The 19th — Role Responsibilities

Tell me about your experience measuring total journalism reach across platforms.

At Spotlight, our CEO wanted a consistent way to answer questions about our total reach, especially from funders. I developed an organizational reach framework that brings together our website, newsletters, social platforms, publishing partners and other distribution channels. The final deliverable is a quarterly visual showing total measured reach and each channel's contribution. The challenging part was that platforms don't measure audiences the same way and partners have different reporting capabilities, so I chose the best available metric for each source and documented those limitations rather than pretending the data was more precise than it was. The framework gave us a much more consistent view of how far our journalism was traveling.

How would you approach The 19th's total journalism reach goal?

I'd first understand exactly how The 19th currently defines total journalism reach, what sources feed it and how those sources are measured. Once I trust the methodology, I'd break the topline apart by source, platform, content type and publishing partner to understand what's driving changes. I'd also pair reach with indicators of deeper relationships where possible, like return behavior or subscriptions. My experience building Spotlight's cross-platform reach framework taught me that distributed data won't always be perfectly comparable, so I'd document those limitations and be clear about what the metric can and can't tell us.

Tell me about a time you had to measure something using incomplete or inconsistent data.

My organizational reach framework at Spotlight includes publishing-partner data, and not every partner has the same reporting capacity. Some can give us detailed numbers while others can only provide more directional estimates. Rather than forcing unlike data to look equivalent, I document those differences and communicate the limitations when I present the results. We still get a useful view of total journalism reach and partner contribution, but leadership understands what is precise and what is directional. My approach is to use the best available information without overstating what it can tell us.

How would you build a reporting system to monitor The 19th's audience goals?

I'd start with the goals and the decisions the reporting needs to support. Then I'd define the metrics, sources and business rules behind each one so we're measuring them consistently. From there I'd determine the right reporting cadence and level of detail for different stakeholders. Leadership may need a concise view of progress against goals, while an audience or product team may need more diagnostic detail. I'd automate recurring pieces where possible, document the methodology and periodically revisit whether the reporting is still answering the questions people actually need answered.

Tell me about a time you identified an audience behavior pattern and turned it into a recommendation.

I analyzed six months of our newsletter and website performance together to understand where there were opportunities to improve readership. I found that Thursday consistently performed well for deeper readership and recirculation, while Tuesday was also strong. I also found that although overall site traffic dipped on Wednesdays, a larger share of visitors came directly from the newsletter, showing strong engagement from existing subscribers. Based on that analysis, I recommended prioritizing larger stories or featuring them prominently in the newsletter on Thursdays, when they had the strongest opportunity for readership and recirculation.

How do you analyze audience behavior beyond topline metrics?

I try to distinguish between measuring how something performed and understanding how the audience actually behaved. Knowing that a story received a certain number of views is useful, but I’m usually interested in how people found it, what they did next and whether they returned.

One example was a five-part reporting series at Spotlight. Rather than just comparing traffic across the five stories, I used GA4 to analyze how readers progressed through the series. I wanted to understand whether people who started with the first installment continued to later parts and where we were losing readers along the way.

So for me, audience behavior analysis is really about moving from “How did this perform?” to “What did people actually do, and what does that tell us about their relationship with our reporting?”

Tell me about a time you turned audience data into an actionable insight for stakeholders.

Our annual audience research showed that Spotlight's existing audience tended to be older, highly educated and particularly engaged with public affairs and policy. That reinforced a larger question we were already asking about how to broaden who we were reaching. The research helped support our decision to invest in a dedicated Arts + Culture initiative as another entry point for people who might not initially come to us for policy coverage. The analysis didn't just describe our current audience; it helped identify an audience gap and informed where we invested in growth.

Tell me about a time you explained data to people with different levels of technical knowledge.

I lead audience research at Spotlight, and rather than presenting teams with demographic tables and percentages, I translated our survey findings into audience personas. That gave people across the organization a shared language for thinking about who we're trying to serve. One persona, Jean, represents parents interested in family-friendly activities, and our First State Fun Day was developed with Jean in mind. That showed me that effective communication isn't about giving people every number. It's helping them understand what the data means, why it matters and what they can do with it.

How would you communicate the same analysis to editorial, product and senior leadership?

I'd start with the same underlying analysis but tailor the presentation around the decisions each group needs to make. Editorial may need to understand what audience behavior means for coverage or commissioning. Product may need more detail about journeys, conversion points or tracking. Leadership may need the larger trend, implications and recommendation. I wouldn't change what the data says, but I would change the level of detail and framing so each stakeholder can understand what matters for their work and what action the analysis supports.

Tell me about your experience conducting audience research.

I lead audience research at Spotlight, including our annual audience survey. I've used that research both to understand who we're currently reaching and to identify gaps in who we're serving. One of the most useful things I've done is translate the findings into audience personas rather than leaving them as demographic tables and percentages. Those personas became a shared language across the organization and influenced actual decisions, including developing First State Fun Day with our family-focused persona Jean in mind. I've also used survey findings to support broader strategic decisions like our Arts + Culture expansion.

How would you approach The 19th's annual audience survey and interview project?

I'd begin with what we need to learn about The 19th's key audience segments and what decisions that research should inform. I'd review what we already know so we're not asking questions just because we've asked them before. I'd use the survey for patterns at scale and interviews to add context and understand why those patterns may exist. Then I'd connect those findings, where appropriate, with behavioral data. Most importantly, I'd translate the research into something teams can actually use rather than letting it end as a presentation.

How would you contribute to The 19th's audience needs model?

First I'd understand how The 19th defines its audience needs and how content is classified. I'd make sure the tagging is consistent enough that we can trust the analysis. Then I'd compare the mix of journalism produced across needs with outcomes like reach, engagement, subscriptions and differences among important audience segments. I wouldn't declare one need "best" because it gets more traffic. I'd look for things like a need with smaller reach but unusually strong loyalty or differences in how segments respond, then bring those patterns back to editorial as commissioning insight rather than a traffic mandate.

How would you use audience data to inform coverage without telling journalists what to cover?

I see audience data as another source of evidence for editorial decision-making, not a replacement for editorial judgment. I'd focus on questions like whether certain journalism is reaching a new audience, generating unusually deep engagement or serving a particular audience need. I'd also look beyond pageviews because the biggest story isn't automatically the most valuable one. My job would be to explain what the audience behavior suggests and give editors useful context while respecting that there are editorial considerations the data can't capture.

Tell me about a time you used audience research to represent audience needs in an organizational decision.

Our annual audience survey showed that our audience skewed older, highly educated and strongly interested in public affairs and policy. That was useful not just as a description of our audience, but because it helped us see who we weren't reaching as well. The findings supported our decision to launch a dedicated Arts + Culture initiative as a new entry point into Spotlight. That's how I think about being the voice of audience needs: bringing evidence about who we're reaching, who may be missing and what audiences tell us into larger organizational decisions.

Tell me about your experience with first-party audience data.

At Spotlight, we create direct audience relationships through newsletter signup forms, our registration wall, events and other acquisition efforts. Once someone becomes a known contact, I'm interested in more than the signup itself. I want to understand acquisition source, what product brought them in, how they engage afterward and whether that relationship deepens over time. My registration-wall work especially reinforced that acquisition volume is only the beginning. First-party data becomes much more useful when we preserve enough context to understand the relationship after acquisition.

How would you expand The 19th's first-party data strategy?

I'd start with what The 19th wants its first-party data to help answer. Then I'd look at where anonymous audiences can become known audiences through newsletters, SMS, events or other products and whether we're preserving enough source and behavioral context to understand how those relationships began. Once someone becomes known, I'd want clean acquisition-source data, consistent engagement definitions and the ability to see whether the relationship deepens. I'd also think carefully about privacy and value exchange. The goal isn't collecting data just because we can; it's responsibly building a better understanding of readers while reducing reliance on third-party data.

Tell me about a time you discovered a metric wasn't actually measuring what you thought it was.

At The Daily Memphian, we used no recorded opens within 90 days to define inactive newsletter subscribers. After a re-engagement campaign, some readers we removed told us they were actually still reading. I investigated and found that some of them had click activity despite having no recorded opens. After researching email privacy changes, I realized opens weren't reliable enough for that decision. For the next campaign, I changed our inactivity definition to no clicks within 90 days. It reinforced that just because a metric is available doesn't mean it's the right metric for the behavior we're trying to measure.

How would you approach data governance and metric definitions?

I think good governance starts with making sure people are using the same definitions and can trust where the data comes from. I'd identify the key metrics teams rely on, document their definitions, source logic and business rules, and establish clear tracking and QA processes. I'd also make that documentation accessible rather than treating governance as something only the data team understands. When definitions or tracking change, I'd want a process for documenting and communicating that so people don't unknowingly compare unlike numbers.

Tell me about a time you improved a reporting or analytics workflow.

Our editorial performance reporting used to require hours of manual preparation before our biweekly meetings. The information I needed lived across WordPress, GA4 and our email platform. I used the WordPress API and Google Apps Script to automate much of the story-level data collection and create a structured dataset, reducing preparation from hours to about 20 minutes. I then connected the data to Looker Studio so editors could access updated performance information themselves. That reduced repetitive work and gave me more time to focus on insights and recommendations.

Tell me about a time you built a system that made data more accessible to other teams.

Our editorial performance workflow is a strong example. Editors needed updated performance information, but the process depended heavily on me manually gathering data before our meetings. I automated much of the underlying collection using the WordPress API and Google Apps Script and connected the structured data to Looker Studio. Editors gained self-service access instead of having to wait for me or for a meeting to answer routine questions. It made the information more useful because it was available when they actually needed it.

How would you approach building an automated dashboard for The 19th?

I'd start with the decisions the dashboard needs to support and who will use it. I wouldn't begin by putting every available metric on a screen. I'd work with stakeholders to identify the questions they regularly need answered, define the metrics clearly and make sure the underlying data is reliable. Then I'd automate the recurring data flow where possible and design the dashboard around usability. I'd also train people on how to interpret it and periodically revisit whether it still reflects their needs.

How would you align audience measurement across partnerships, events, marketing and membership?

I’d first understand what each team is currently measuring, how they define success and where their data lives. Then I’d identify where audience data overlaps so we can create shared definitions and consistent tracking across the organization, while still allowing each team to have program-specific measures. That gives us a more consistent picture of the audience across the organization.

Tell me about a time you worked across teams to solve an audience or data problem.

At The Daily Memphian, I regularly monitored subscription activity in Piano and started noticing patterns that didn’t look like normal subscriber behavior. There were similarities in things like payment activity, geography and email addresses, so I investigated further.

I found that those suspicious records were also flowing into SendGrid and inflating our audience counts, so the issue was affecting our downstream reporting as well.

I documented what I was seeing and worked with our technical team, who then worked with developers to address the underlying issue. Afterward, I continued monitoring the data and confirmed that the abnormal influx had stopped.

What do you do when someone asks you for an analysis?

I start by understanding what they're actually trying to learn or decide, because the metric someone asks for isn't always the best way to answer the underlying question. Then I identify the data and measures that can answer it, validate the analysis and bring the results back to the original decision. For example, our CEO asked me to compare Southern Delaware donor count with the previous year. I realized donor count alone didn't capture the financial impact, so I recommended looking at donor dollars too. Together, the two measures gave us a much more complete answer.

Tell me about a time you realized a stakeholder's original data request didn't fully answer their question.

Our CEO asked me to compare donor count in Southern Delaware with the previous year after our expansion into the region. When I thought about what we were actually trying to understand, donor count alone didn't capture the financial impact of that growth. I recommended looking at total donor dollars as well. I analyzed both measures, and together they gave us a much more complete picture of the expansion's impact. It reinforced for me that part of an analyst's job is helping refine the question, not simply pulling the number someone initially requests.

How would you create a system for fielding and prioritizing analysis requests?

I'd want a lightweight system that captures the question, the decision it supports, the requester, deadline and status so requests don't disappear into messages or meetings. Before accepting the scope, I'd clarify what the person is actually trying to learn and what deliverable would be useful. For prioritization, I'd consider urgency, organizational impact and whether someone else is blocked waiting for the analysis. I'd also look for recurring requests that could eventually become self-service reporting rather than repeatedly doing the same analysis manually.

Tell me about a time you empowered someone to use data independently.

A colleague's role expanded into newsletter production, so I walked them through GA4 using actual questions they would need to answer in their work. We went through traffic acquisition and how to break down traffic by source and medium. I focused not just on how to build the report but on what the metrics meant and how to interpret them.

For example, traffic labeled as email doesn’t automatically mean it came from our newsletter, because other organizations may link to our stories in their emails. I wanted them to understand both how to find the data and what conclusions they could actually draw from it. The goal was for them to be able to answer those questions independently afterward.

How would you train a newsroom with different levels of data experience?

I'd start with the decisions people actually need to make rather than trying to turn everyone into an analyst. I'd use real newsroom questions and teach people which metrics help answer them, what those metrics mean and where their limitations are. I'd also provide documentation or self-service resources they can return to afterward. Success wouldn't be everyone knowing every feature in GA4 or Parse.ly. It would be staff feeling confident answering routine questions themselves and knowing when a question requires deeper analysis.

How would you prevent yourself from becoming the bottleneck for audience data?

I'd separate questions that genuinely require analysis from routine questions people can answer themselves. For recurring needs, I'd invest in dashboards, documentation and training so teams have appropriate self-service access. That's something I've already worked toward at Spotlight with our editorial dashboard and GA4 training. I still want people to come to me for deeper questions, interpretation or unusual analysis, but they shouldn't need an analyst every time they want to understand basic performance.

How would you design and implement an audience data strategy at The 19th?

I'd start with what the organization needs to understand about its audience and what decisions the data needs to support. Then I'd identify the metrics and sources that can answer those questions and make sure we have clear definitions and reliable tracking. From there, I'd build the systems that make the information usable, whether that's recurring reporting, dashboards or audience research. I'd also build in documentation and training so people know where to find the data and how to interpret it. Then I'd continue refining the strategy as the organization's needs evolve.

Walk me through your general process when you’re approaching an analysis.

I always start with the business question and what decision the analysis needs to help us make.

From there, I determine which metrics and data will actually answer that question. Before analyzing it, I clean the data and validate it against the source systems to make sure the numbers make sense.

Then I do the analysis, which might involve comparing groups, aggregating data or looking for patterns over time. Depending on the project, I may already have a hypothesis I’m testing, or the analysis itself may surface something I want to investigate further.

Finally, I present the results and bring it back to the original question: What did we learn, and what should we do next?

How do you communicate data to non-technical stakeholders?

I start with what the data means for the person I’m talking to rather than walking them through the technical details. I give an overview of what the data is telling us, the key insight, and then the practical takeaway.

I used to communicate data by spending a lot more time on the technical details and how I got to the answer. Over time, I’ve learned that most people don’t need all of that. They need to understand what the data is telling us, why it matters and what they can do with it. I’ve also learned how important visuals can be in making data more accessible and memorable, especially when I’m communicating with people who aren’t analysts.

What does it mean to you to use data effectively in a newsroom?

I think using data effectively in a newsroom means giving journalists information that helps them make better decisions.

That requires understanding the question an editor is actually trying to answer. Sometimes the useful insight might be that a particular story is reaching a new audience or that readers are showing unusually strong interest in a topic. 

I also think you have to understand newsroom workflows. Editors are making decisions quickly, and the data isn’t always going to exist in a perfectly structured dataset. Part of my job is figuring out how to make the analysis useful within those realities.

I’ve run into that at Spotlight with editorial performance analysis. We needed story-level editorial information that wasn’t readily available in our analytics data, so I used the WordPress API to create a structured story dataset that could eventually be connected with performance data.

Ultimately, I want data to give journalists another source of evidence for making decisions

How would you help people across a newsroom use data to inform their decisions?

I’d start with the decisions people actually need to make rather than trying to teach everyone to become an analyst. The goal is to give them enough understanding and access to use data confidently in their own work.

At Spotlight, I played a major role in defining what we needed from our editorial analytics dashboard and worked with our technology contractor to get it built. That gave people across the organization self-service access to the data instead of having to come to me every time they had a question.

I also want people to understand what the metrics can and can’t tell them. For example, editors naturally pay a lot of attention to pageviews, but the story with the most pageviews isn’t necessarily the story readers were most engaged with. Another story might have fewer pageviews but much stronger engaged time or a higher newsletter signup rate.

So I’d focus on helping people understand which metrics are useful for the decisions they’re trying to make and how to interpret them. Success to me is people becoming comfortable enough with the data to use it when it can help them make a decision.