Category Archives: Storytelling

The Lazy Person’s Guide to Text Wrangling

Writing in plain text is increasingly popular. Writers of all kinds are adopting Markdown, the plain text writing format. They swear by the zen-like benefits of writing in plain text. WYSIWYG isn’t cool any more. Yet little attention has been given to how to work with plain text when editing material originating from many different sources.  The growing popularity of plain text opens new opportunities for the creative reuse of text content, because plain text is inherently portable, able to move between different applications easily. This post will describe how to edit plain text content when different sources provide raw text that is used to develop new content.

Reusing and repurposing text is helpful in many situations.  For a personal side project, I’ve been exploring content design options, using representative content, for a prototype.  I want to reuse public information from different sources and in different formats (e.g., lists, tables, descriptive paragraphs). This content offers possibilities to combine and remix information to highlight specific themes. Even if I were a fast and accurate typist, retyping large volumes of text that already exists is not desirable or feasible.  Cutting and pasting text is tedious, especially if the text is formatted.  I wanted better tools to manipulate text encoded in different formats, and change the structure of the content.

Digital text is generally not plain text.  Digital formats can, however, be converted to some flavor of plain text.  Content designers may acquire digital text that exists as HTML, as CSV (a barebones spreadsheet), and even as PDF.  Each format involves implicit forms of structure, at various levels of granularity.  CSV assumes tabular information.  Plain text assumes linear information.  HTML can describe text content that contains different levels of information:

  • Names, words, dates, numbers and other discrete strings
  • Phrases that combine several strings together
  • Sentences
  • Paragraphs
  • Headings and other structural elements

We can edit and fine tune all these levels using plain text.  Text wrangling makes it possible.

What is Text Wrangling?

Text wrangling converts and transforms information at different levels of granularity.  For example, information in a list could be converted into a table, or vice versa.  Text wrangling can restructure and renarrate information. It can also clean up content from different sources, such as standardizing spelling or wording.

Word processors (Word, iA writer, Scrivener, etc.) are designed for people writing fresh content.  They aren’t designed to support the reuse and repurposing of existing content.  When trying to manipulate text, word processors are rather clumsy.  Word processors fail when wanting to:

  • Generate content variations
  • Explore alternative wordings
  • Make multiple changes simultaneously
  • Ingest content from different sources that may be in different formats
  • Clean up text acquired from different sources.

Text wrangling differs from normal editing.  Instead of editing a single document, text wrangling involves gathering text from many sources, and rewriting and consolidating that text into a unified document or content repository.  Text wrangling applies large scale changes to text, by automating some low level transformations.  It uses functionality available in different applications to reduce typing and cut-and-paste operations.  This editing occurs during a “pre-drafting” phase, before the text evolves into a readable “draft”.  Editors can wrangle “raw” text fragments, to define themes and structures, and unify editorial consistency.

Tools for Text Wrangling

Many applications have useful features to wrangle plain text.   Ironically, none of these applications was designed for writers; most were designed for coders or data geeks.  As writing in plain text becomes more popular (in Markdown, Textile, AsciiDoc, or reStructuredText), more people are using coding tools to write.  These tools have enhanced editing features lacking in word processors, particularly the many “distraction free” apps designed for writing in Markdown.

Because none of the wrangling applications was designed specifically for text prose, no one application does everything I want. I use a combination of tools, and switch between them, depending on which is easiest to use for a specific purpose.  That sounds complicated, but it isn’t.  Plain text can be opened in many applications, and can be copied easily between them.

I use three kinds of tools to rework text:

  1. Spreadsheets (Google Sheets, Excel)
  2. Text editors that are primarily designed for coders (TextWrangler, Brackets, Sublime Text)
  3. Global utilities that are available to use within any application (TextSoap, Paste).

Before I share some tips, a few caveats (remember, I promised a lazy approach). These tips are not a comprehensive review of available apps and functionality.  Other apps provide alternative approaches, and some will be unknown to me. Because I use a Mac, my experience is limited to that platform  My preferences are motivated by a desire to find an easy way to perform a text task, without needing to learn anything fiddly.  Most developers would use something called Regex scripting to clean text, which is a powerful option for those comfortable with scripting.  I’ve opted for a quick and dirty approach, even if it is occasionally a messy one.  Lastly, apologies to my tech writer friends for my bloggerly presentation — I’m assuming everyone can locate more specific instructions elsewhere.   You’ll learn more from a Google search than I can provide you through a single link.

General Approach to Wrangling Plain Text

The basic approach to text wrangling is to work with lines of text.  Text editors organize text by line, as do spreadsheets (which call them rows). To take advantage of the functionality these tools offer, we convert text into lines.  Everything, whether a word, a sentence, a paragraph or a heading, can become a line that can be transformed.  Lines can be split, and they can be joined, to form different levels of meaning.

All text can be considered as a line, that can be transformed into different structures
All text can be considered as a line, which can be transformed into different structures

Some lines of text are just words or phrases — these lines are lists.  Some lines are complete sentences.  Working with sentences on separate lines is flexible.  It is easier to perform operations on sentences, such as changing the capitalization, when each sentence is on a separate line.  It is also easier to reorder sentences.  When finished with editing, individual sentences can be joined together into paragraphs.  Brackets has a “Join Lines” function (an extension) that makes it easy.  Just highlight all the lines you want joined together, and they become a single line of several sentences forming a paragraph.

Stripping out HTML and other markup

The first task is to get the content into plain text.  Working in plain text makes it easy to focus on the text.   If you have text content that’s encoded in HTML, you’ll want to get it into plain text, without all the distracting markup.  Even if you are comfortable reading HTML, you’ll find CSS and Javascript markup that’s irrelevant to the text.

Sometimes you can get plain text by selecting the “Reader View” in your browser, and copying or emailing the text and saving it.  Alternatively, you may be able to acquire tabular or structured text on websites from within Google Sheets, using “ImportXML” or “ImportHTML”.  These functions take a moment to learn, but can be very helpful when you need to get a little bit of text from many different webpages.

When you are working with text files instead of live content, you want a way to directly clean the text without having to first view it in a browser.  Open the file in text editor, highlight text, and use TextSoap to strip out the HTML markup.  TextSoap is a handy app that can clean text, and can be used with any Mac application.

Breaking Apart Phrases

Phrases are the foundation of sentences, labels, and headings.  If you need to wordsmith many words or phrases, it may be easiest to get these into a list, and work with them in a spreadsheet.  A list is essentially a one column spreadsheet.  By breaking apart the text in one column into different columns, you can modify different segments of the text, changing their order, standardizing wording and formatting, or extracting a sub-string of text within a longer string.  A common, simple example relates to people’s names.  Do you want to list an author’s name as a single phrase (“Ellen B. Smith”)? Or would you like to separate given name(s) and family name?  What order to you want the given names and surnames?

Spreadsheets can split or extract the text in one column into one or more new columns.  You can create a new column for each distinct word, which allows you to group-edit distinct words.  Or you may want to extract and put into a new column what’s distinctive or unique in each line.

Google Sheets has a function called “Split” that breaks the text in a column into separate columns.  When words are in separate columns, it can be easy to make changes to specific words.  “Substitute” allows specific words to be swapped.  “Replace” allows a sub-string to be changed.

Consolidating Phrases

Spreadsheets are good not just for breaking apart words, but for combining them.  You can take words in different columns on the same row and combine them together using “Join.”  “Concatenate” allows words from different columns and different rows to be combined.   This is an even more flexible option, because it lets you try unlimited combinations of words in different cells.  For example, you could play with different word hierarchies (broader or narrower words on different rows within a single column), or array a range of related verbs or adjectives across a singe row in different columns.  “Concatenate” can enable simple sentence generation.

A different situation occurs when you want to take information that’s within a table, and express it as a list.  A matrix table will have a column header, a row header, and a value associated with the column-row combination.  Suppose you have a table listing rainfall, with the columns representing months, and rows representing years, and the cell representing the amount of rainfall.  This information can be transformed into a single row or line.  In Excel, a function called “Unpivot” does this (Google Sheets lacks this functionality).  It presents all the information in a single row, such as “May | 2017| 2 cm”, which can be joined together.  These values can be transformed further into a list of complete sentences, such as “In May 2017, the rainfall was 2 cm”.  That list of sentences could become the beginning of separate paragraphs that discuss the implications of each month’s rainfall on the local economy.

Removing Redundancy

It’s helpful to put each discrete idea in the text on a separate line.  These may be names of topics, phrases, or facts.  During the text development phase, you’ll want to collect all text strings of interest.  Text wrangling tools can help you collect everything of potential interest, and worry later about whether you’ve already covered these items.

Suppose you need content about all of your products.  You can create a list of all your products, with each product on a separate line.  If you have many product variations that sound similar, it can be confusing to know if it’s already in the list.  If the list is a spreadsheet, it is easy to remove duplicates.  If the list is a text file, TextWrangler has a feature called “Kill Duplicates”.  To spot near-duplicates, sorting the lines will often reveal suspiciously-similar items.

Spotting duplicate or redundant paragraphs takes an extra step.  To compare alternative paragraphs, put them in separate files.  TextWrangler allows you to compare two files using the function “Find Differences”.  Both files are displayed side-by-side, with their differences highlighted.  This approach is more flexible than a word processor, which can assume you want to merge documents, or choose which document is the right one.

Harmonizing Style

A big task when assembling text from many sources is harmonizing the style.  Different texts may use different terminology.

Text editors, like word processors, have “Find and Replace” functionality, but they offer even better tools.  Find and Replace is inefficient because it assumes you have one word you already know you need to replace with another word.  Suppose instead you have many different words referring to the same concept?  Suppose you aren’t sure what would be the best replacement?  This is where the magic of “multiple selections” comes into play.

Brackets’ Multiple Selections feature can let you make many edits at once.  (Sublime Text has a similar feature).  All you need to do is highlight all the words that you want to change.  Then, you type over all the highlighted text at once, and see the changes happen as you type.  Words change on many lines at once, and you can try out different text to decide which works best across different sentences.  And to repeat: the highlighted words being changed don’t need to be the same word.  If you have some sentences that talk about dogs, some sentences that talk about canines, and some sentences that talk about mutts, you can highlight all these words (dogs, canines, mutts), and change them all to “hounds” — before deciding to say “man’s best friend” instead.

Brackets has a related feature called “Multiple Cursors” that is also amazing.  It allows you to place your cursor on multiple lines, and edit multiple lines at once.  Suppose you want to decide the best construction for some headings.  You want to know if saying “{Product X} helps you {Benefit Y}” is best, or “{Product X} makes it possible to {Benefit Y}”.  You list all the products and their respective benefits on separate lines.  Then you can edit all the headings at once, and try out each variation to see which sounds best.

Shifting Perspective

Many wording changes involve changing voice, or flipping emphasis.  Do you want to discuss a task using the imperative “invest” to emphasize action, or by using the gerund “investing” to emphasize a series of activities?  If you have many such tasks, you might want to put them in a list of statements, and try out both options.  You can then decide on a consistent approach.

In addition to the multiple cursor approach, you can edit multiple lines of text using the “Prefix/Suffix” functionality available in TextWrangler.  This allows you to either insert or remove either a prefix or a suffix to a line.  This could be useful with deciding on the wording of headings.  Maybe you want to see what the headings would sound like if they begin “Case Study: ” or whether they should end with “(Case Study)”.

Skeleton Frameworks

Plain text tools can help you reuse text elements again and again.  This can be useful if you have a template or framework you are using to collect text.

Sublime Text has a feature called “Snippets”, where you can store any text you want to reuse, and inject it into any file you are working with.

Another option is a small utility called Paste, which works with any application on a Mac.  It is like a huge clipboard, where you can store large snippets of text, give these snippets names, and reuse them wherever you may need them.

Adding Markup to Plain Text

Plain text is great for writing and editing.  But eventually it will need some markup to become more useful.  Several options are available to turn plain text into web text.

Many writers have adopted Markdown. You can add Markdown syntax to the text, and convert the text to HTML.

You can also add basic HTML elements to plain text using TextSoap, which is a utility that can be used in any Mac application. You simply highlight the words you want to tag, and choose the HTML element you want to use. This option may be desirable if you need elements that aren’t well supported in Markdown.

The most robust option is to use the tagging functionality available in some text editors.  You can add markup using Brackets’s “Surround” extension, where you highlight your text, then define any tagging you want to place around the text.  Sublime Text has a similar feature: “Tag > Wrap Selection.”  These features let you add metadata beyond simple HTML elements; for example, to indicate in what language a phrase is.

Limitations of Text Wrangling

Text wrangling techniques can be handy in many situations, but will be inefficient in others.  They are intended for early content development work.  As I’ve discovered, they can be helpful for assembling text to prototype content.

These techniques aren’t efficient for editing massive content repositories, or editing single documents that aren’t very long.  If you need to migrate large volumes of content, you’ll want some custom scripts written to transform that content appropriately.

Text wrangling focuses on redrafting raw text, rather than collaboration, which will generally get delegated to another platform such as GitHub. Word processing apps offer better support for the review of well-defined drafts, where comments and change tracking are important.  If you are editing or rewriting individual documents, especially in collaboration with others, tools such as Google Docs that track comments will be a better option.

Someday, I hope someone will develop the perfect tool to edit text.  Until then, using a combination of tools is the best option.

—Michael Andrews

Helping People Make Tough Choices

People face tough choices in their lives everyday — choices without one clear winner. Perhaps available choices have different advantages, or all choices involve some kind of compromise. Content should illuminate what’s at stake when making these choices. In many situations, content fails to do this.  It hides what’s at stake, doing so in the name of simplicity.

When Simplicity is a Ruse

Many of us, myself included, don’t like fiddling with complicated financial decisions.  We wish there was a page we could go to and tap a button that says “Don’t bother me with this again — just take care of it” and we’d never have to think about it again.  Instead, if we find such a page with such a button, we later find out that we’ve made a decision that is nearly impossible to undo. That same website offers another page with an endless loop of popups asking us: Are we sure we want to cancel, given that there is a $1000 penalty?  It might as well ask us if we want to suffer lots now, or suffer a good bit indefinitely.

Financial news channels like CNBC and Bloomberg advertise smartphone apps seeking to manage your money.  Some apps promise to find you the best credit card, while others promise easy and cheap mortgages.  Tech investors fund countless startups who claim their app offers a new level of simplicity for financial management.  These apps have slick UIs, and appeal to prospective customers who imagine themselves as take-charge and in-control.  With just a couple of taps on the app, you can execute your decision without the annoying fuss that other firms make you go through.  Target users know what they need:  an app that lets them make big decisions wherever they are, even while running a half marathon during their lunch hour.  Other ads show day traders thumbing their apps, making big buys while walking toward their private jet.  Miraculously, the chaotic world of finance becomes a zen-like zone of clarity.  Everything is now simple.

These two-tap solutions suggest they offer everything the user needs to know, even if sometimes that’s not true.  The apps imply financial decisions can be boiled down to a couple of numbers, and content never has to get in the way of taking action. The best deal never involves a teaser rate that will change, or draconian terms and conditions invoked if you change your mind later.  Making a rapid decision never requires one to think about the customer service experience after the fact.   Users are asked to trust that all these issues have already been accounted for, and that the app has only chosen options that are free-of-problems, and are the absolutely best ones available anywhere.

Many businesses repeat the mantra of simplicity in connection with their customer proposition. Instead of being a genuine belief, the mantra is often just empty words.  Simplicity, a rock-solid idea, can be cynically manipulated to trick customers into believing the convenience they experience is in their best interest.

Simplicity became a marketing buzz word at the same time as user experience (UX) became a commodity. They became features to brag about, rather than experiences to faithfully deliver. Simplicity is a good thing, but that statement needs to be qualified. True simplicity is paradoxically difficult to offer.

Easy Now or Easy Later?

When people working in UX or behavioral economics talk-up the ideal of simplicity, they often are actually championing the notion of convenience.  Convenience implies something is easy-to-do, while simplicity implies additionally that something is easy-to-understand When a company boasts of offering an app that lets you make choices on your smartphone, saying what you need is in the palm of your hand, they are appealing to convenience more than to simplicity.  Understanding involves more than taking action in a moment-in-time.  It involves seeing how the future of that action will unfold.

For much of the history of the web, people worried about drowning in information, a phenomenon called TMI (too much information).  Yet a recent Pew survey of American web users revealed that “worries about information overload are not widespread” now.  That sounds like a good thing, and in many ways it is.  People shouldn’t feel overwhelmed by information.  Smartphones have helped people keep up with information, and their small screens have brought discipline to previously long and complicated content.

But part of that progress has occurred at the cost of removing information that is valuable.  The same Pew survey revealed that Americans think that “schools, banks and government agencies expect too much from people”.  These sectors are routinely considered difficult to deal with, due to how much information they expect their clients to provide them.  To understand why banks and government agencies can’t be as simple to deal with as Amazon, we must consider how such institutions differ from a ecommerce retailer.

Compared with financial, educational and government institutions, retailers can automate their processes to a greater extent. Retailers control most choices about supplying, pricing, presenting and delivering goods to customers.  When retailers automate these processes, they remove related complexity from customers.  They handle much complexity on the backend, so that users don’t need to experience complexity on the frond end.  Simplicity is easiest to achieve when offering straightforward transactions, such as booking a concert ticket, because little information is required from the customer, and the customer needs to understand little in return.

An insurance company, on the other hand, needs the customer to make various decisions that involve multiple steps.  Insurance companies can’t presume what the customer wants. Customers need to be actively involved.  It is more difficult to automate customer choices, because the customer needs to understand what choices they are making.

Via Wikipedia
By DiacriticaOwn work, CC BY-SA 3.0, Link

Decisions that have Consequences

Trivial decisions are easy to outsource to other parties, but important ones require active participation.  People are short on time, and often low on energy.  People understandably want choices to be easy.  But they are not well-served when content implies decisions are straightforward, when in reality they involve many considerations.

Many customer decisions are iterative, multi-faceted, or nonlinear.  Customer priorities and circumstances are difficult to reduce to a few common variables.  Imagine being given two or three choices for a hair style when booking a hair cut online, with no further input allowed. While seemingly convenient, such a process won’t ensure a happy outcome. Even though haircuts are generally low-value purchases, they have big emotional consequences for the buyer.

Any transaction that potentially involves “buyer’s remorse” after purchase, where making changes is difficult to do, requires some active involvement from the buyer to indicate what they want.  These decisions include:

  • high-value, feature-rich, or maintenance-demanding products that will be owned a long time such as cars
  • emotionally symbolic purchases such as family vacations
  • personalized, long-term service relationships such as insurance and education.

By their nature, decisions that can involve buyers remorse are tough choices.  We’d never experience regrets if the right choice was obvious before we made it.

Tough choices typically have many inputs, often happening over a period of time.  Certain tasks resist automation, such as deciding on and making medical appointments, or doing taxes.  Many decisions are one-off ones, or are infrequent, so that circumstances will have likely changed since the prior time the decision was considered. People can’t just set preferences that will be triggered each time the decision needs to be made.

Some tough choices involve comparing two or more items that have many dimensions to them.  Other tough choices involve a single decision to do something that involves some risk. Perhaps the perceived cost of the decision is high, or the effort involved to commit to the decision deflates one’s motivation.

Too Much Information, or Too Little?

Publishers resort to generic ideas to determine how much detail to provide customers.  People routinely complain about having to read long documents.  Research shows that people feel overwhelmed when offered too many choices,.  As a famous, widely-cited experiment demonstrated, when people were asked to choose which breakfast jam to buy, offering too many choices resulted in lower sales. Common sense and experimental research both seem to suggest that publishers should limit the amount of information presented to audiences.  Don’t risk making the experience a chore to do — give people what they say they want, which is not having to read much.  The case for information- and choice-rationing sounds  straightforward. Sadly, we can’t afford to hide an important qualification: what’s simple now isn’t necessarily what’s simple later.

Streamlining information and limiting options is a great tactic for simple decisions like buying jam. But it can be a bad tactic for helping people make tough choices.

Too much information is a symptom, not the core problem. The core problem is that making an informed choice is difficult, due to the nature of the decision.  Some businesses eagerly take that chore away from customers, by limiting their choice. They may justify limiting choice by arguing that something that looks complicated does not look trustworthy.  Complicated content obscures what’s best for the customer.  According to this logic, if the proposition looks simple, then it will be trusted.  These businesses consider trust as a  matter of optics, rather than of substance.  People don’t need to be informed, they just need to feel something is simple.

When publishers remove information that’s critical for customers to know, they are offering fake simplicity.  They are restricting vital information, and in so doing, influencing the decisions customers make.  Fake simplicity hides information and choices that have consequences for the audience.  Fake simplicity hides potential problems from users.  Real simplicity addresses potential problems.  Informing users about issues and tradeoffs doesn’t make the content complicated — it makes the content useful.

Let’s look at a familiar example of real simplicity applied to an iterative decision.  The solution reveals information that could be problematic for users.  That solution is the traffic information available on Google Maps.  The user can just look at the map, and see how heavy traffic is.  It couldn’t be simpler from the user perspective.  Behind the curtain, the app draws on complex processes to deliver this information.  Google pools data from all its smartphone locations to produce a composite picture of how quickly traffic is moving.  It is complex from Google’s perspective, but simple from the user perspective.  And it provides the user with vital information they care about in an easy-to-understand manner.  No one complains Google Maps has too much information. Its transparency is what makes it useful when making driving decisions.  Even though in many cases there is one best option when choosing a driving route, the dynamic variability of traffic means that alternatives and adjustments always need to be considered.

Many apps appear to be convenient because they remove things you should be aware of.  Smartphone content in particular frequently hides critical information for the sake of superficial simplicity.  Content may:

  • Hide explanatory information informing decisions
  • Hide choices or options.

Publishers may hide information or options using UI tricks such as saying “more” or “what’s this?” to de-prioritize important information. They may leave out information people would want to know about, because the information is not legally required to be disclosed — at least at that stage. Often, the gotchas are loaded at the final stage of consideration, in the form of an T&C agreement check box.  Sometimes firms offer false choices, believing that having only one attractive option is the best way to get people to make a decision on a complex issue.  Or the user interfaces restrict choices that users might want.  For example, many people are happy to use Instagram to take photos, until they realize that Instagram doesn’t offer the choice of downloading the photo for use outside of the Instagram ecosystem.

Supporting Genuine Choice

User interfaces are easy to simplify.  Human emotions aren’t.  When people aren’t sure what choice to make for a decision that has consequences, they want both to understand, and to trust the information available.  Trust derives from integrity, and telling the truth.

When the choice is tough, content that suggests all options will make you happy, or there is only one best choice, doesn’t inspire trust.  True, customers don’t want too many choices.  Normally, more than four choices at once is too much.  But people want genuine choice, and know how choices differ and they will be affected by each.  They know that all options don’t result in equal levels of long-term happiness.

Choice is an emotional issue.  It involves personal expression.  One of the purest examples where people want to express choices is an online petition. They participate in a petition not to complete a transaction where they get something in return immediately, but to make their views known.  People want choices offered to reflect their real preferences.  “Simplicity has it’s limits. Make the process too simple and you run the risk of encouraging ill-targeted, unfocused petitions” notes David Karpf, a George Washington University professor who has studied petition websites such as MoveOn.org and Change.org.  These sites don’t try to make choice a single action, but break choices into a few simple steps.

“The feeling of choice…promotes wellbeing.  Our wellbeing depends on the ability to exercise choice…” notes Michael Bhaskar in his new book, Curation. People want choice that is meaningful to them personally.  They are skeptical when  sellers present false choices such as having a “best value” package prominently displayed in the center of options that aren’t good deals. They flinch at choices where it is hard to discern the differences among them, or evaluate which is best for a given situation.

Promoting Real Simplicity

If hiding or rationing information doesn’t help customers make tough choices, what does?  There can be too much information — if it isn’t relevant.  Real simplicity highlights the parameters that are most important when making a decision.  It shows what to consider, and provides the means to choose.  Real simplicity doesn’t try to make the choice for the user.  It helps the user know what’s important about the choice to consider.  It knows what issues have had consequences for people in the past, and reveals these to people now making such decisions.

For people who think about simplicity in terms of user interfaces rather than user experiences, simplicity seems like it involves taking stuff away: reducing the number of words, or number of clicks.  Those are sometimes useful tactics, but they tend to confuse task activity with making decisions.  From a user experience perspective, real simplicity isn’t about clicks —  it is about clarity.  As far as possible, people need to feel clear about the decision they are making.  They know what they are getting into, and are confident they’ve made the best choice possible.

Real simplicity doesn’t derive from word choices or screen layouts.  It embodies decisions about the kinds of information to feature, deciding what choice parameters to prioritize because they ultimately matter most.  Moreover, real simplicity offers a framework for understanding and comparing these parameters.  Readers have a context to see how different parameters influence their overall choice.

Some DON’Ts:

  • Don’t limit choices unless that is part of your up-front proposition
  • Don’t condense several decisions into a single package of choices that are hard to discern
  • Don’t offer choices few people will want to make, to make one seem better
  • Don’t bury explanations of important criteria

Some DOs:

  • Provide several meaningful choices that will appeal to different kinds of people
  • Emphasize what is most different about each choice compared with others
  • Break discrete issues into separate decisions, instead of bundling unrelated issues into difficult-to-grasp decisions requiring a leap-of-faith
  • Emphasize trade-offs involved with different choices when they exist
  • Offer ways for people to compare different aspects of a decision across different options, so they can focus on issues of greatest importance to themselves personally

Some of these guidelines trespass on decisions relating to product offers and pricing.  Content can’t be separated from these issues. Product managers should understand that attempts to drive customer behavior toward specific outcomes that are optimal for the business are not necessarily how customers consider their choice.  We see this issue now in the cable TV industry, where cable operators want to force customers to accept bundles that aren’t valued by the customer.  Unless there is alignment between how customers consider and value choices, and how firms offer choices to their customers, the firm’s reputation will suffer, and customer retention be vulnerable to market challengers who offer choices more aligned with customer needs.

Tough Choices are Individual

Tough choices aren’t categorical, divided into good choices and bad choices that apply to everyone.   Tough choices involve right choices and wrong choices as they apply to individuals.  Take the example of vacations: going to Aspen is not a “good” choice.  Aspen will appeal to skiers who can afford it, but won’t to people on a budget or who don’t care for mountains.  The individual’s situation and context shapes what’s ultimately best.  Even mundane decisions, such as purchasing a new home water heater, involve significant individual variation in circumstances and preferences.

Some of the information needed to support tough choices needs to be disclosed prominently on forms and in tables.  Features, specifications, applications, add-ons, exclusions, replacement and servicing costs, renewal pricing, incentives, projected maintenance, refunds and cancellation fees. Everyone can benefit from knowing such details.  People can notice these parameters and assess how they may be affected by them personally.  Yet content can do more than simply add transparency.

Tough choices, at heart, involve human stories.  Why did someone make the choice they did?  Different people will make different choices.  Each person will have their own reason for their choice.  These considerations can be highlighted in stories.

Stories are a powerful form of content, and are well suited to helping people make tough choices.  Stories can show why others made the decisions they did, and how that decision worked out for them:

  • What was their situation at the time of their decision?
  • What did they consider important, and how did they evaluate the choices available?
  • How happy have they been with the choice they made?
  • What do they wish they had done differently?

A great story might be one where a person gets advice from a friend about a decision.  The friend is happy with the choice they made, so the person decides to make the same choice.  Later, the person realizes they aren’t really happy with the decision they followed.  He realizes that his situation is different from his friend in an important respect, and he would have been better off making a different decision.  Some of the most powerful stories happen when people learn from mistakes.

Stories aren’t always simple — especially if we equate simplicity with minimalism.  They can involve many words, and twists and turns.   But stories can make complex situations understandable and memorable.  They can help individuals identify what’s personally important about a situation.

—Michael Andrews