Designing for Retention: Product Development Insights
AMA with Neil Katz, Chief Customer Officer, Advance Local
NEIL KATZ, ADVANCE LOCAL
“We have an open culture of information sharing. And a large percentage of our workforce are journalists who spend their lives finding things out. So nobody waits for data to come from the high castle. Leaders are awash in data. Our greater challenge is agreeing on what insights they provide.”
Full Transcript
*Questions submitted by community members
Transcript and highlights from our Ask Me Anything.
Neil Foster is the Chief Customer Officer at Advance Local and has amassed his subscription expertise across roles at major news publications (traditional and digital). Neil has done it all in publishing has been a journalist, producer, editor, and executive lead. Please join me in welcoming Neil to our SubscriptionWorks AMA.
Hello, House of Kaizen community. Happy to be here. Looking forward to interacting with all of you. And learning from you.
Okay, let's get started with our first question.
Question: For the people who may not be familiar with Advance Local can you please give a quick intro?
Advance Local is a holding company of 25 newspapers and 10 regional websites. Our print products like The Oregonian, Cleveland Plain Dealer, and NJ Star-Ledger, typically serve medium-sized cites. Our digital sites, http://NJ.com NJ.com, MassLive, http://Mlive.com, AL.com, and others are usually statewide news properties. Our digital content reaches 50M people on-property each month, double that if you include AMP and Facebook. We’re owned by the Newhouse family, the same group that owns Conde Naste and Reddit and has been publishing news for almost 100 years.
As a resident of NJ, I appreciate your contribution...I read NJ.com regularly
We appreciate it.
Question: You have an educational background in Journalism and Digital Media, coincidence?
I was actually a computer geek before a journalist. I studied digital media in college and built a boutique design firm creating CDROMs (yeah that long ago) and web 1.0 sites. I sold that shop to an online education company, which quickly imploded in the dot com bubble, and realized I was bored with computers. I retooled, became a journalist, and got to report in India, Iran, and other interesting places. I wrote for papers. I produced for TV. I tried to avoid being the tech guy. But sure enough within three years of my new career, I was working on the digital side of journalism. And today I find myself solving the same kind of problems many tech shops have: how do we build an audience that loves us; how do we use technology to understand who is willing to pay for our services; how do we use lean startup techniques to test out new editorial product ideas; how do we mix first and third-party data in a privacy forward manner to understand what content readers want. Don’t know if it’s a coincidence. But it’s certainly a full circle.
I love how you built your experience wearing many hats in publishing and now lead the charge.
Love this. The dawn of digital media wisdom is upon us as we see similar challenges to those faced at the beginning of this WWW thing
Question: Assuming you have learnings by product and brand, how do you prioritize resources (money, time, or people) between products and brands to capitalize on the opportunity?
The simple answer would be focus on NJ.com (our biggest property) or Oregonlive.com (our best converting property) but it really comes down to market participation. Each market leader comes with a unique POV and we partner with markets where we have alignment. Some want to focus heavily on starts, others on retention. Some are focused on in-market versus out-of-market segmentation. Some are heavy sports markets. Others are politics driven. And so on. It’s been really fascinating to see how each leader wants to operate and also how the audiences behave so differently.
What is the general flow of learnings across brands? Who owns this effort at Advance Local?
Learnings flow in many directions (the market presidents know each other a lot longer than they know me but we have set up a very interesting group, called StoryLab, designed to run content, marketing and tech experiments across newsrooms and then share that information across the org. Each newsroom nominates a representative to participate. They meet bi-monthly with our chief content officer and members of the subscription team. Participants cook up ideas, we bring in research to size the audience, we fund new content creation, often put paid media behind the ideas and see what rises with our readers. Last year, our experiments found that certain types of data-driven explainers were massive traffic and subscription hits. We discovered readers would pay for advanced weather maps and we learned that we could scale high school sports across the enterprise. This year we already have a pipeline of half a dozen interesting ideas. A few will live, a few will die and one or two will become mega hits.
Sounds like you could start selling StoryLab to other local news orgs...they could benefit.
Sounds fun and fascinating!
That's a great setup to share learnings and ideate on new features!
Question: What KPIs are towards the top of your cross-product/brand executive dashboard and why?
At the very top are high-level outcomes: total readers, total page views, total subscribers, new starts, stops, and 13-month retention. Underneath that, we track dozens of important variables that drive those outcomes. We track users by platform and by engagement levels. Advertising tracks ECPMs, RPMs, and sell-through rates across platforms and providers. Newsrooms track story performance and trends. The subs business tracks blocks (how many times people are stopped from reading a story), conversion rates (how often people pay when stopped), and the unique performance of every topic on every property. Underneath that is a customer data platform where we experiment with building segments by a combination of variables (i.e. show me the reading habits of highly engaged high school sports fans who live in NJ and pay by the month) and use machine learning to predict propensity to pay and churn.
Question: What types of research have the team scoped? and does it generally span multiple markets or start as a POC?
Most of our research starts with a single or a few markets joined together. We've looked at everything from how people use crime information (safety more than spectacle), why people use us for weather information despite having a sea of options (they like to compare forecasts and they like our hyperlocal understanding of their neighborhood), to what's stopping a power reader from buying (mostly they enjoy the large amount of free-to-read content we have). We also are constantly polling new signs up and stops for issues we can quickly address.
Very true, I don’t think I would ever trust the local news reporting of a national paper...they feel too disconnected from the community.
I like having different sources for trusted information local, national, and even hyperlocal (towns).
This is a great insight and really speaks to that local nuance to content and control you mentioned previously.\n\nOn average people subscribe to 1.1 news sources, typically national first and many a local on top of that -- so knowing the local value is critical to building and sustaining that audience while the national/international news is in the spotlight
Question: Do you create unique personas and journey maps for each of your products/brands?
Yes and no. We map out journeys that are mostly enterprise-wide, for example how we move you from FBIA to our website or how we move you from one visit to retargeted visits to registration to the newsletter to purchase. We then use our customer data platform to customize the audiences and messages we target based on the content strengths of each market.\n\nWe do have a few cases that are market specific. For example, NJ.com has a very robust high school sports business. That journey often starts with SEO to our high school sports data platform, then drive to content, newsletter, trial, etc.
This must be hard to balance broad priorities and efficiencies with locally nuanced priorities.
Question: When you get a learning from one specific brand, do you test it with the other brands or do you go straight into implementation?
It depends on what the markets want to do. In our company, the enterprise sets goals, brings technology, offers services, and teaches best practices, but ultimately a local market runs its business. Here’s a perfect example. This year we’re rolling out a machine learning software that chooses which stories are gated and which are free to read. We worked with two local markets that wanted to test the idea. The results were very positive. Now we’re rolling it out company-wide, but each market has its own flavor. Some want to dive right in. Others want to test it again to make sure the results are as good in their market. And still, others want to launch the service testing various gating heights. Doing things this way can sometimes make us slower. There is only so much capacity to handle variations. But it also helps us learn in a more robust way, as we build results from many different market conditions.
Question: How do you decide what learning should be shared and what should NOT be shared?
We have an open culture of information sharing. And a large percentage of our workforce are journalists who spend their lives finding things out. So nobody waits for data to come from the high castle. Leaders are awash in data. Our greater challenge is agreeing on what insights they provide. That alignment is a messy, argumentative process that feeds into the tensions in our business. Ads versus subscriptions. Local versus enterprise. On-property audience versus distributed. And so on to make the signal from noise, this year we’ve built a cross-disciplinary group (finance, subs, ads, data, locals) tasked with owning the source of truth on around 30 vital signals in our business. Talk to me in six months. I’ll tell you how it’s going
Sounds like a great setting for a TV show newsroom office.
If only we were in the same room these days. TV will need to find a way to dramatize millions working from home. Not a ton of drama in that.
Ha! great point....perhaps Zoom will be the next great content production source for reality tv
Question: Understanding that local markets have their own content priorities -- eg food in Charlotte and football in Philly -- is there a one-size-fits-all approach to subscriptions across all your markets?
Our tech stack is mostly enterprise driven: paywall; CMS; customer data platform; ad server and so on. But our tactics tend to be quite local. Food is huge in NJ. Sports are huge in Cleveland. Crime and politics matter everywhere. Outcomes are driven by the history of the brand, the size and wealth of the audience, the composition of competition (both free and paid) and the unique expertise of certain reporters to name a few. And the biggest factor is one we don’t control at all… NEWS. Is there a hotly contested local election? Did the team make the Final Four? Was there a horrible shooting? An inspiring young leader? And so on. We say that news is our chief revenue officer. Our job is to be ready when it breaks.
This sort of hierarchy really resonates and do a similar approach here with identifying the
“North Star” KPI, but also defining and understanding the optimization kpis that funnel into that so that we can better measure and understand how different tests and tactics bubble up into a bigger impact
Question: Local news and independent journalism are critical right now. How do you empower your brands to pursue this worthwhile goal and does cross learning play a part in this effort?
Luckily this is the beating heart of the business. Unlike some others, the company has wisely not strayed from the local mission. Sometimes that’s been hard. For example, nationals got four years of Trump Bump. We got a Trump Slump. A similar thing is happening with Ukraine. People have a finite time to read news and sometimes sacrifice locals for bigger global stories. However, by staying true to local, when weather, sports, politics, crime, corruption, and a thousand other priorities break at home, we continue to be the top source in almost all of our markets.
Makes sense. Slow growth is still growth and may be more sustainable in the long run.
I wouldn't say we're slow growth. It's tailored to market needs. If a market wants to jump ahead with results they've seen elsewhere, totally fine. And they often do. If they want to be more cautious, replicate results first or try something else, that happens often too.
This makes a lot of sense!
Question: Is there a benchmark across all markets for the "right” balance of advertising vs subscription revenue?
There are a few ways to look at this. If you merge print and digital, we are a subscription-forward company with rapidly declining circulation. If you look at digital alone, we are an advertising-forward company with rapidly increasing circulation. But really how our markets look at the world, is by what year can they support their newsrooms with digital revenue only. The existential issue in the business isn’t really whether ads or subs is dominant. It’s making sure we have a healthy economic future when print is gone. And for that we are trying to build a balanced attack: advertising (both direct and programmatic), subscriptions, events, premium podcasts, affiliate revenue, sports betting has launched in a few of our states and we are doing pay-per-sign-up deals around that, we’ve even launched professional subscriptions around the legal marijuana industry.
So smart to tackle the emerging Cannabis industry at the local level.
NJ is about to get very interesting.
We actually are in the market for what must be one of the greatest jobs in America.... A full-time marijuana reporter and critic.
Please post that job description in our job board channel!
Question: What are some of the unique challenges local/independent publishers are up against that major publishers often don't need to consider?
I guess we have exposure to both. Each of our properties is hyperlocal, but as an enterprise we reach 50M readers. From the local perspective, it’s harder to find enough audience to build a big subs business. NYT and WSJ can choose any topic in any location and try to build around that. Locals can’t. We’re also counter cyclical. As I mentioned before, often the largest national and international events shift audience away from. We also have stiff competition from free traditional sources like local TV.
I know we are at our 30-minute mark right now but I am going to throw up the few remaining questions we have.
Thank you for all of these fantastic insights.
Question: With a generation (or two) of people who use free sources (social) as their primary source for news, especially as it pertains to local content, how has Advanced Local properties tried to position themselves when offering a paid subscription? Have you seen this trend shift at all in recent years?
There’s no easy answer here. When readers start on Google, FB, Insta and Tik Tok, that puts us at a disadvantage. Even if we get the traffic, those aren’t our customers. Google (via AMP) and Facebook (FBIA) have allowed us to build a great advertising supported content businesses on their platforms, but their support for subscriptions are weak. That said we do have a path. Content we publish off platform can bring people on platform. From there, cookies help us retarget those people to come back again. We try to drive those readers into newsletters and later registration by tailoring content to their needs. Some small percentage of those readers will eventually become paying customers. But we shouldn’t sugar coat it. These are long funnels and require lots of free-to-read scale to get a subscriber number that’s interesting. Ultimately, we feel our secret sauce (and everyone’s really) is to be original. Each of our markets has unique content that you really can’t get elsewhere. NJ is famous for high school sports, food and politics. Cleveland is famous for Browns coverage. The Oregonian has been the premium news source in Portland for decades. And so on. Not everyone appreciates that enough to pay. But the news hounds do.
Question: What subscription businesses do you follow for best practices? Aka what does good look like?
NYT for program diversity (news, games, cooking). Sirius for call center excellence (stop saves, upsells, choice framing), Duolingo for great engagement mechanics and seamless layering of paid features on top of a robust free offering. Spotify for retention lock-in (good luck moving all those playlists to another service).
Question: What is the one subscription (besides one of your own) that you value highly?
I really like RealVision Crypto, a video interview series by macro analyst and crypto investor Raoul Pal. He does 1-2 hour interviews with just about everyone in crypto large and small, as well as major institutional investors. The content is completely unique and Pal brings insight and expertise to a confusing and fast changing topic. Prices run from free to $6,500/year.\n\nOn the completely other end of things, I like RV Life, a mapping app for RV road trips. It’s surprisingly hard with Apple and Google maps to figure out if you’ll get ticketed for driving a camper on a particular highway or smash into a low hanging overpass. $60/year.
Please join me in thanking Neil Katz for sharing his experience with our community. This is a very revealing discussion and a rare peak into a unique business.
My pleasure.
Thank you to our SubscriptionWorks community for your questions as well.
I hope you will stick around the community and join us for our next AMA, Neil!
Will do!