One of the biggest themes of my spring interviews is that digital transformation is a non-starter when you're coping with dirty data. Another one that's been prominent at Salesforce Connections: there's no escape from email.
If anything, email is strengthening its hold in the core of the marketing mix - albeit in a modern way that brings topics like hyper-personalization, next best offer, AI, and embedded email commerce into the conversation.
The themes of dirty data and email's persistent relevance converged during my interview with Mark Nylec Sr. Director, CRM and Email Strategy at Gannett. Our chat validated email's important while calling into question lazy thinking about "AI" as the magic ingredient for email engagement.
Gannett is the publisher of USA Today, which is in the throes of a digital re-invention. Gannett also owns over 100 daily newspapers, and nearly 1,000 weeklies in six countries. Gannett's reach opens up interesting "synergies," if you will, between USA Today's marketing and localized content.
Email marketing and the digital push
In the midst of that is Nylec and his email marketing team. Nylec joined Gannett from his email marketing role at AARP. Nylec knew he was jumping into the deep end of the pool with Gannett. As he told me, he liked the "messiness" of the data challenge:
It seemed like it was a mess, which is something I like cleaning up. Obviously I could have stayed At AARP forever. But it's kind of fun to get in something new, and start tearing it up.
When Nylec joined Gannett, email was not a marketing emphasis. The data needed was tucked away:
At the time, email was not a big part of their solution because the data in their environments was difficult to access, and there was no clear leadership in what was going on in the channel.
But given USA today's digital emphasis, that had to change:
The big push is to get digital. It's to be more digital as print subscribers are reducing.
In theory, USA Today had the right tools at their disposal. When Nylec joined Gannett, they were already a long time Salesforce Marketing Cloud customer. But they weren't getting much out of it:
At the time, they were ready to drop the Salesforce tool and go with somebody else because they were just having such a hard time even getting simple things like data out of the environment that was in there.
Nylec wanted to clean up a mess - and he got one. He couldn't tell how many unique emails were in the system:
It's really the way the system had been set up. Gannett's been customers of Salesforce for many years. Stuff was just jumbled - people did not know how to get to it.
Another problem? Gannett's users had built a custom data extract for data tracking with Salesforce. It was causing more problems than it was solving:
They were creating such a huge extract of data on a daily basis. It would just fail and you're totally messed up at that point, once the data's screwed up.
Nylec rolled up his sleeves. First step? Go back to a standard tracking extract for reporting:
I moved them to using the standard out of the box tracking extract, which should've been the direction from day one, to be honest, which is something you can turn on in five minutes - as opposed to nine months worth of work.
Another classic data quality problem was also in play: siloed data.
When I got there we had, I think, three-hundred and fifty different business units in the system and that's why the data was so disparate. We had data across every one of the business units.
Consolidation within Salesforce solved that:
I consolidated actual business units in the system. I deleted or deactivated inactive users, and created nightly processes to take the data out of the business units to put into a shared environment, so that I could access them and run reports.
Clean data pays off
After two months, that phase of the cleanup was done. Nylec had one thing going for him: Gannett's email lists are highly engaged, with 30 to 40 percent open rates for daily sends. Bottom line: readers want the newsy emails, and are all confirmed with double opt-ins. Now that the email lists are humming, the fun of business model change begins:
We know how exactly many email addresses we have. We know how many are active and inactive. I can look at any point in time and say how many people have opened and clicked during that point in time, because I've got that all dropping into a table. Data's accessible, it's able to be used and consumed and we can run campaigns against it now, which we just couldn't do before... So now we're at a point where we can start actually building and creating something pretty cool.
With a whopping 1.5 billion email messages sent to subscribers last year, email is an integral part of Gannett's digital plans. Nylec's team is still primarily using the daily emails to drive readers to the site, but he sees that changing. Newsletter monetization is high on his list, as well as getting folks registered with deeper web site profiles.
Evaluating AI and personalization
One interesting thing about Nylec: he isn't as interested in hyper-personalization "AI" type features that are heavily hyped these days. His subscribers are happy with their content sends and daily is the right rhythm, so "best day of the week to send" algorithms aren't relevant. That has led him to be skeptical of the AI emphasis we hear from just about all vendors lately, including Salesforce. As he said to me:
The thing I struggle with is the AI talk. People are talking about, "Oh! The AI, the artificial intelligence, great!" They talk for two minutes about a grandiose picture of AI, but nobody actually talks about how AI works.
That doesn't mean Nylec isn't interested in some functions that fall under the "AI" umbrella. Embedding personalized recommendations in newsletters via "next best offer" functionality is relevant to him. At Connections, he was on the lookout for customers that have successfully done this.
He's also looked at "AI" technology that can automate steps in subscriptions. For example, enabling subscribers to sign up to the web site without leaving the newsletter, and auto-populating the fields to smooth the process:
The AI could then recommend registration on the site, or it could say, "Hey, let's get you digital or print subscription," based on all the factors we know about you. So, that's what I'm looking at from an Einstein or AI integration angle, and trying to build that out as well.
With the lists clean and easy to manage in Salesforce, Nylec is ready to rock. He plans to aggressively push the email sign-ups via the web sites:
Once we get that, I'll have a nice funnel of people coming into the channel, so I'll be growing my list better than I am right now.
GDPR isn't a big issue for Nylec because he doesn't have many European subscribers. He also maintains good list hygiene, which should bode well in case other privacy rules are put into place in the U.S. Gannett's lists are all double opt-in; inactive subscribers are regularly purged. Deleting data upon request within Salesforce is easily doable.
Nylec is also working on enhancing Gannett's "inbox placement" so they hit even more eyeballs. An integration with Return Path is in the works to further improve their inbox placement and deliverability.
All of that should support the meat of a digital effort, which is: adding to the bottom line. If AI can help to automate the process of selling paid subscriptions, that's the kind of AI Nylec has plenty of time for:
If I could start doing that and getting some value selling subscriptions... Moving the needle digitally within our emails, that's huge.
Now that Gannett has their email program in the digital fast lane, I'm looking forward to his updates next year.