Research by Erin Provey at Sirius Decisions shows companies typically spend more than 50 percent of their marketing budgets on content creation, yet organizations rarely have a detailed plan for how to manage that spend. Too often teams focus on how to publish, run campaigns, and measure ranking, clickthroughs, conversion, and so on … all the while assuming the content to feed all this activity just magically appears.

But where does the content come from? The trend in the industry is to pull genuine voices from all over the organization, such as product teams and subjectmatter experts; and companies scale content efforts by buying it from agencies or content marketing platforms. You have enough content. How do you know if it’s good enough?

One company I work with has over 100 (one hundred!) agencies creating content for it in addition to internal content teams spread across the organization. Think about this from the customer’s perspective: What are the chances the customer will get a high-quality, consistent customer experience when there are so many disconnected moving parts? How can you know before you publish whether your content is going to be effective?

One traditional way to address these challenges is to develop written branding and style guidelines for content. These often include bold exhortations to be “friendly, approachable and responsive,” or “sincere and human,” or even “expressive.” These are great high-level statements of intent, but how, concretely, can you use these ideas to decide how to write? How can I make them actionable? And how can you be sure everyone has a common understanding of how to write in a way that reflects these goals at scale?

Many organizations devote significant effort to defining personas for their content, but here again the same issues remain: how exactly do I write differently to engage a CIO versus a CMO?

Well … it’s really hard. You can’t do it just by publishing a style guide on the intranet and hoping everyone reads it. You can’t just train people (although training is a good start), since people forget their training, they move on—and you can’t be training people all the time.

Fortunately, as in every other area of marketing, AI can help. The latest advances in artificial intelligence make it possible for machines to read content and provide insights on brand and tone of voice, which allow you to scale your content operations while still staying on brand and on target.

Three things characterize an effective approach to content: goals, guidance and governance. We’ll look at each of these in turn and explain how they work together to make your teams successful.

Setting Goals

Like any other plan, nothing gives your content-creation strategy a better chance of success than writing it down. But in doing so, make sure the goals you document are actionable. It’s fine to start with high-level aspirations like “be human” … but you have to go beyond that.

The best guides I’ve seen go into rich detail about the audience. They make clear there may be different target audiences you want to reach with your content, and they map the buyer’s journey for each segment. Good guides also capture nuances in tone to adopt in different types of content; you want to “speak with one voice” but, especially in B2B communication, you generally want blogs and social content to be lighter and more conversational than technical white papers. Even in a B2C environment, you probably want to take a slightly different tone within different areas of your content, depending on the context.

Delivering Guidance

Once you set goals, the next challenge is to get the team to follow them. If your strategy is to send people a link to the style guide (or worse an attachment), then good luck.

It won’t work. It won’t work because the guidance isn’t delivered in bite-sized chunks when and where content creators need it. As writers, we need help on all aspects of what makes content effective: findability, precision, branding, tone, word choice—but only one piece at a time. And we also need it right where we’re writing because we aren’t going to go somewhere else to look for it.

It won’t work because it’s passive—writer or editor has to decide to go and look something up.

And it won’t work because things change. Updating a PDF version isn’t going to make a difference if everyone has the old PDF version on paper at their desks. Even a searchable web resource that content creators refer to isn’t viable because they still need to follow it, learn it all, and apply it every day. That’s a big ask.

There are generic apps that can help with guidance, such as Flesch-Kincaid, the new editor feature in Word or the lightweight Hemingway version. But these will only go so far since they are by their shrink-wrapped nature not designed to give advice on the intricacies of your brand and story—nor how to choose the right language to engage your target audience.

This is why Acrolinx has worked hard to fill that gap. We focus on using our AI engine to provide you with expert guidance aligned with your goals—in nice bite-sized chunks, right in your editor of choice. It’s like having your best editor friend with you every step of the way.

Instilling Governance

The third piece of the puzzle is governance. Governance means lots of things to different people, but it boils down to knowing what content exists, and making sure it’s doing what it’s supposed to. The most popular content quality analytics look at the past performance of published content—but relying on post-publishing analytics is like looking in the rearview mirror.

The scoring and analytics you access before publication show not just a generic metric of content quality, but whether the content is really fit for the intended purpose. You have to calibrate the scoring so it correlates with your post-publication scores, but once that is done you can see the future—you have a leading indicator for success.

There are great tools for measuring likely performance—Google Analytics or even generic scoring on readability—but it is hard to make these actionable without going a step further. Making insights actionable means getting them to writers and editors, and giving them guidance on how to fix the content. As they create ever more content, at ever increasing velocity, it is critical to be able to adapt and stay agile with the content rather than knowing what just happened.

This post was originally published on Chief Content Officer (CCO).