We hook a reader when we actively engage her in making predictions.
Once, I would have said the answer is suspense, but that begins too late. To explain why, let me reference that quote from Hitchcock which explains the difference between suspense and surprise. He begins with us seeing a bomb beneath the table, but for our purposes, that’s where we stop. We see the bomb and that provokes a question. Will the people at the table discover the bomb before it’s too late? The reader, even if unintentionally, answers that question and then reads to see if she’s right. She makes a prediction, and, like a man betting on a greyhound, she’s invested in the outcome. Everything Hitchcock talked about was the suspense between the prediction and the reveal, but it begins with provoking the prediction.
The investment is gone once the reveal is made. We’ve been screaming at Hitchcock’s guests, but after the bomb goes off, then what? The reader made her bet, and at the end of the race, win or lose, the investment in the outcome is gone. We either catch her with the next race, give her new dogs to bet on, or she’s going home.
It’s prediction after prediction after prediction, and our story must start with the first race and end with last.
As E. M. Forrester said, “[The story] can only have one fault: that of making the audience not want to know what happens next.” If the reader doesn’t want to know, it’s because she’s not invested, and that happens for a few particular reasons. There’s no intriguing question to provoke a prediction; the question’s been answered without a new question to take its place; or the story has presented a series of questions, but their answers were too obvious.
When we compare good and bad action films, we see a variation of this. In Die Hard, we get a wide shot of John McClane and various villains on the roof before they come in contact with one another. We understand the geography, the threat, and the characters involved. Our minds anticipate the danger and predict possible outcomes. In bad action films, we’re not sure where the characters are. Bad guys jump out of nowhere and throw a few punches before being dispatched. The viewer isn’t actively engaged in the set up of the action scene.
Staging such a shot seems unrelated to writing stories that hold a reader’s interest, but the underlying principle is the same because that roof-top shot was telling its own story in a few seconds. Visually, it did exactly what we’re talking about here, setting up characters and stakes and asking a question the viewer must answer.
On a macro scale, survival is a possible question, but on a micro scale, the possibilities are limited. Maybe McClane gets captured. Maybe he gets injured or someone else dies. Maybe he escapes. In each scenario, the how becomes more important than the what. A surprising escape we never anticipated, that thrills us, but remember, that thrill is short lived. When McClane ties himself to a fire hose and jumps, that’s a surprising solution, but then he’s stuck on the side of the building. He shoots his way through the glass, but then the hose wheel disconnects and what had been his life line now threatens to drag him to his doom.
It’s a series of quick predictions, and as the last one ends, it leads into another where McClane is ambushed by a terrorist hungry for revenge. In an action film, the questions posed generally involve action, but the principles apply to all storytelling. In My Dinner with Andre, the questions posed are centered on the conversation. For one movie, the ultimate stakes are survival for McClane and his estranged wife. For the other, the ultimate stakes are a philosophically-imposed social alienation and an escape from the captivity of the mind. Throughout each film, questions are raised for the viewer to answer, small questions which resonate with the story’s big question. It’s the same for your story. It’s the same for mine.
Is calling it a prediction meaningful compared to the old phrase of “wanting to know what happens next?” Yes, it’s different in an instructive way. It tells us that our writing must present enough information, and withhold enough information, to compel a reader to predict an outcome for the quantum of story they’re reading.
I can think of 007 reasons not to call it a quantum of story, but a quantum is a small, discrete package of information. In our usage, it might be a sequence of events or it might be a single moment within that sequence. It fits.
Our focus on prediction reveals that this is the mechanism that sustains interest and keeps the reader invested. If the reader doesn’t have enough information to propose a possible outcome, that quantum needs a rewrite. Tell your story in such a way that the reader has an emotional or logical understanding of what’s at stake for whom, and psychology will drive her to predict how the problem might be resolved. We fail to ask a question when our quantum is vague or confusing. A reader can’t predict an outcome if they don’t understand what’s happening.
If you’ve asked yourself why successful stories fail in so many ways, I’ll predict that provoking predictions is the answer. Mind you, a reader’s investment is still connected to personal taste. Taste determines whether the question we ask inspires any particular reader to make meaningful predictions. If the story’s not a good fit for a reader, there’s nothing to be done. We shake hands and move on, but if the story and the reader share the same tastes, prediction is everything.
— Thaddeus Thomas


Another great, brief explanation of what it takes to keep the reader reading! However, I do think it's different for movies, which must keep the watcher watching. I'm in the process of re-watching the British police procedural/detective series Morse ---having seen all 33 episodes "cold." After researching John Thaw and Kevin Whatley, and having seen the how-it-came-to-be (it's from a series of books by Colin Dexter) and interviews with Whatley post-mortem of Thaw's death, I'm now watching it for analytical purposes. The Morse series is not like The Sweeny, a 70's era Thaw vehicle that includes "action" sequences a-la Streets of San Francisco and Mod Squad, and most Americans probably now couldn't stay engaged for the 2-hour! format.
I think it must be said the RULES of writing or film making depends both on the form of story delivery and on the audience. Morse is beautifully introduced by romantic scenes of English life and scenery before the crime takes place. It works for me because I happen to love scenes of English history. Show London now, and I'll vomit because it no longer reflects that kind of historical beauty. I digress.
"Hooks" will vary according to genre. There are those who don't like Hitchcock films or or action films or literary writing because of their hooks alone. One person's hook is another's hokum. The problem is, as I see it, is that the reader and the watcher can not only make accurate prediction,
s/he can recite the dialog before the words are spoken. Prediction has become predictability.