Does journalism need marketing?

Does journalism need marketing?

Or is the journalism the marketing?

This is the question I asked back in February when I started this newsletter. With all the excitement of journalist-run newsrooms popping up, I wanted to see if I could help move subscription upgrades faster.

Long story short, I started working with a small newsroom in April to do just that.

Our goal was to double their new monthly subscribers—from 100 to 200.

Did we do that? Well, it depends on who you ask.

I made several mistakes on this first project and it’s got me going back to the drawing board on my role in supporting journalist entrepreneurs.

If you’ve been thinking, “maybe we need to hire a marketer…” read this first.

P.S. I’ll be at SRCCON in Minneapolis in August and at ONA24 in Atlanta in September. If you’re coming to either, let me know! Would love to meet you.

What I got wrong: my first newsroom marketing project

In March, a journalist from a worker-owned newsroom reached out to me because I’ve been talking about this whole journalism revenue model thing on Twitter. After a couple chats, we decided to give a test project a try.

I’ve been doing growth strategy for subscription businesses for years but mostly for tech companies, not news outlets.

About our project

• Three months long (April-June)

• Set out to run 12 marketing experiments (we ended up running 8)

• Point of contact: Business Manager (also a worker-owner)

• About 1-2 days a week of my time

Starting and ending point

• Started with 100-150 new subscribers a month (on average)

• Our goal was to get them predictably over 200 new subscribers/month without increasing churn (they did get here, but towards the tail end/in the month after our project and I don’t think they would attribute any of that to my help for all the reasons I share below)

Here’s five things I got wrong about subscription marketing for newsrooms and what I’d do differently now.

Too heavy on experiments

🧪 What we did: I pitched them the idea of running 1-2 experiments a week. Targeting things like their paywall, their upgrade screen, in article upsells and their email marketing.

❌ What went wrong: Experiments come in once you’ve exhausted “best practices.” The baseline of good subscription marketing is well established. You don’t need to experiment on sending new subscribers a welcome email, for example. Just start doing that. We had a lot of catching up to do on that baseline and unfortunately some of those things (like the email I just described) are foundational and not immediately measurable.

➡️ What I’d do now: Run a checklist of foundational marketing needs and get more of that stuff into place before beginning experiments. Don’t make everything an experiment. It overcomplicates stuff we already know works.

Not getting better tools

🧪 What we did: We used their existing tools which included their website platform, their email tool, Stripe, Google Analytics and Google Drive.

❌ What went wrong: This stack has a lot of drawbacks and it’s extremely barebones by today’s online business standard. I’m mindful that lean newsrooms don’t want to add a bunch of software subscriptions, but we were really trying to race a car made out of sticks.

Google Analytics in particular wasted a ton of our time and not being able to control much on the website meant we mostly ran email experiments. That’s a pretty limited arena especially when most of those emails are primarily about, well, the news.

➡️ What I’d do now: Compose and fight for a list of tools that make marketing decisions faster or tools that market on your behalf. That includes:

• A web analytics tool like Plausible or Amplitude

• A Stripe analytics tool like ChartMogul or Baremetrics (needed for better churn analysis)

• A plug-in call to action tool like RightMessage, GetSiteControl or HelloBar

• An email referral program like Partnero

• A social proof tool like Senja

• A social listening tool like Mentionlytics or Syften

• A social media autoposter and autobooster like Hypefury

I’m not saying you need all of these but having low software support in our year of automation 2024 is putting way too much manual labor onto your team. You don’t have to live and breathe AI to know that some stuff can just be handled by computers now.

The light version of everything I listed above is about $235/month. If you pay contractors at least $50/hour and those tools save you more than 5 hours a month (which they will), they’re worth it.

[Also see: 3 ways you might be slowing down upgrades]

Not enough discussion about attribution

🧪 What we did: We attributed subscribers to our marketing project only if their upgrade could be immediately and directly tracked from an experiment.

❌ What went wrong: This sounds like a logical way to measure the effectiveness of a marketing test but it only works if you have high quality, detailed attribution analytics (which we didn’t.)

Tracing every new subscriber back to a marketing test is called “last touch attribution” and it doesn’t account at all for everything that lead to that moment. If a marketing test plays an earlier role, it gets no credit.

➡️ What I’d do now: Develop a more mature model of attribution.

Most of our tests produced new subscribers but I didn’t do a great job measuring the compounding effect of our experiments. Marketing is in play end to end throughout your reader’s journey with you. But you need better analytics to fully see that picture.

Too isolated from the newsroom

🧪 What we did: I worked directly with the Business Manager, and while I interfaced with a few of the journalists, particularly the ones working on marketing efforts, we were not well integrated. This was my idea (I think) and it made the engagement very easy to run because I wasn’t coordinating several people’s opinions.

❌ What went wrong: Sadly this failed because we were playing a game of telephone with what we learned on the marketing side and what the rest of the team understood about that. That meant that when journalists had to execute a marketing task (like “add this call to action into your article”) it was executed in a fairly haphazard and uneven way.

It also meant that many things we learned did it fact work to drive subscribers are no longer in practice less than a month later because they were not widely adopted by the whole team.

➡️ What I’d do now: Build in more newsroom training related to our marketing efforts. Get the team on board with the marketing efforts and integrate them early and often.

Missed opportunity on copywriting

🧪 What we did: I wrote a lot of the copy for our marketing experiments and then, it was all rewritten by the team in their publication voice and tone.

❌ What went wrong: Marketing and sales copy is different than editorial content. These rewrites meant that copywriting patterns for sales (like the AIDA model: attention, interest, desire, action) got lost and watered down.

➡️ What I’d do now: This is probably the most delicate space at the intersection of journalism and marketing.

I see two routes here: get a pro copywriter and onboard them into the newsroom styleguide OR get the editorial team some sales copywriting training so they can preserve the structure needed to sell subscriptions while injecting their brand into it.

Our takeaway

Marketers still have lots of value they can offer your newsroom, particularly if you scope their success metrics well and look at their work as a value add for your readers not just your revenue.

Still, it’s ideal for journalists to build marketing capabilities into their team without requiring a full time marketer. I believe we can solve some of these marketing problems at an industry wide level or at least across several publications but I really do think we’ve got to fix the software first.

I’d encourage you to audit your manual tasks and see if you can replace some of them with software. Add a bit more tooling than you’re comfortable with for a season (say 3-6 months) and see if you can notice a reduction in time spent on marketing tasks OR an improvement in revenue decisions.

Let me know if you need a specific recommendation and let me know which ones you try!