Sourcing
Primary source
The original document: a filing, a standard, an official help page, a published study.
Citing a blog that cites a study is not sourcing. When the chain breaks, and it does, the mistake is yours to defend.
What editors say to each other, in plain terms, and what each one changes about the work.
One page written before anybody starts: the question, the audience, the figure the piece must find and what you measure afterwards.
Most bad drafts are not wrong, they are wide, because nobody said what was out of scope. One line naming what the piece does not cover saves a full rewrite.
The sentence in a brief naming what the piece deliberately will not cover.
Cheapest line in the document. Without it the writer covers everything adjacent and the reviewer has no ground to object from.
The person who decides a piece exists and signs off what it must contain.
When nobody holds this role, topics arrive by whoever spoke last in the meeting and no piece can be judged against anything.
Finding what is already published and verifiable before writing a word.
Forty minutes here regularly kills a topic. That is a good outcome, not wasted time, and it is why the research happens before the outline.
The original document: a filing, a standard, an official help page, a published study.
Citing a blog that cites a study is not sourcing. When the chain breaks, and it does, the mistake is yours to defend.
A number you produced yourself: counted, calculated or collected.
It is the only thing in a piece that cannot be found elsewhere, and the only part other people cite. A piece without one is a summary.
The named source printed next to a number, with its year.
Anything you cannot attribute gets cut, however good it sounds. This rule is what makes the rest of the piece believable.
The short paragraph under the headline that says why the piece exists.
It is where most readers decide. Written last, because you do not know the argument until the piece is finished.
The paragraph, usually third or fourth, stating the point of the whole piece in plain terms.
If a reader stops after it they should still have the answer. Hiding the point until the end is a habit from essays, not from journalism.
The final line, chosen to land rather than to summarise.
A summary ending tells the reader the piece is over and teaches them nothing. An unresolved detail does more work.
Payment per word, the most common freelance arrangement.
It pays for length, which is the opposite of what an editor wants. Day rates and per-piece fees align better with cutting.
A separate pass where every number and name is verified against its source.
Done by someone who did not write the piece, or it is not a check. Twenty minutes, and it is the difference between a mistake and a retraction.
The document settling recurring choices: dates, numerals, capitals, terminology.
Its real value is ending repeated arguments, not consistency. Two pages that get followed beat forty that do not.
The plan of what gets published, when, by whom, and against what measure.
Two columns are usually missing: the review date and the number the piece is judged on. Without them the calendar is a list of deadlines.
The date, set at publication, when somebody looks at the piece again.
A page with no review date decays quietly for two years and then gets blamed on the algorithm.
The full list of what you have already published, with traffic and links against each.
Most teams do not have one, which is why they publish a fourth page on a question they already half answer three times.
Combining two thin pieces on the same question into one stronger URL and redirecting the loser.
Pick the survivor by links pointing at it, not by which one you prefer. Redirect rather than delete, so what it earned is kept.
Updating a piece with new information and saying what changed.
Changing the date and reordering paragraphs is not a refresh, and readers notice within a screen.
Deleting pages with no traffic, no links and no purpose.
Not neutral to keep them: they spend crawl attention and blur what the site is about.
Where a piece goes after publish, on which dates.
Deciding this after the piece exists means it was written for nothing in particular. Put it in the brief and it changes the draft.
The planned reappearance of a piece weeks after publication.
Almost nobody sees a post on the day it goes out. Three dates beat one push every time.
Republishing a piece on another platform, usually with a canonical back to yours.
Without the canonical you are competing with your own page, and the bigger domain wins.
Whether the person finished the piece able to do something they could not before.
Harder to count than traffic and closer to the thing you want. Replies and questions are the cheapest proxy.
Everything a piece cost: writing, editing, research, design, and the page it lands on.
Teams that do not know this number cannot tell an expensive failure from a cheap one, so they repeat both.
A draft, a published page, or the topic you are stuck on. We reply with what to cut and what is missing. No call.