By Josiah Fisk
It would only be logical to assume that the simplification of all-text documents should begin by focusing on the language. Logical, but not the way to get the best results—or so I believe and hope to show.
What I suggest instead is a process that is organic and evolutionary. It resembles the solving of a puzzle without any image of the finished product to consult along the way. In this process, there is both a constant proposing of ideas and a constant questioning of ideas. Rather than a traditional step-by-step process, it’s more iterative—a matter of gradually morphing the raw content into something that works.
In this process, as we practice it, the design and the writing aspects are highly integrated and always in dialogue (whether that dialogue is between individuals or inside a team member’s head). The creative process is also in constant dialogue with the needs and constraints of the project, most of which are identified in the project’s discovery phase. Visual solutions are actively sought and generally preferred. Everything remains fluid and conditional for as long as possible. The actual writing of text is virtually the last step, taken only when everything else that can be done is done.
To make this organic process even more powerful, we use an expanded skill set. Along with writing, we include information design, marketing, and advertising (both copywriting and art direction). Why marketing and advertising? Because even though we generally don’t want our finished products to look or read anything like marketing documents, some of the tools that marketing and advertising professionals use can be extremely valuable in figuring out what resonates with readers.
The Case Study: Simplifying a Workplace Policy
Large financial services firms are notorious for the volume and complexity of their employee policies. In addition to the normal HR policies, there can be more than 100 pages of disjointed and not always logical compliance and information-security policies. These may cover anything from procedures about how to anchor a company-provided laptop to policies about a spouse’s personal stock trades.
The policies are required reading—literally. When you’re hired, you sign a statement saying you’ve read and understood the policies. Yet it’s widely acknowledged that almost no one actually reads these policies, simply because they are basically unreadable.
The traditional response to this state of affairs is not to fix the policy, but to beef up employee training. This works well for raising awareness of the main points of policy and of the possible consequences of violations (for the employee, being fired; for the firm, millions or even billions in penalties and lost business). The problem comes when a real-life circumstance prompts an employee to check the policy to see exactly what it says. Instead of a ready reference source, they find a thicket of legalese, with neither language nor structure that corresponds to what they saw in the training.
Recognizing the inefficiencies and risks of this situation, our client, a large asset manager, asked us to make their policies as simple as possible. (At the client’s request, we have substituted an imaginary brand name and identity, but the documents are otherwise identical.)
In addition to making the policies look and read better, the target was to make them about 20% shorter. In the case of the policy discussed here (the first one we worked on), we were able to make the document 47% shorter, measured by word count (79% shorter by page count), while also significantly improving standard readability metrics (see Figure 1). We find that this degree of simplification is rarely possible with editing alone. It requires a complete reconception of the document.
Step 1: Finding Out What’s Required and What’s Desired
With this project we began, as we usually do, with a fairly non-directed discovery phase. We reviewed most of the client’s existing policies, as a set and individually, and we asked a lot of questions. These ranged from legal and regulatory concerns to employee perceptions to document production and maintenance. Our goal was to get a clear picture of every stakeholder’s needs and preferences. We were also concerned with the process by which these documents would be managed. We sought to uncover any latent issues that might emerge later on and disrupt the course of development.
Once we had most of the content and constraints on the table, we began mentally playing out the various scenarios in which employees were likely to encounter the policy. There are a number of ways this can be done, but the most effective, in our experience, is how it’s typically done at ad agencies:
- Look at the relevant demographic research and other information about the target audience(s).
- Forget about the research (consciously, that is) and, almost as a writer or actor would, place yourself inside the motivations, fears, beliefs, preoccupations, and other psychological and emotional conditions of typical target audience members.
- Go back to your data or other resources (including talking with people who represent target audience members) and do a reality check.
If this sounds unscientific and imprecise, it is. Yet it works remarkably well (as generations of Don Drapers can tell you). This simple, sloppy heuristic lets you reach into the vast sea of information and inferences about your target audience and fish out the tiny handful of factors that actually matter.
The reason it works is that those factors tend to be situational rather than demographic. That is, they’re about the user’s likely need (“what’s that vendor gift rule say again?”), the user’s mood of the moment (“ugh, I don’t have time for this”), as well as the user’s general attitudes (“I don’t really know why these rules are here; they seem like busywork”).
A mid-career IT professional, a veteran portfolio manager, and a newly hired admin can all reasonably be expected to have the reactions described above. They’re basic human reactions.
A word of caution: this is a skill that takes some practice to learn. It’s a bit trickier than using personas—more about situational dynamics and the search for widely shared psychological ground, less about the attributes of any given persona. If you don’t have ad agency copywriting experience, you might want to consider adding someone to your team who does, even just for this phase of the project.
Step 2: Finding the Right Structure (Or Letting It Find You)
Structure has more effect on comprehension than any other single factor. Research in information design supports this point, and our own experience strongly confirms it. It’s not hard to understand why: imagine taking a chapter from a favorite novel and reading the paragraphs in random order. Not a word of the original has been altered, yet the result will be pure chaos.
With legal documents that respond to a particular regulation, the structure often follows that of the regulation, point by point. That’s perfectly sensible from the drafter’s point of view, but usually a disaster for the reader, because regulations tend to be organized around intellectual concepts rather than around the experience of complying with the regulation.
At More Carrot, we tend to spend a lot of time on structure, actually “auditioning” a lot of different possible structures and variants to see what works best. In the end, the right structure often seems to choose itself, so to speak. If we look closely at all the needs and constraints of the various stakeholders, the nature of the content, and the various usage scenarios, one or two commonly used basic structures (such as progressive instructions, segmentation by user, narrative with sidebars, or index) clearly stand out as being better adapted to the circumstances at hand than the others.
When evaluating structures, it’s good to remember that the number of logical choices is always much greater than the number of helpful choices. For instance, it’s not uncommon to find documents that are divided into sections according to which department controls the content. That’s logical from the standpoint of the different departments—each one gets its own dedicated “real estate”—but from the user standpoint, it amounts to being handed a kit of parts and having to figure out how it all goes together.
Partly because of this, we actively go looking for constraints, because they can help us sort out ideas that may look appealing but in fact are impractical. That’s why, in this case, although the initial assignment was to simplify a single policy, we went ahead and looked at more than a dozen other policies. We wanted to be sure that we wouldn’t be boxing ourselves in if the client decided to ask us to simplify all of their polices (as they eventually did).
Working together, the writing and design members of the team settled on a fixed, linear architecture for the basic structure: an introductory page, always with the same three subsections, followed by one or more sections that always contained numbered rules. We also created two styles of sidebars (to handle content that logically belongs outside the flow of the main narrative) and a table style for when a table is the best presentation method.
Simple as this setup is, it has enabled us to find sound solutions to all of the communication challenges and information configurations we have encountered to date (seven policies and counting). In general, we find that close collaboration between the design and writing disciplines in the structure phase greatly increases the odds of coming up with a structure that is simple, effective, and not in need of constant amendment or exceptions.
Step 3: Developing a Design that Engages Users without Calling Attention to Itself
Looking at the design of the simplified policy, nobody is likely to find it terribly exciting (see Figure 2). But that in itself is part of the design. Policies are not there to entertain. Their job is to help the user discharge an obligation as quickly and painlessly as possible. An essential part of that is making every effort to keep attention focused on the content and not on the document itself.
This is the main difference between graphic design and information design. Effective graphic designs tend to be those that attract attention, set a mood, create an impact. Information design uses the same tools but for nearly opposite goals.
Nearly every visual characteristic of the design of our revised policy—color, tint, size, shape, orientation, proximity—is doing work (in addition to adding a degree of general visual appeal). Most of the work is navigational, either at the macro level or the local level. Some involves giving emphasis. Some involves both.
Even the logo is doing work: it’s part of bringing the policy into conformance with brand standards. As minimal as the branding is, it’s enough to make the document look more professional, more organized, and more approachable. It’s the difference between a document you feel you can use and one you can’t—which in the world of policies is the difference that matters most.
This may seem like a minor point, but it’s actually both crucial and poorly understood. One of the most frequent missteps we see in informational and transactional documents comes from equating “more professional” with “more like a marketing piece,” whether in language or design or both. In informational and transaction documents, a professional yet plain look is not a sign that the designer was having an off day. It’s an intelligent and effective reflection of user-centered goals and priorities.
Step 4: Developing Local Structures within the Larger Structure
Once the general structure of the document was determined, we moved on to looking at how to organize the content within that structure. We began by looking for the overlaps and repetitions within the content. There were plenty. For instance, in the original document about 80% of the information that was presented in a table was restated in the text. At the same time, neither presentation was complete.
There was also the two-page FAQ. This material was almost totally redundant with content presented elsewhere. This is a common pattern in legal or disclosure documents: create a document no one can read, then instead of clarifying the document, distill the main points into an FAQ. This, of course, only makes the total amount of material even longer. We’ve seen this pattern so often that we consider an FAQ something of a red flag, in that it may be a symptom of a document that is failing to communicate. On the plus side, if you fix the document, you can generally get rid of the FAQ.
A similar problem is what might be called “emphasis inflation.” Because of poor drafting, important points become lost in the confusion, leading commenters to add sternly worded repetitions of these points, typically in bold or all caps (or both). Since commenters from different departments often have different ideas of what needed to be emphasized, the document is eventually covered with redundant language, all of it “yelling” at the reader—with the important points getting lost in the confusion.
In this project, by reducing document volume and by giving key points appropriate emphasis within the ordinary context of the document (rather than as additional labels or warnings slapped on top), it was possible for those points to come across very clearly.
Once again, structure played an important role—points became clearer when presented in context rather than as isolated thoughts. Visual cues were also important. Within the fairly restrained system of hierarchies we used in this document, there were still elements that gave emphasis where it was needed.
Step 5: Closing the Distance between Content Requirements and Readability
There is an inherent tension between the goal of the simplification expert and the goal of the legal drafter. The simplification expert is trying to help the reader get as much meaning as possible as quickly as possible. The legal drafter (usually an attorney) is trying to write something that will endure scrutiny—specifically, something that leaves no room for any possible meanings other than the intended meaning. Said another way, the simplification expert is focused on the first reading, while the attorney is focused on the last reading.
This focus on the last reading is why attorneys expend considerable labor spelling out what may seem obvious. The crossing and recrossing of the same territory from different angles and the verbatim repetition of large chunks of language may seem like conspicuous vices to a plain language writer or designer, but to an attorney they can appear harmless, even a virtue.
It is common to fault attorneys for writing this way, but the criticism misses the point. An attorney’s job is to prevent legal damage. If in the process they happen to use more words than might strictly be necessary, that is not a concern for them—nor should it be. Their concern is that the job gets done, not that it gets done in an especially elegant or efficient manner. Making things efficient is a separate process.
The attorney is quite welcome to undertake this process—although most, to be sure, are neither inclined nor adequately skilled to do so. This is why the job of creating good policies and procedures usually falls to a professional writer. As long as that writer can figure out a way to say what needs to be said in fewer words, most attorneys will be quite happy with it (and many actually end up preferring it). There is no inherent reason that the first reading and the last reading cannot be the same reading. In fact, when this is the case, it helps every type of reader. In our experience, the client’s attorneys almost always acknowledge that the simpler documents we develop are easier to maintain, once they have become familiar with them.
Step 6: Do the Actual Design, and, Finally, the Writing
Only after all of the steps described above did we put pen to paper (so to speak). We find that it’s much easier to work productively when there is so much already in place before we embark on the actual design and writing.
Even at the execution stage, there’s a lot of collaboration between writing and design, including use of tools from the world of marketing. This type of collaboration gives each discipline a chance to show which issues it’s best at solving. The writing and design work hand-in-hand to communicate with the reader. The marketing tools help to ensure that the result will signal to readers that their concerns are understood. Although the partnership among the different communications disciplines and the different stakeholders places additional constraints on the project, paradoxically acknowledging those constraints early on makes the process easier and the result better for all users.
We also find that the user-friendliness introduced by good structural and visual aspects gives us a little more leeway on the language side. Readability, after all, is not merely a matter of how much “resistance” the text offers, but of whether a reader’s willingness to engage with the text is sufficient to overcome whatever resistance may be encountered. In a document like this, where the legal team insisted on keeping certain language that really cannot be called plain, the combined value of good structure, good design, good visual cues, and significantly reduced length helps the reader readily negotiate the more complex content. We think the design can keep readers engaged through some of the “bumps in the road.”
This, again, is a useful tactic borrowed from the world of marketing: when one part of your message presents an irremediable difficulty, your audience is more likely to stay with you if other aspects of the communication are good. We’ve all seen ads that place a strong emphasis on attractive features or on status value as a way of trying to get us to accept a high price. It’s Marketing 101, and marketers do it because it works. There’s no reason they should be the only ones to benefit from techniques like this.
When a reader encounters a communication (whether a document, website, or anything else), that reader does not distinguish between the different factors that make the communication appealing or unappealing on first impression, nor between the factors that make it easier or harder to process once they engage (assuming they do). For this reason, we approach document creation by paying attention to all of the dimensions that can affect both initial appeal and continued engagement. Only by doing this can we claim to be creating communications that are truly working as hard as they can.
Josiah Fisk is founder and president of More Carrot LLC, a plain language and information design firm with offices in Boston and Luxembourg. He is in frequent demand as a speaker and is on the board of the Center for Plain Language in Washington, DC.
