Expert Interview Series: Part 3
What really goes into making big data work for email marketing
Big data has marked the beginning of a major revolution. It is not about teaching a computer to think as humans; it is more about applying math to huge quantities of data so that you can infer possibilities and predict every customer’s behavior.
Although it can get complicated and messy at times, big data is transforming the way businesses work. It has made marketing more effective and emails are no exception.
We got in touch with Ryan Phelan, co-founder at RPE Origin, to get deeper insights into data and its significance in emails.
Let’s kickoff this interview by going back to the basics. What is big data?
Ryan: The concept might be basic – on the surface you have a lot of data – but the implication of big data is more advanced. It’s data that is organized, accessible, usable and actionable.
A lot of peope have a lot of data. They can see email opens (for now), clicks, where they clicked, whether they’re active or inactive, and what they did after they clicked. Big data is an advanced strategy because it makes vast amounts of data actionable beyond a single use or app.
Do you have big data in your ESP? No, you have lots of data. Some of it is actionable, but only in relation to your email response data.
If you have other data, great. Now you have the start of big data. But more important than that is what you do with it. That should be your focus.
We have seen a surge in the use of big data in the realm of email marketing. What makes it so popular?
Ryan: But have we really seen this? I don’t know that we have.
We have seen in the last 12 to 18 months with the pandemic how much data an email marketer can have, or have access to, but I still see marketers doing list-based segmentation and campaigns. That’s not a “surge.” That’s routine.
The advantage to Big Data is that we expand the possibilities of our marketing. We’ve said the gateway to greater success is segmentation, production based on propensity, and other natural reactions to customer behavior.
All of this data is available if marketers take the opportunity to get it and organize it for use. Ninety percent of the email in my inbox is still “batch and blast.” It tells me these marketers still don’t know who they’re talking to and sending the wrong content.
The secret fact of email is that it’s the only channel where you can make money even if you don’t do it well. But you’ll make a lot more money if you can organize, use, and expand on your data beyond the basics.
If you had to share the top 3 applications of big data in emails, what would they be?
Ryan: First up, the most advantageous email programs are triggered emails – marketing automations right out of the gate. You can see a customer’s action or behavior and react in real time.
That could be a product upsell after a purchase, or special messaging when you detect signals that a customer is moving (see my latest Only Influencers column for more on this overlooked opportunity).
It could be browse behavior or buyer history. Your email system is looking at consumer signals and acting on them.
Next, I would add transactional emails. I buy Product X, and you know people who buy Product X usually go on to buy Product Y. That’s called “next logical product,” and it’s an email you can send to follow up on the first purchase.
These are emails that are constructed automatically at time of send. It’s an advanced strategy that relies on models, propensity, and inventory to support the customer profile. That complexity takes time to build, but the additional money you can make far outweighs your time investment.
It also makes building emails much easier because you can rely on your data to tell you what to send, not the merchandising team. You’re more likely to include products that your customers actually buy, not just what comes close based on your merchandise assortment.
The more sophisticated your approach to using data, the better results you’ll have. That’s why it’s not enough just to have data – it’s what you do with it that makes the data valuable.
Can you share with our readers how exactly it helps in more effective personalization?
Ryan: I see personalization as not just putting someone’s first name in a subject line or message copy. That’s entry level personalization. Anybody can do “Hello [first name].” I see personalization as how personal that email communication is to the end user. Rate your own emails from 1 to 10 on how personal they are to your own interests and activity.
Do the products in your emails relate to your own predisposition or propensity, or your purchase history? That’s level 8 to 10 personalization.
Data helps you get smarter in how you market. It’s akin to the websites you visit every day. They should be smart and relevant experiences if the site is built to take in all types of data and present you with a highly personalized web experience but it takes a lot of work on the back end to make that happen, especially at scale.
Effective personalization means the consumption of communication happens more regularly because the communication is more relevant.
Marketing Automation continues to be a goal for marketing professionals. In what ways can you leverage big data in email automation?
Ryan: The lifeblood of marketing automation is data, and the more data you have, the more decision trees you’ll have in that automation. Take the abandoned cart. It’s easy to detect when a customer abandons a cart. That signal triggers an email, where you say, “You left something in your cart. Come back and check out.”
But, if you increase the data in your email, you make it more relevant and give your customer more reasons to act. You can say, “Hey, you left these cool shoes in your cart. Did something go wrong? Do you need more information? Let us help!”
The more data you use in your marketing automations, the more intelligent and effective they will be. But don’t overlook the planning process. Always think it through first. Ask yourself, “What other pieces of data can I bring in to make this even more effective?”
Go back to your own inbox, and look at the automated emails you receive. Rate these from 1 to 10. Are they fancy? Or did the marketer just phone it in?
Can big data help boost email deliverability rate?
Ryan: No. Many people believe this, but the only thing that’s important to deliverability is having the right email address. That’s data, but not necessarily big data. What’s more important is how valid your dataset is.
From a deliverability standpoint, did you validate the email address at opt in? Did you use double opt-in?
Another issue is how recent your data is. How old are your records? A 6-year-old email address might not still be in use, but it’s not a hard-and-fast rule. I still use the Gmail address I got back in 2004 when Gmail launched.
The perennial challenge with any large dataset is the validity of the data itself. Your first questions to your CRM group should be “How old is this data? How recent are the responses?” That plays into deliverability.
Subject lines, as we all know, influence the email open rate. Do you think big data can help to craft more effective subject lines?
Ryan: Big Data can help you craft an amazing subject line. Should you use it to craft the perfect subject line for each subscriber? No. It’s a waste of time.
You’re better off spending that time and effort on crafting the perfect message that they’ll open and read. Yes, there’s a science to creating good subject lines that persuade people to open your email. But we’re addicted to the subject line. With the recent news that Apple will be killing off open rates, who cares?
Yes, the subject line is important. But let’s focus on the things that make us money – and those are in the message itself.
How have you applied big data in emails through all these years? It would be awesome if you can share some real life examples with our readers.
Ryan: The biggest big data project I ever worked on was a massive telecommunications company with more than 2 billion email records. Our goal was to reactivate customers who had moved to other brands. With that big of a dataset, we could do fun stuff.
We took a methodical, data-driven approach with those email records to match them to third-party data on factors such as where they live, what their interests were, how old they were, if they had children in the home, if they were married, watched the news, played golf, and more. Then we used data pieces to fun complex models that grouped these records into seven cohort groups, all similar to each other.
Once we had that, we developed custom strategies for each cohort group, such as people who were attuned to news and technology, people with college-age kids, people who were the tech support for their families, and so on.
Building that dataset changed my perspective on how data can be used effectively and responsibly to craft messages. But it was a lot of work. Nothing in this world is easy, so buckle down and get it done. The bonus for all of our work? Effective campaigns that could be replicated time and time again!
What can we expect in the near future as far as application of big data in email marketing is concerned
Ryan: The propagation of consumer data platforms (CDPs) have given businesses the benefit of the single-record view of a customer. That’s what we’ve been talking about forever. A CDP organizes the data to show us everything we know about this customer in one place. That makes it easier to use the data for effective messaging.
To be effective email marketers, we need to know not just what our customers do with email but also how they behave on our websites, in social media, with videos and SMS. We need to consolidate that data to draw conclusions and make decisions.
That’s the CDP’s promise: to be able to query the database more effectively to find good segments. If more companies adopt CDPs, email will finally become more sophisticated.
Any insights that you would like to share with marketers who wish to incorporate the power of big data in their emails?
Ryan: Big data starts with small data. Take whatever data you have, organize it, use it, and then add to it. There’s no threshold that tells you you’re now using big data.
Big data is multidimensional. You see that someone bought something, but what did they buy next? Someone browsed your website. What did they do after that? Someone watched one of your videos. Which one, for how long, and what did they do next?
The hardest part is just getting started. Tell yourself, “I refuse to participate in the “batch and blast” mentality any longer.” Then work out a road map to move beyond that muscle memory.
The gateway to that achievement is through greater use of data. You might have a lot of data, but if it’s specific to one event or one channel, you don’t have big data. You just have a lot of data.