Read for structure before content
Before reading a long document line by line, spend two or three minutes scanning headings, subheadings, and any summary or conclusion section. This builds a map of the argument before you invest time in detailed reading, and it means unfamiliar vocabulary later in the document is easier to guess from context because you already know roughly what the section is about.
Set a comprehension target, not a translation target
Trying to understand every single word of a technical document slows reading dramatically and is often unnecessary. Decide in advance what level of comprehension you actually need: can you summarise each section's main point, or do you need to extract specific numbers and decisions? Matching your reading depth to the actual task saves significant time on long documents.
Flag instead of stopping
Stopping at every unfamiliar word breaks comprehension of the sentence and paragraph around it. Instead, mark unfamiliar words lightly, whether with a highlight, a saved-word tool, or a quick note, and keep reading to the end of the paragraph or section. Often the meaning becomes clear from later context, and you can review the flagged words afterward as a batch.
Review flagged vocabulary by document, not randomly
After finishing a document, review the words you flagged together, since they usually share a topic and register, such as financial terms or regulatory language. Grouped review is more efficient than mixing them into a general vocabulary list, because the shared context helps you remember not just the meaning but how the words are typically used together.