I had the pleasure of meeting Jacqueline Boltik at this year’s Litmus Live in Boston. She gave a wonderful talk on using Python to improve your email marketing data. I was so intrigued by her work and history that I cajoled her into being the inaugural interviewee for The Better Email. Enjoy!
Jacque Boltik is a consultant, speaker, and educator at the forefront of new digital strategies. She’s been scraping the web since 2009, and is the founder of Kingrail Consulting—a boutique firm specializing in data science and email. She also teaches a course on email marketing at General Assembly and volunteers as an instructor with Girls Who Code. Jacque graduated from Harvard College with honors and in another life was a figure skater.
How were you first introduced to email marketing and design?
I was first introduced to email marketing while working in brand marketing at PepsiCo. The use of email was part of a larger product marketing strategy I was overseeing. I remember being pleasantly surprised by how cost effective it was.
Tell me about your work at the LA Times.
As Director of Strategy and Business Development my role was to conduct due diligence on new ideas, form partnerships, and launch and staff new initiatives. On the surface the initiatives under my purview seemed fairly diverse - launching a joint venture in China vs. email newsletters - but at their core it was the same concept: finding new ways to package and make money from existing work. The use of email was proven to be a very effective way to promote a wide range of initiatives.
How does working in a newsroom compare to working in other email marketing roles?
Prior to working on the business side of journalism, my experience had mostly been in traditional brand management, marketing, and product development at PepsiCo and Starbucks. I worked in sports marketing on NFL player, team, and league sponsorship at Pepsi (Super Bowl XLVI!) and launched a new line of Starbucks branded coffee products (that you can find in the chilled aisle of your local grocery store). My role focused more on developing 360 marketing campaigns, of which email had a role but was not the main focus.
The biggest difference between working in consumer packaged goods (CPG) marketing and the business side of journalism was the hands on nature of working in journalism. In CPG there were an army of agencies standing by on retainer to help, and your physical product takes years to develop. In journalism, budgets are smaller and your product is created every day by people in a very direct way. Working in business development in journalism requires a different level of flexibility, and also allows you to create and test products in a very nimble way. Whereas in CPG we would develop new products 3 + years out from launch, in journalism we were able to test ideas and new products in a matter of months, not years.
In journalism, budgets are smaller and your product is created every day by people in a very direct way.
What challenges are unique to email for journalism?
Digital media in the past has largely been driven by chasing traffic; although that’s starting to change… In journalism, it’s fairly easy to grow a large email list. However, the focus on having a large list size is a challenge. Over half of internet traffic is from bots, and email is not immune. While most media companies clean their email lists, there’s a pressure in general to have the largest lists rather than the cleanest lists. I think this is an industry wide challenge in email, but is probably worse for journalism and industries that rely on the ability to attract outside advertising revenue.
What do you think the future holds for digital media?
There’s a huge opportunity for data science powered by email to strengthen online media revenue models. For paid subscriptions - having someone’s email allows you to combine previously siloed databases and use data science to build more sophisticated models to reduce churn and improve acquisition. For digital ad revenue - I think email holds the key to changing the losing digital CPM game. When doing online digital ad buys, it often costs the same to reach a bot and high level executive. This doesn’t make any sense, and I think we can do better - although some of this involves advertiser education away from CPM quantity and toward CPM quality.
You’ve worked with some big newsletters like Lenny, The Ann Friedman Weekly, and Clover Letter. How did you get involved with those projects?
I became involved with these projects due to my work at the LA Times which was relatively early to launch high quality editorial newsletters. I oversaw the launch of over 10 LA Times newsletters and was able to bring my skills and newsletter expertise to these projects.
What lessons did you learn working on such high profile projects?
I think it’s really important to have some level of technical understanding of email—even if you are on the business or editorial side. I learned Ruby on Rails about 6 years ago (shout out to Michael Hartl and ended up learning the technical side of email out of necessity to get the LA Times newsletters off the ground quickly. Through that process I learned email is very nuanced and it’s easy to run into issues that seem mysterious and hard to crack if you don’t have a technical understanding of, say, how spam filters work or how HTML email is rendered by different email clients. For higher profile projects it’s important to get these details right. I think it’s surprisingly hard to find people who have this knowledge.
I think it’s really important to have some level of technical understanding of email—even if you are on the business or editorial side.
How important do you think design is in email marketing?
I’m a fan of the one arm, one eye, one thumb rule for email design—where you hold your phone one arms length away, close one eye, and try to use one thumb to navigate the email. I think I first heard about this technique from Fabio Carneiro at MailChimp. I’m delighted when I see emails that are well designed and push the envelope from a technical perspective, but there’s still so much work that needs to be done even to get to this basic level of mobile readability. I tell this rule to pretty much everyone I meet :)
Some of Jacqueline's clients.
You gave a great talk at Litmus Live about data mining in email with Python. When did you first start combining programming with your email work?
Thank you, glad you enjoyed it. I think it helps to solve a common challenge. Most companies have a lot of data, it’s just not in an easy format to analyze, or it’s in different databases. I began using Python and pandas at the LA Times because the out of the box reports and KPIs did not answer all the questions I wanted to ask. Sometimes data I wanted to include in a report was in another database. Using python and pandas allows you to combine various data sources and answer pretty much any question you can dream up.
What are some ways people can get started with using Python (or any other language) in their own email programs?
In the past I’ve used STATA, R and Excel for data analysis. In my opinion Python / pandas is easier to use and faster. It’s continued to grow in popularity in the data science community. Beyond analysis, you can build a full-scale application. Or spin up a server. There’s a lot you can do to improve email data by combining your data sources using Python. If you want to learn more, I recommend Wes McKinney’s new book Python for Data Analysis, 2nd Edition. You can order it on amazon here. I read a pre-release version and can’t speak highly enough about the book (not a paid endorsement, I’m just a HUGE fan of pandas and Wes is the creator of pandas).
How did you get involved with Girls Who Code?
I was looking for a volunteer opportunity and think learning a technical skill is extremely valuable, no matter what career you want to pursue. Girls Who Code had a good reputation so I applied to teach at a local chapter.
What does it mean to you to teach at Girls Who Code?
I’m self taught, so learning to code was a real struggle at times. I try to be the resource I wish I had for the girls. They teach me so much and it’s amazing to see how quickly true digital natives pick up new skills.
What’s the most memorable experience you’ve had at Girls Who Code?
It’s not one moment in particular, but overall what’s made the biggest impression on me is how supportive the girls are of each other. They are a joy to work with!
You also teach at General Assembly. What are the most common questions you get there about email marketing?
Common questions related to how to develop a strong email strategy, how email relates to other marketing efforts, how to assess the success of your email program—or a given campaign—what good design looks like, general best practices, how to clean your list, there are many questions… The audience is usually a mix of B2B and B2C so I try to cover as much as possible and provide links to resources for topics we don’t have time to go in depth on.
What is the one thing you constantly see people getting wrong in their email campaigns?
There are a surprising number of emails that are not mobile responsive. So that one probably bothers me the most since it’s just such a lost opportunity. This is something people in the industry are well aware of but for marketers who are being asked to wear more and more hats it’s sometimes overlooked.
The one thing I’m most passionate about helping to fix is the issue of poor data integration. As a consumer I receive emails and ads on social that are not relevant to me from companies I know have my data. For example, I get ads asking me to subscribe to services via email and online from companies where I already pay for the exact service they are advertising. This happens because their databases are not well-integrated. Using Python and pandas, it’s possible to build powerful but lightweight solutions that can help to address this issue.
What are some opportunities you’ve seen in email that you don’t think enough people are exploring?
I think there’s a huge opportunity on the data side. Beyond opens and clicks, by collecting email addresses companies can use related data to better understand their target audience across the board. One of the challenges many companies face is that data often lives in silos. You can use email data to combine various data sources, to build a richer data set, then data science to build models to analyze your richer data set. It’s hard to understate how valuable this is—there’s a famous example in the data science world with Netflix. Netflix had a Kaggle challenge where the team who could write the best algorithm to predict user ratings for films would win $1,000,000. Later Netflix released a new feature when you login that asks if you or your child is watching. A basic model with the field that indicates you or your child was ultimately better than the fancy Kaggle model (which lacked this information). Once you have someone’s email address you can find out a lot about them and use that information to build a richer data set.
Beyond opens and clicks, by collecting email addresses companies can use related data to better understand their target audience across the board.
What advice do you have for anyone looking to get into email marketing?
My best career advice is to become a T-shaped person: learn a little about a lot of disciplines, and develop a deep expertise in an area you find interesting. If you’re looking to get into email marketing, start by learning about the different areas email marketing touches, then research the areas of expertise within email marketing that excite you. I also recommend building technical competencies. Even if you are not in a technical role, having a basic understanding of how things you touch work will help you to be successful and stand out.
Even if you are not in a technical role, having a basic understanding of how things you touch work will help you to be successful and stand out.
How do you stay sane outside of the email world?
I enjoy a good cocktail. Each season I try to explore new go-to drinks in order to discover new favorites. My go-to cocktail this summer was the classic Daiquiri: 1 ½ ounces light rum, ¾ ounce fresh lime juice, and ¼ ounce simple syrup. Best with Havana Club White Rum and fresh lime juice. Right now I’m sampling fall cocktails, so far the Mount Pelee is my favorite… Labor of love cocktail is the Ramos Gin Fizz but it takes so much work I feel bad ordering it out, so mostly make it at home when I want a good arm workout - you have to shake it for about ten minutes!