze, test, and validate! Marketers are seldom the most scientifically minded folks, but a little rigid logic can really go a long way. You can’t manage what you don’t measure.
  • Make sure your tests are statistically backed. You need statistical relevance in order to be able to draw conclusions. The larger the sample size, the more accurate your observations will be.
  • Tight Controls are required to make sure you’re changing the right things. Keep what works. Lose what doesn’t.

Personalization is priceless in this cold, dark digital age when consumers are skeptical and selective. Marketing has become less personal than ever, but technology is finally enabling that to change. Put down your megaphone, step away from the podium, and have that one-on-one conversations your customer crave. It’ll be a lot of work, but ultimately, it’ll pay for itself.

If you are looking to add more strategic, personalized experiences to your email marketing tactics, let’s chat about it today.


Author
Jamal Khan
Sr. Solution Architect, Marketing Cloud Practice