"Clarity" is the mantra for many science communicators. Science must be made clear to the masses.
Clarity is also the greatest challenge for the field as a whole - not clarity of content, but clarity of goals and metrics. In other words, we know what science communication is, but we can't clearly articulate why we do it - and we have absolutely no idea if we're doing it right. This is the source of so much frustration among many science communicators.
First, let's be clear about something: if your primary interest is simply writing great science articles, you already know your audience. It's you. There's clearly nothing wrong with this, but this blog post isn't about you. Here, go read about radio waves in space or something.
If your goal is to communicate science for a reason beyond enjoyment there are some things to keep in mind:
1. Know your audience. This is the first rule of communication. Identify who you want to reach, and be as specific as possible about it. The more specifically you define your audience, the more likely you are to know how to influence it. Too many science communicators think "the public" is an audience. It's usually not. Maybe your real audience is state legislators in North Carolina who criminalized sea level rise, or moms who have questions about GMO's, or farmers who see antibiotics as the difference between a profit and a loss.
2. If you have a reason, you should have a goal. The easiest way to know if you've influenced an audience is to ask it to do something - and this is where goals and metrics become real. If your goal is "soft," such as general awareness or information, ask your readers to share your article, and count the number of social shares. If your goal is policy oriented, put a link in your piece to the email address of a policy maker and count the number of outclicks. If your goal is giving parents the information they need to make smart decisions for feeding their families, put together an infographic or a .pdf fact sheet and count the number of times it is downloaded.
3. Replicate the results. Keep track of how many times your audience did what you asked of it, and the approach you took. Learn from it. Try different methods. If this sounds familiar to scientists... well, that's kind of the point.