How we built a newsletter dashboard in two hours — and the lessons we learned
The Economist set up a dedicated newsletters team in September 2017 as part of a broader push to improve our owned-and-operated products and to build a more direct relationship with readers. Since then, we have redesigned our flagship daily and weekly newsletters so they are not merely megaphones for our journalism, but stand-alone products that aim to meet readers’ needs.
Behind the scenes, we have also improved the tools and processes that help us make more informed decisions. One of them is an internal newsletter data dashboard called NED (Newsletter Engagement Dashboard), which we built in two hours in March and spent the next two months improving.
Here’s what we learned.
Move fast
Our email service provider (ESP) is used by a handful of people who are directly responsible for producing newsletters. This meant that many colleagues did not have access to the relevant data that live in the ESP. So we decided to build a dashboard that is connected to the ESP and easily accessible to team members.
Spelling out this immediate goal gave us a clear cutoff. We knew the first iteration of the dashboard could be deployed as soon as the data were transferred from the ESP to a shareable and configurable platform. We also stuck to the metrics that are readily available: open rates, click-through rates, referral traffic, and performance by position. And we left room for additions.
Using Google’s Data Studio, exporting and customisation took about two hours. A link to the dashboard landed in the inbox of a small group of guinea pigs the next day.
Shadow users
We then hovered over the shoulders of a few editors and observed how they interacted with the dashboard. We discovered that some found the setup with multiple modules disorientating. This prompted us to spend the next few weeks improving the design and the placement of the lead module to make navigation easier. It now contains a weekly summary of newsletter performance, followed by various benchmarks for added context.
We also noticed that the mobile version of the dashboard was fiddly, so we introduced a PDF snapshot which is now attached to the weekly email to our colleagues. Over the next few weeks, we will add signposting to each module of the dashboard, so we don’t simply present the numbers, but provide more context on how the numbers are measured and why we include them. The easier to use we make the dashboard, the more insights our colleagues can extract from it.
If a dashboard answers all your questions, you’re asking the wrong questions
While our current metrics paint a decent picture of individual newsletter or story performance, we need to dig deeper to find out whether readers are getting real value from our newsletters.
For example, once readers click through to an article and land on our website or app, how much time do they spend reading the article? How often do they come back to the newsletter to find another interesting piece? How likely are they to subscribe to The Economist and renew after reading our newsletters? And the big question for the team is: how do we know whether we are fulfilling the product vision of using newsletters to help readers discover the best of our journalism?
To answer these questions, we need to go beyond the dashboard and look at the reader journeys. After all, even though our teams are structured around individual digital products and editorial formats, what matters to readers is a frictionless, end-to-end customer experience.
So while the immediate goal of building NED is to democratise data internally, we hope to use the dashboard to spark conversations between various teams to better understand reader journeys and behaviours. A successful digital product that truly serves readers isn’t built on data alone; fostering a culture in the newsroom to examine data alongside editorial judgement and user research is crucial. If a dashboard can answer all your questions, you are probably asking the wrong questions.
Since the launch of NED, we have increased the frequency of regular catch-ups between the data team and the core newsletters team. We also hosted a “data deep dive” with a wider group of colleagues, during which we connected the dots between data from different products and compared data with research. NED isn’t designed to drive our editorial and product decisions. Rather, it informs the tools and processes we use to make decisions.
The dashboard is an output. A better experience for readers is the outcome
Despite the brightly-coloured charts and the addictive interactives, a dashboard is simply a collection of numbers. Building a dashboard is not the end goal; turning data into actionable insights that lead to a better experience for our readers is. With iterative improvements, we hope NED eventually becomes a destination that our colleagues visit for all things newsletter data-related. But when it comes to building a better reading experience across all of our digital products, this dashboard is just the start.
Sunnie Huang is the newsletters editor
Molly Devine is the editorial data analyst
Sam Li is the product data analyst
