Beginnings of stories are very difficult. I think it's partly because of the way I write - I like to put together a puzzle that comes together into a certain type of constellation at the end. This means I usually know what I want the end result to look like - but it also requires that I start all the beginning pieces in the right place so they will progress into that final desired relation.
I wrote the opening chapter of For Love, For Power probably upwards of six times. The same thing happens pretty often with short stories. I spend a lot of time walking around with different kinds of beginnings rattling around in my head, until insight strikes and I can see the optimal starting configuration. Sometimes, as with my current novel Work-In-Progress, I write the opening in one way, and then before I get too far out of the starting gate I realize it's not right. In the case of Secrets That Bind, I realized that the chapter I'd written as chapter 1 could not be chapter 1. I therefore promoted what had been chapter 2 to the beginning, and took that original chapter 1 and integrated it with the chapter that had been chapter 3.
Let me get concrete. I have two undercaste characters in Secrets, Meetis and Corbinan. Each one has a very particular role. Corbinan is the person who first stumbles into the big conflict. Meetis is the one who drives the solution to that conflict. In my original draft, I thought, "jump into the main conflict as soon as possible." It's often an approach that works well. In this case, however, it was not working, for two reasons. First, Corbinan is not the solution-driver. He stumbles in and is swallowed by the conflict - in a sense, he's the "damsel," though he does act on his own behalf in ways that damsels don't. Meetis is a person in trouble, but she's all about the solutions, and the solution to her first (seemingly unrelated) problem soon delivers her into an identifiable relation with the main conflict, so readers can anticipate her entry into it. Second, Corbinan speaks in undercaste dialect. So does Meetis, but her language is far more accessible (because she's younger, and because she's someone who is good at placating Highers). By encountering Meetis first, we're able to get a more gradual entry into the language, and learn the context surrounding the way Corbinan talks, so that he's less of a jolt when we encounter him.
This is one of the reasons why I like to have an "entry" beta-reader. I'll write the first chapter of each point of view character, and then run it by someone who can tell me if they are feeling very confused, and which characters they relate to most easily. It's a kind of critical orientation to the story that allows me to make sure the jumping-off point is solid. Jumping off stone is far more effective than jumping off sand.
When I write short stories, I like to think around until I find a first sentence that gets me jazzed to jump in. I'm willing to spend concerted time on this, even up to a few days. I've gotten better at it as time has gone by.
Now, however, I am about to attempt two short story sequels, and I find I'm hesitating (in the case of one of the two, this "hesitation" has amounted to several years of not being able to start). The crux of the problem is bridging - creating an opening that not only sets up the ending properly, but also evokes the first story for those who have read about it, and provides enough backstory for those who haven't read it, without bogging down the current story.
I think I need to start thinking about these stories a little bit the way I think about novel openings - that is to say, to enter in a place where I feel familiar with the world, and excited about the plot, and worried about other people's entry later. Backstory can be added. Revisions are possible - and more than that, they are inevitable. What better opportunity to create a bridge between two islands than when both ends of it can be anchored on solid ground?
Wish me luck.