How to use UX principles in writing.

June 18, 2025

Filed under: AI in Marketing, Writing Tips — @ 8:09 am

“UX” (user experience) is one of those industry terms that can sound like fancy jargon to outsiders and requires highly specialized knowledge. In fact, in my experience, even seasoned marketers can throw the term around to sound like they know what they’re talking about without understanding how UX principles truly work.

But if you have to communicate things as part of your job, whether to internal or external audiences, you should be using UX principles in your writing. It’s a secret weapon that will ensure you get what you want out of those communications. And this is true even if you’re using AI to assist with drafts.

First, some quick background, and then I’ll show you how to easily apply UX to your writing.

The term UX refers to the overall experience a user has when interacting with a product, system, or service.

In simplistic terms, for example, a website with “good UX” is easy to use…speedy loading pages, clear navigation, visually attractive, and a seamless experience with no annoyances. A website with “bad UX” is difficult to use…it can have pages that take forever to load, be confusing to find what you want, have harsh or distracting visuals, have broken links or missing information, and so on.

UX is a term heavily used in design work…graphics, visuals, and products. Letting UX principles guide their work enables designers to create things that have smooth, frictionless interactions with minimal (if any) negativity for the user.

But imagine how much better your life would be if all your written communications created smooth, frictionless interactions.

Here’s how UX applies to writing.

In this context, I don’t mean becoming a “UX copywriter,” which is a specialized job in itself.

I mean the everyday communicator – in marketing, sales, the executive suite or whatever – using UX principles to make their written communication frictionless.

This can apply to emails between colleagues/bosses/direct reports, documents of any kind, marketing emails to key audiences, and especially anything that has a call to action. Because if you want the reader to DO something, you’ll want to make it as easy as possible for them to do it.

To achieve that, let these three tips guide you:

Spend enough time before writing to clarify in your own mind exactly what this communication needs to achieve.
Anticipate questions and hesitations your readers may have and address them proactively.
Make strategic choices about how the information is organized and – if necessary – comes across visually.

Let’s break each of those down further so you know exactly what to do.

1) Clarify your goal before writing.

Believe it or not, most folks don’t do this…and I say that from my vast experience hosting writing workshops and working with tourism executives on their writing. Most people sit down and just spew out words. If you’ve ever tried to write an email or document and spent an absurd amount of time backspacing over what you wrote before you get into a groove, you know what I’m talking about.

You might think you know what your goal is, but in reality, people most often come at writing with the goal of “I have to write (this thing).” They’re not really thinking of the writing part as just a tool that’s going to help them achieve an end goal, which is communicating information and/or getting the reader to do/think/feel something.

When you aren’t crystal clear on your goal, it usually results in “bad UX” writing, such as:

Wordy, repetitive information that never quite makes the right point
Any calls to action getting lost in a sea of meandering words
Confusing the reader as to what they’re expected to do/think/feel
Losing the reader’s interest quickly

Take a beat – or as many beats as necessary – before you start writing to answer this critical question: what do I want the reader to do/think/feel after reading this? And then, keep that goal front and center to guide your thinking for the next two steps.

But keep in mind that clarifying your goal is particularly essential if you’re using AI to help with your draft. If you’re a regular AI user, you know that the clearer your prompt, the better the outcome. “Garbage in…garbage out” is an apt phrase here.

2) Anticipate and address questions proactively.

This one might be the hardest part because it’s not always easy to put yourself in someone else’s shoes. But I have found that often times, people write with blinders on…filling in detail gaps in their own mind and not realizing key points are missing from the finished piece of writing.

If you want a good laugh that proves this point, watch this dad try to follow his kids’ written instructions for how to make a PB&J sandwich.

When your audience is reading your communication, whether that’s a meeting invitation to your team or a marketing email to potential guests, “good UX” writing anticipates and addresses all their questions in the initial communication.

Why is this helpful? Because they don’t get distracted away from your call to action by all the questions that pop up in their brain as they’re reading. Even better, it reduces (or hopefully eliminates) the need for them to come back to you with questions, which would delay you getting the result you want…which is getting them to do/think/feel something.

This example will give you an idea of what I mean.

You’re inviting colleagues to a three-hour strategy meeting from 1-4pm. So you write an email that says:

Hey all. We’re going to meet about the 2026 strategy next Thursday from 1-4pm. Please come prepared to discuss your thoughts based on the budgets and programs that John shared earlier this year. I’ll send an invite. Thanks!

If I’m the recipient of that email, I’ll immediately wonder things like:

Should I eat lunch before or will food be served? (yes, my first thought is always about food)
Do I have those docs from John? Did I even get them? Where can I find them if I don’t have them?
Am I going to be asked to make a presentation here or is it more of an informal discussion?
Who’s running the meeting and what will the agenda actually entail?
Three hours is a long time to sit in a meeting…will there be a break?
What’s the goal of the meeting…are we making decisions here or just discussing things?
What else do I need to do to be prepared for this meeting?
Is this virtual or in person?
And if it’s not clear by the email distribution, who’s coming to this meeting?

If that email went out to 10 people, all 10 of them are likely to have at least a few of those questions…which means you’d be stuck volleying back and forth with 10 people to answer questions.

A “good UX” approach of that same email would be:

Hey folks.

All department heads will meet in person next Thursday from 1-4pm in the 18^th floor conference room to discuss the 2026 strategic plan. Please come prepared to discuss your thoughts based on the budgets and programs that John shared earlier this year, which are attached again here for your convenience.

Sue and I will lead the meeting and each person will be asked to spend 5 minutes sharing how the budget impacts their department. Then we’ll brainstorm together on how to achieve the goals shared on the last page of the attached program document. We’d like to leave this meeting with at least five workable ideas for discussion at the next leadership meeting.

Drinks will be provided and please feel free to BYO lunch. We intend to use the full three hours, but we’ll take a break midway through.

Please accept the calendar invite coming your way shortly.

Is that email a little bit longer? Yes. But it’s not TOO long and it will eliminate at least 98% of the questions the recipients are likely to have. Thus, allowing them to just accept the calendar request when it arrives with no additional time wasted on either side. THAT’S good UX.

What if you don’t yet know the details but need to quickly get out the invite to lock it in on everyone’s calendar? Send this:

Hey all!

We’re planning a meeting of all department heads in person next Thursday from 1-4pm at HQ to discuss the 2026 strategic plan. More detailed information will follow tomorrow to ensure that you’re prepared, but for now we just wanted to get the time blocked on everyone’s calendar. Please accept the meeting invite coming your way shortly!

So…even if you don’t have all the details, you can anticipate that they WILL have questions and just let them know from the start that those questions will be answered. It still eliminates friction and volleying.

3) Make wise organization/layout choices.

If your goal is to make it as easy as possible for a reader to embrace your message, then organization plays a HUGE role in how they receive/interact with your words.

Here are just a few facts to keep in mind:

Giant, long blocks of copy are desensitizing. Break long paragraphs up into shorter ones. Better yet, use bullets wherever possible. The human mind LOVES bullet points…they make information more digestible.

If you’re giving a deadline – whether a booking deadline for a vacation package or a response deadline on a colleague’s email – ALWAYS put that at the beginning so the reader knows the time sensitivity. If you put it at the bottom, or worse, hidden in the middle somewhere, there’s a high risk they won’t see it. Don’t assume they will read all the way to the end of whatever you sent.

Especially if you’re communicating a lot of information, or complex information, put down everything you need to say and THEN go back and reorganize. It’s hard to organize as you create, so let it all flow out and then reread it using the lens “what’s the best order to unfold this information for the reader?”

If you’re using AI to help with writing, you still need to be aware of facts like these to be sure you’re getting an effective result. You might have to prompt it with commands like “this copy is too dense, break it up into more easily readable paragraphs and use bullet points where possible.”

Also, sometimes to be most effective, written words need a little assist from visual layouts. Pete & Gerry’s did a marvelous job with this email to their database talking about the rising costs of eggs and the bird flu epidemic. The subject is serious and written words alone would have come across flat, like a crisis statement or a formal press release. The combo of visuals and simple, brief, and distinct text boxes make this email SO easily digestible by the reader. It’s not arduous to read it so you actually DO read it.

Overall, using UX principles in writing is the best way to ensure you get the result you want from readers. Does it take more time/thought on your part (even if you’re using AI) to approach writing this way? Sure. But that’s the whole essence of UX: you work harder/smarter to make it easier on the user.

And THAT makes it easier for them to do/think/feel what you want.

The world is getting judgy about AI usage.

May 14, 2025

Filed under: AI in Marketing, Observations From the Field — @ 12:46 pm

Is AI a useful tool, a cheater’s crutch, a superpower, or the scourge of the earth? Well, according to our recent survey about AI usage at work, it’s all of those things…and more. One responder even vehemently proclaimed, “AI is absolutely awful. Burn it down.”

Benefits aside (and there ARE benefits, arsonists notwithstanding), the arrival of AI tools like ChatGPT, Gemini, Claude, and others has kicked up a massive cloud of negative emotions along with it – fear, guilt, and anxiety among them. And many of these emotions come from peer pressure, which is especially prevalent in the marketing industry. You can feel judged for using it and judged for NOT using it.

For example:

AI champions think non-users are dinosaurs going extinct.
AI haters view users as lazy cheaters whose brains will turn to mush.
Some people hide their usage because they don’t want people to think they’re incapable of producing their own work.
Some people hide their non-usage because they fear people will think they’re not on the cutting edge of technology.

So how many people really ARE using it for work? And how often?

If you believe the media hype, EVERYONE is using it for work. The topic of AI and how to use it saturates media coverage, industry conference agendas, and business-focused social media platforms like LinkedIn. It’s no wonder it feels like the world is getting judgy about AI usage. The push to use it is everywhere you turn.

Yet, we recently surveyed 419 adults to ask for their anonymous response to the question, “How often do you use an AI assistant like ChatGPT to help with your work?” And despite AI tools being readily available for more than two years, the results aren’t as one-sided as the media-hype-peer-pressure would have you believe.

In this graph about how the world is getting judgy about AI usage, the data shows that 36% of people never use it, 45% of people use it occasionally, and 19% of people never use it.

So according to this survey:

36% of people never use it
19% of people use it every day
45% of people only use it occasionally

AND OH, THE COMMENTS! We added a box for optional comments and didn’t expect many. We were so wrong. People had a lot to say about it, all of which was quite illuminating.

From the comments (which you can read for yourself here), we learned things like:

Cool (simple) ways to use it we hadn’t considered, like correcting formula errors in Excel that you can’t figure out.
Many people won’t use Large Language Model (LLM) AI platforms like ChatGPT, Gemini, and Claude, and others because of their extensive drain on the environment. This is a particular concern for people and companies that have made a commitment to sustainability.
At the moment, there seem to be more “passionate haters” than “passionate lovers.” Like the colorful commenter that said, “Don’t trust AI, it’s anti-human and generally produces slop.”
It seems to be gaining traction as simply a helpful tool…many people use it for polishing up their own writing, getting them started with a framework for a document, and – as one responder put it – “using it as an extension of my own brain.”

That observation in the last bullet? That’s pretty much the stance of the “45% of people who use it occasionally.” They just see it as a tool…one they use when it’s needed and don’t when it isn’t.

And that’s really what it comes down to in the end.

You are not a dinosaur if you DON’T use AI.

And you are not a superhero if you DO use AI.

AI is just a tool. Tools are used for a variety of reasons, like to make things easier or to solve problems. And people adapt to using newly emergent tools at different paces. There’s no good or bad here, and no right or wrong. There are only choices and outcomes.

There are plenty of people out there leading perfectly happy lives without smartphones. Are there efficiencies and benefits they’re not taking advantage of because of that choice? Sure. Does it matter? No. Who says that doing more and faster is better? YOU may not be able to live without those efficiencies and benefits, but that’s YOUR choice.

We are still in the very early stages of LLM/AI usage by the general public, and it’s reasonable to expect that more people will come to adopt its use over time. Humans aren’t always quick to change and often resist learning how to use new tools and technology. Eventually, the new tech settles into whatever place it belongs in each of our lives, based on our own personal needs/likes/style. But initial resistance is normal.

I wasn’t around when the telephone was invented, but at the time, many people saw it as “the instrument of the devil.”

I was around, however, when the internet was invented and I fully remember having to counsel our marketing clients at the time on whether or not they should have a website. That too was a polarizing choice for people. And over the years, I saw clients cautiously come to embrace website marketing… then fully embrace it and strike all printed marketing materials from their budget. Then – whoa – realize that plenty of people out there still want/need/crave printed materials and scramble to rebuild sales lost from banishing them. And then eventually make peace with some sort of hybrid of print/digital mix.

And more examples…

E-books like Kindle didn’t replace books for every reader.

Food processers didn’t replace hand-chopping for every chef.

Task management apps didn’t replace handwritten to-do lists for every planner.

Hell, music streaming services didn’t even replace vinyl for every music lover.

And AI will never replace “writing” or “thinking.” Too many people get pleasure from brainstorming, strategic thinking, and crafting the perfect combination of words that articulate their point. Might those folks use AI as a tool for other things, or to help them occasionally with some parts of writing or thinking? You bet. But just because you use AI for SOME things doesn’t mean you have to use it for EVERYTHING.

So let’s all stop being so judgy about people’s AI usage. Use it or don’t…it’s totally up to you.

My Great Aunt Pauline – bless that woman’s wisdom – always used to say, “Things take just as long as they need to.” Which means, when it’s the right time for you to embrace AI for various things, you will. Until then, don’t stress about it.

And if you’re worried that we are all going too far into embracing automation and depersonalization (especially in hospitality), these stories from the tourism world will warm your little heart.

We’re all just suckers for homemade cake.

Masking tape: the unsung hero of authentic branding.

10 unexpected (and fabulous) tourism guest service stories.

Using AI in tourism marketing requires secrecy.

January 17, 2024

Filed under: AI in Marketing, Digital Marketing, Writing Tips, Effective Communication — @ 12:34 pm

The successful use of artificial intelligence (AI) in tourism marketing rests on one key factor: secrecy. Not secrecy from your colleagues or organization…secrecy from your audience. You need your use of AI to be invisible to them for two main reasons:

Blatantly apparent AI-generated copy is generic and robotic-sounding. And even if it were written by a human, that kind of soulless, impersonal copy doesn’t engage the audience and wastes your time and budget. The rest of your marketing will have to work that much harder to spark a sale, let alone close one.
If they sense you’re using AI to influence their decisions – and done incorrectly, they will – you’ll lose their trust. People are increasingly sniffing out AI-generated copy and then doubting its credibility. Humanity just hasn’t yet reached the point where we’re comfortable being persuaded to do something by a robot.

Let me be crystal clear. Using AI in tourism marketing is smart. Tools like ChatGPT and Google’s Bard can be useful in many ways. But YOU have to be smart in how you use them. Until you become proficient in prompting and re-prompting to produce successful output, you are at risk for – essentially – sending out a stoic, unfeeling salesperson to promote your offerings.

And that won’t end well for you.

As a tourism marketer, you’re trying to sell potential guests an experience that makes (hopefully) everlasting memories and creates (hopefully) a lifetime relationship with them. Robotic-sounding, generic marketing copy is incapable of igniting the spark required for that love affair. And when used incorrectly, that’s EXACTLY how AI-generated copy sounds…like a robot wrote it.

To guard against this, you need to be aware of six telltale signs that content was written by AI.

Jodie Cook wrote a fantastic piece for Forbes on the subject, calling out these six dead giveaways:

Lengthy introductions (aka “throat clearing”)
Inclusion of ethical considerations
Generic thoughts and advice
Lack of personal stories
Specific go-to phrases
Signature structure

You can read the piece for more information on all six, but the three most prevalent ones in tourism marketing are lengthy introductions, generic thoughts/advice, and lack of personal stories. Let’s take a look.

Lengthy Introductions

I call this the “blah blah blah” introduction and AI is famous for churning it out. A rudimentary AI-generated piece, like that of many unskilled human writers, puts a bunch of fluff at the start and takes a bit of time to get to the meat of the content. Often it includes clichés (“Once you arrive, you’ll never want to leave!”), or broad-sweeping statements that mean nothing in particular (“Come experience the magic of the outdoors!”). It’s usually filled with a lot of long sentences and densely packed with a quantity of adjectives that would make a thesaurus blush. This is the written version of speakers who begin their presentation with excessive throat-clearing…it’s buying them time to get into the rhythm of their speech.

Generic Thoughts & Advice

OMG, if I see one more tourism organization proclaim “We have something for everyone!”… I’m going to scream. Even if you truly DO have something for everyone to enjoy, that sort of vanilla claim has zero chance of actually luring a potential visitor. Without specific prompting and sculpting on your part, AI programs like ChatGPT are likely to generate generic content like “breathe the fresh air as you wander through our beautiful forests,” and “hop on a boat to get out and feel the ocean mist on your face as you watch an orange-hued sunset,” and “sip and taste your way through our vibrant dining scene,” and – my personal favorite – “come away with memories that will last a lifetime.” None of that is specific and unique to YOU, so why should it compel anyone to choose YOU? It’s just…uninspiring.

Lack of Personal Stories

First cousin to “generic thoughts & advice,” a lack of personal stories isn’t meant so literally as in “a person telling a story about their experience with you.” That’s part of it, for sure. But on a broader level, it’s about your BRAND making a personal connection with the audience. Doesn’t matter if you’re a destination, hotel, cruise line, attraction, or even just a travel service…whatever. Sharing your quirks, your variety of unique Instagram-worthy experiences, and other stories that inspire them to feel a personal connection to you… THAT’S essential in successful tourism marketing. And it’s something you won’t get from AI-generated content without training it to write that way. It lacks the ability to do that on its own because by default, its process delivers one-size-fits-all content that’s impersonal.

And that right there is the problem. In the world of tourism, people are choosing where to spend their precious time and money, and this is VERY personal to them. Tourism is a passionate and deeply engaging purchase decision that goes way beyond transactional. They may not care if a robot wrote their appliance user’s manual (which is simply delivering information), but they sure as hell want – say – their honeymoon suggestions (which requires the dance of persuasion and has a lot riding on the outcome) to come from a credible source.

Which brings us to the best news of all, and a hilarious silver lining for tourism marketers. A brand is now no longer the LEAST credible source for promoting its own offerings. The “credible believability” pecking order currently stands like this, from most believable to least believable:

Someone I trust.
Someone I know casually.
A stranger unaffiliated with the product/service, which could be a media outlet or a random person on social media.
The organization itself.
A robot.

Y’all, we’ve moved up a notch. So don’t squander that gift by making it obvious you’re using AI in tourism marketing. Make that your little secret.

Not sure how to get started doing this properly? Check out these ChatGPT tips for tourism marketers.

Feel like you suck at writing and so you can’t properly judge AI’s output? These two quick reads will help you:

How to stop being an impatient writer.

Write better copy with patience and a thesaurus.

ChatGPT tips for tourism marketers.

February 13, 2023

Filed under: AI in Marketing, Public Relations, Digital Marketing, Website Strategies, Customer Service, Crisis Communications, Effective Presentations, Writing Tips, Marketing Strategies, This is Cool, Social Media, Effective Communication — @ 11:24 am

Here’s why tourism marketers need tips for using ChatGPT, an online program that engages in human-like dialogue based on a prompt: because we’re all too damn busy to waste time. And messing around with a new tool we’re not sure we’d even use feels a lot like wasting time. Is it worth it? Should you invest the time to get familiar with ChatGPT because that investment of time will pay off?

Short answer: yes. You need to know what this tool can REALLY do before you decide to embrace or reject it. It has some uses that may surprise you.

So let’s jumpstart your learning curve with some practical tips for how tourism marketers can explore the benefits of using ChatGPT.

First, let’s get one thing straight. ChatGPT is just a resource and a tool. You’ve got a lot of tools to help you do your job. Google is a tool. Adobe Illustrator is a tool. Semrush is a tool. But the relentless media frenzy around ChatGPT has given it near-mystical properties that make it seem more potent than that. Chill out, y’all. It’s just a tool. It’s one more resource in your toolbox to potentially help you do your job better, smarter, and faster.

And like all tools [she says sheepishly, aware that she barely knows how to use 5% of the available apps on her iPhone], its usefulness is only as powerful as your knowledge of how to harness it. I’ll never forget years ago when an accounting mentor said to me, “If you’re doing any manual calculations whatsoever or taking a long time to manipulate data in an Excel spreadsheet, then there’s a shortcut, command, or function you just don’t know about. Excel is designed to make life easier. If it’s making it harder, go learn more about Excel.”

ChatGPT is the same. When you first try it out, you won’t be savvy at knowing how to coax the most effective results from it. So you’ll plug in a few basic things and the outcomes will be unimpressive. And then, because you’re super busy and there’s no mandate that says you need to use ChatGPT, you’ll dismiss it as unhelpful and go back to the familiar tools in your toolbox.

But what if I told you that…

You could paste a particularly legalese-sounding section of a vendor contract into ChatGPT and say “explain this to me like I’m an 8^th grader”…and it does?

It could produce a style guide for all your team members to follow, after you feed it several samples of a brand’s voice to analyze?

It could take your 400-word bio and make it fit that directory listing’s 100-word requirement in just one click?

It could give you a substantive list of story ideas for your content calendar…and then organize them into a seasonal schedule…and then create first drafts of each piece of content, in different formats for social channels, blog posts, email newsletters…and even website copy that’s optimized for the keywords you require?

It can indeed do all those things and more…if you know how to prompt it effectively.

Janette Roush is Executive Vice President, Marketing and Digital, for NYC & Company, which is the official DMO/CVB for New York City. And she’s one of ChatGPT’s early adopters and passionate champions who is learning to master the “art of the prompt.”

“If you want to get ChatGPT to give you useful answers, the key is in how you formulate your prompt,” Roush told me. “I was once advised to think of it like an omniscient three-year-old. It knows everything under the sun, but it doesn’t know who YOU are, WHY you need to know, and WHO you’re trying to talk to. You need to prompt it with details like that for it to return a result that’s written in the context you need. Otherwise the result will be very generic and way less useful to your purpose.”

Roush has honed her prompting skills through persistent trial and error. In fact, she even documents her journey with ChatGPT on LinkedIn, making regular posts about prompts she’s tried for a wide variety of uses and the results they’ve produced. (Pro tip: Connect with or follow her there. You won’t regret it.)

Inspired by Roush, I took ChatGPT for a three-hour test drive one morning, just giving it prompts for various tourism-marketing-related things. One thing I quickly learned is that a generic prompt yields a generic answer and specific prompt yields a specific answer. Case in point: Look at how it adjusted its responses for social media captions based upon my specificity:

And it did the same thing as I sought its help to generate story ideas for Northern California:

Are those story ideas all perfect with no need for tweaking? Perhaps not. But did it give me threads to follow where before I had none? Absolutely. And some good ones too.

So, in addition to writing copy, one use of ChatGPT is to think of it like you would a sounding board. Or a brainstorming partner. It can’t ideate on its own (it’s not designed to innovate) but it can work with the prompts you give it to hit you back with starter threads.

Roush shared some spectacular direction on how to prompt ChatGPT as a sounding board in one of her recent LinkedIn posts:

You may be thinking “well, why can’t I just Google stuff like that instead of using ChatGPT?” And you can. But Google (“regular” Google, not the emerging Google Bard version that’s trying to infuse AI into its experience but not quite succeeding as of this writing) will give you a slew of different links for you to go explore and assimilate all the information on your own. And ChatGPT will just…answer you. Not with “here are ten sources you can read to find story ideas” or “here are ten sources to see how other destinations are making themselves an attractive esports destination.” It delivers YOUR story ideas, and tells you how YOUR destination can achieve an attractive esports destination profile.

And then – mind blown – you can direct it to actually WRITE that story about ice skating in Northern California or OUTLINE that strategic plan to develop esports tourism in NYC.

Again…will they be final drafts that need no tweaking? Absolutely not. They will be FIRST drafts, but if you’ve prompted with care, they’ll be pretty damn good first drafts.

And THAT saves you time, which is the whole point of using ChatGPT for marketing assistance.

But wait, you say. When I use Google as a resource tool, I can handpick from among sources on the results pages that I feel are legitimate and credible. Without such references, how do I know the information I’m getting from ChatGPT is accurate?

Folks, I remind you again that ChatGPT is not supposed to be a mystical tool that sees all and knows all. You’ll need to check your facts, just like you would using any other source. Do you really think that something is accurate just because you got it from a source on Google that you consider “credible?” News outlets get details wrong, websites have outdated information, and inaccurate stuff has a way of floating around and perpetuating online. So, ChatGPT is no more nor less credible than any other source you use. And you should do your due diligence on its output when necessary.

And while we’re at it, I should also remind you that most of the output you get from ChatGPT will need tweaking and polish. Even with the absolute best of prompting, there will still be nuances and phrasing you’ll need to infuse. So it can’t hurt to brush up on your writing skills, and these tips will help.

If you want to explore how ChatGPT can potentially help you with your tourism marketing needs but you’re not sure how to begin, Roush offers these four tips to get started:

Commit to a finite time period for practice. You won’t learn how to use any new tool unless you devote time to using it. Roush recommends setting a challenge to yourself, with some kind of accountability built into the period. Take two weeks or a month or whatever, during which you commit to prompting ChatGPT on at least one topic every day. “I challenged myself to post a new ChatGPT insight on LinkedIn every day for a month, and it forced me to think of that tool daily,” she says. “It didn’t come naturally to me at first, but after a while, as various needs arose throughout the day at work, I’d automatically say to myself ‘let me see how ChatGPT would handle that.’ And then I’d dive into prompting.”
Don’t think of it just for help with writing. With accurate prompting, ChatGPT is an excellent resource for organization, explanations, curation, and more. Roush says it’s helped her structure her lesson approach for her work as an Assistant Professor at Hunter College, and it’s helped flesh out her vacation itinerary in Montreal by finding cool things to do nearby to her already-planned stops. “I’ve also used it to help it explain things I don’t fully understand,” she says, “like when I understand 80% of a technical proposal and I want to understand 100% of it. I can ask ChatGPT to explain it to me in layman’s terms.”
Learn to become specific in how you prompt. You won’t be good at this right out of the gate. It takes time and practice to master the art of prompting. When Roush first dabbled in using ChatGPT, she – like most folks – prompted it with “silly things,” just trying out generic questions and commands, and receiving lackluster responses. “It wasn’t until I stumbled upon how to start being more specific that I began to see the possible uses of ChatGPT,” she says. “I had asked it to create an itinerary for my vacation in Montreal and it was pretty vanilla, just hitting all the major tourist sites. But when I fed it my existing itinerary and asked it to suggest enhancement additions using the right prompts for specificity, it really impressed me.”
Let ChatGPT create a style guide for you, so it learns to deliver responses in your own voice. Roush fed it around 40 of her previous LinkedIn posts and asked it to create a writing style guide for her… which it did shockingly well. Now she can instruct ChatGPT to use that guide when asking it to write stuff on her behalf. “It was surprising how well the style guide captured my voice,” she said. “If I had tried to analyze my own work and write up my own style guide, it would have taken forever and probably been less accurate.”

The bottom line is that the more you use it, the more uses you’ll discover for it. And with practice at the art of prompting, you can make ChatGPT something akin to a full-service virtual assistant who brainstorms, writes, organizes, and educates.

Or… not. You may end up hating it, but until you REALLY take it for a lengthy and diverse test drive, how will you ever know?

Related reading: Issac Asimov’s I, Robot. It was written in 1950 and well…here we are, folks.