Once the pages are illustrated, the words underneath do a specific job — and most people get it wrong in one of two directions.
The two failure modes
Describing the picture. The illustration shows you both on a beach; the caption says "us at the beach." The reader now has the same information twice and nothing new.
Writing a paragraph. Three or four sentences of context under every page turns a storybook into an annotated album. The pictures stop carrying the story.
A good caption sits between: short, and about something the picture cannot show.
What a caption is for
The illustration handles what it looked like. The caption handles what it meant, what happened just before, or what somebody said.
Compare:
Our first holiday together.
with:
You had planned every hour of this trip and then we missed the train on day one.
Both sit under the same picture. The first labels it; the second is a story, and it tells you something about the person the illustration cannot.
Length
One or two sentences. Occasionally three if the moment genuinely needs it.
The test: read the page with the caption covered, then uncovered. If the caption adds nothing you did not already have, cut it or replace it.
Four things that reliably work
A line of dialogue. What one of you actually said. Even mundane speech dates and places a moment better than description does.
"Five more minutes," you said, for the fourth time.
The thing that went wrong. Friction makes a moment specific; smooth days blur.
The reservation was for the wrong Tuesday.
The detail nobody else would notice. Small, true, unglamorous.
You wore my jumper the whole weekend and never mentioned it.
The consequence. What this moment led to.
We have argued about that route every year since.
What to avoid
Dates as captions. "June 2019" tells a reader nothing. If the date matters, say why.
Generic affection. "You mean everything to me." True, but it could sit under any page in any book. Attach the feeling to a fact.
Explaining the joke. If a page illustrates an inside joke, the caption should not unpack it. The one reader who matters already knows.
Present-tense narration. "Here we are at the lake." Captions read better as the memory itself than as a tour guide.
Keeping a consistent voice
Decide early who is speaking. Options that work:
- One person to the other — "you" throughout. Most common, and the warmest.
- Both of you together — "we". Good for a shared keepsake rather than a gift.
- Neutral narration — rarer, but works for books meant for a wider family audience.
Whichever you pick, keep it for the whole book. Switching between "we" and "you" partway through is the most common thing that makes an otherwise good book feel disjointed.
Captions that carry the story forward
In a book of ten to twenty pages, the captions collectively are the narrative. Read them in order with the pictures covered — they should still make sense as a sequence.
If they read as a list of unconnected labels, the book will feel like an album. A few small connective phrases fix this: "by then," "the year after that," "which is how we ended up."
Where to start
Write captions last, once you can see the pages. It is much easier to write one true sentence under an illustration that exists than to imagine one in advance.
Then start your book, or read how to write your how-we-met story for the opening page specifically.
FAQ
How long should a caption be?
One or two sentences. If the caption only describes what the illustration already shows, cut it or replace it with something the picture cannot say.
Should captions include dates?
Only when the date does work — an anniversary, a day that matters. "June 2019" on its own tells a reader nothing.
Should I write "we" or "you"?
Either, but pick one and keep it for the whole book. Switching partway is the most common thing that makes a book feel disjointed.
When should I write the captions?
Last, once the pages are illustrated. Writing one true sentence under a picture you can see is much easier than imagining it beforehand.
What if I cannot think of anything to say?
Use a line of dialogue, or the thing that went wrong that day. Both produce more specific captions than trying to summarise a feeling.