Since late February, I’ve been feeding the bullet points from every day’s Lunchtime Espresso email into a detailed spreadsheet. The goal of this project is to determine what exactly makes a good midday update email, and to turn
The data points I’m tracking are:
- The actual content of each bullet point. Statistically the least useful column, but potentially useful for corpus analysis.
- The exact date on which the bullet point was published.
- One lead topic, and then a number of additional topics from the same list. These are usually broad categories like ‘British Politics’, ‘Business’ or ‘Culture’.
- Some tags, which are essentially low-volume topics. A tag might be as specific as ‘Starmer in China’ or as general as ‘Immigration’.
- Each Lunchtime Espresso bullet point links to at least one other source. Some of these are news articles, from outlets like the BBC or Bloomberg. Others are social media posts, liveblogs etc. I’m also saving the actual links, which are not statistically useful but important for record-keeping.
- The time the linked story was published. This is