Remember project briefs?

The project brief is law.

In my early career, that’s what I had been told. If a brief came across my desk that was unclear, imprecise or missing details, the client needed to take a hard look at what they were asking for before I could even begin to work out ideas. We’d complain about how their objectives were so far-reaching, that their audience included everyone, and that there were uncountable ways to execute on it — which meant there’d be countless comps and revisions to try and get inside the clients head.

We’d send them away to work out their ideas and to solidify better plans on their own, with instructions on how to write better project briefs in the future.

When they did have their ideas worked out, we’d never get things in front of users, mostly building design by consensus. Launches happened with little fanfare from actual users, with lots of stakeholders patting themselves on the back. Was success ever measured? It was rare and what metrics were tracked were spun to show their successes.

I’m so glad things have changed.

Getting stakeholders to involve experience designers earlier

The clients never changed. Instead, I looked at my role in the process differently — my clients weren’t equipped to answer a lot of the questions that would help me make informed design decisions — I had to teach them what was necessary — do what they thought was right and then validate if they were. Sometimes that would come at the very end or after a product launch, through analytics, surveys, or qualitative interviews.

As I showed them to examine their products and ideas through a user’s perspective, I was continually asked to become involved earlier and earlier in new projects — I moved my way backwards through the design process introducing new tools and techniques to my clients. What was solely UI design then became wireframing and visual design, then user flows and journey maps, and then to personas, ideation and strategy.

With each new method introduced in the process we become more confident in what we set out to create. Generally, most companies now have a better understanding of experience designers, but it took a while to get here didn’t it? If you’re working with people that don’t quite get it, I understand your struggles - the only way to convince them of the design process, is to just show them what their missing.

Eliminating ideas too soon

But there is still one thing I sometimes struggle with — clients still have a general idea of what they want to do. They might not know how to pull it off, but they’ve already sold themselves on their idea (or bought into an idea to early) — that it’ll work and people will use it. I want to challenge them to start generating ideas that they don’t fall in love with — to become even more user-centric in their thinking and practice by testing those ideas as early as possible.

A crumbled up piece of paper.
What's more 'on-the-nose' than a crumbled up piece of paper here.

I’ve seen rooms of people come to consensus — either through lengthy discussion, debate, or voting on an idea that they believe will be worth pursuing — eliminating nearly all of the ideas that were generated. Some of those ideas deserved to be tested but there is a perception that testing is time consuming and costly. So the winning idea is chosen because it balances its feasibility to be built, its cost, and the belief that it will succeed because it exists - with those other ideas becoming lost in the ether.

We need to stop that belief.

Methods of validating ideas and building what people will actually want to use

If you haven’t read it yet, The Right it by Alberto Savoia introduces a handful of pretotyping (Yes, pretotype tools that focus on this exact issue — clarifying ideas through quick and inexpensive testing, collection, validation and analysis of that data to make objective and informed decisions at the beginning of the design process. His professional experience in researching the failure of products has led him to create a toolkit of methods that can be used to verify if people will actually use your idea once you bring it to market.

A whiteboard drawing of a lightbulb and a chart.
This is way too conceptual to find an image, so I drew my own. Yeah, it sucks. I don't care.

Although some of his pretotyping techniques focus on services, a number of them are applicable to digital products. The techniques stress the importance of bringing your idea to a very small slice of your potential market — to gauge the interest in using the product, by faking its existence, by running it for a short period of time, or repurposing an existing similar product. Along with the techniques, he offers up measurable ways to validate the tests to give more confidence in your potential ideas — an approach that could save time, cost, resource, and potentially bring better ideas to market.

It’s a fun book and I’m going to reference it a few more times here - I’ve actually gifted this to some coworkers in the past and probably will again.

The more unclear, the better

Getting clients to buy into using ideation validation methods is another story.

A long fence along some farmland with a tree in the distance that's barely seen because of the fog.
Not my photo but I wish I had taken it.

If anyone approached me today with a project brief, I wouldn’t complain about how their objectives were far-reaching and that their audience included everyone — I’d be excited to work on a project that hasn’t been solidified in their mind, that we’d get the opportunity to validate our assumptions through pretotyping and to get real data before ever committing to an uninformed idea.

In fact, for me, the more unclear the ask is, the better.

The problem with some oft used metrics

There is a general understanding of the importance of metrics — daily active users, click through rates, conversion rates and NPS scores are at the heart of what makes business progress and what differentiates good products from mediocre ones.

In addition, specific user experience metrics, like perceived usefulness and engagement scores are starting to take hold alongside the more traditional metrics, but often these UX metrics struggle to define the reason to why they are what they are and what causes them. There’s not a lot of correlation between the value of a metric and what may actually be contributing to those values, under the surface.

Interpreting attitudinal and behavioral metrics

A group of people at a concert in front of the stage with their arms above their heads.
This is human behavior that I can get behind and understand.

Even the most useful metrics need to be interpreted — the problem is that they could relate to any aspect of how it was designed, explained, or implemented.

As an example, take the number of user submitted photos to an online photo sharing site that’s struggling to increase their user base — if the site has a low number of uploads per user, that metric does nothing to inform on whether there is an issue in the task flow, the visual design, or even the core idea of the product. These types of metrics can act as benchmarks for improvements to your product (make the upload process easy, incentivize the user to upload) but there’s a ceiling to where you can make improvements if your product isn’t something that people actually want to use in the first place. Possibly, there wasn’t enough real measurement on the original idea before creating a site that would attempt to solve a problem that didn’t really needing a solution.

You can continue to make those improvements to your dying website, with added cost and little return, or, in hindsight, you could have examined how you generate and test your ideas before you ever launch a product that people have little interest in.

Embracing not-so-great ideas and their failures

Although we have these baseline metrics, which can be useful in making improvements to good products, we have inconsistent or nonexistent metrics for evaluating ideas. We only look at what others have done through competitive analysis, make ‘best-practice’ decisions, and take a ‘good enough’ approach to new product designs. There’s often little thought or understanding on how to measure the value of an idea that doesn’t involve outright asking people what they’d think if it existed — which, more often than not, leads to false positives and reinforced (false) confidence in what you are creating.

A very old car with a broken wheel, in black and white.
I don't know anything about cars, but the whole concept of this car seems like a terrible idea. And it's broke, so it fits here.

Those ideas aren’t truly tested until prototypes get in front of potential users after much time, money and resource has been spent. Scrapping a bad idea that late into the process means having to spend as much or more time regenerating ideas and eventually testing with more prototypes. The amount of rework, at that point, usually informs the decision to just increment on a mediocre execution of a less-than-ideal product.

We often say we embrace failure, but in practice, we don’t know we’ve failed until it’s past the point of no return.

So how can we identify failure early on and gain confidence that our good ideas can succeed in the market? Savoia’s “TRI Meter” stresses the importance of having multiple ideas and a willingness to iterate over your ideas quickly — not when you’ve committed major resources, but when you first need to know if someone would be interested in using your new product or service — real metrics that are more reliable than the outcome of focus groups or surveys.

Learning to truly embrace failure comes much more easily when you’ve only committed an insignificant amount of time and money. The more ideas you can test and the more you can change those ideas for more testing, the more confident you can be that the idea you land on will have a higher success in the market.

Seriously, go read his book.

The future of market validation metrics

It’s easy to generate ideas on how to solve a problem, but we need a way to more accurately gauge which ideas are worth pursuing and better methods for eliminating ideas that aren’t heavily influenced by stakeholder opinion.

And, of course, we need to convince these stakeholders to let multiple ideas have their chance before anyone has latched on to a solution.

More importantly, we need to have better metrics for validating ideas, and measuring their influence on the outcomes of actual solutions.

I look forward to a day where I put together a report on market opportunities using tried-and-true viability metrics.

in ideation user-experience

Artificial intelligence, specifically language models such as ChatGPT, has the potential to revolutionize the way in which user experience (UX) designers conduct research, design interfaces, and write microcopy.

Spoiler Alert: This was written by chatGPT and the accompanying image was created by Stable Diffusion. AI is finally starting to make things weird.

A painting of a persons hand typing on a computer keyboard
A person's hands typing on a computer keyboard, possibly chatting with chatGPT, as imagined by Stable Diffusion

User Research

First, let’s discuss how AI can aid in research. One of the most time-consuming tasks for UX designers is user research. This includes conducting surveys, interviews, and focus groups to gain insight into how users interact with a product or service.

AI, such as ChatGPT, can assist in this process by simulating user interactions and providing valuable data on how users might respond to different design options. This can save time and resources for the designer, as well as provide more accurate data than traditional methods. Additionally, ChatGPT can assist in analyzing large amounts of data from research, making it easier for designers to identify patterns and insights.

Interface Design

Next, let’s consider how AI can aid in interface design. One of the most important tasks for UX designers is creating interfaces that are intuitive and easy to use for users. ChatGPT can assist in this process by providing suggestions for interface elements and layout based on a provided prompt or context.

For example, a designer can provide ChatGPT with information about the desired functionality of a website or app and the target audience, and ChatGPT can generate several options for the designer to choose from. Additionally, ChatGPT can assist in testing and evaluating different interface designs by simulating user interactions and providing feedback on usability.

Microcopy

Finally, let’s discuss how AI can aid in writing microcopy. Microcopy refers to the small bits of text that provide crucial information and guidance to users in interfaces such as websites and apps. These include labels, button text, error messages, and tooltips. Writing effective microcopy can be a challenging task for UX designers, as it requires a deep understanding of the user’s needs and goals. ChatGPT can assist in this process by providing suggestions for microcopy based on a provided prompt or context. For example, a designer can provide ChatGPT with information about the desired tone and message of a label, and ChatGPT can generate several options for the designer to choose from. Additionally, ChatGPT can assist in providing accurate translations of microcopy for internationalization.

“Conclusion”

In conclusion, AI, specifically language models like ChatGPT, have the potential to greatly assist UX designers in their tasks. By providing valuable data and insights from user research, suggesting interface elements and layout, and providing suggestions for microcopy, AI can help designers save time and resources while also improving the overall user experience. As AI technology continues to advance, it is likely that its role in UX design will become even more prominent. "

My Quick Take

What you just read was created with ChatGPT in about 10 minutes on a Saturday morning before I even started drinking my first cup of coffee.

Is it amazing? Not exactly, but it’s enough of an outline that I could take this an expand it out into something a bit more value. It actually reads like something I would have written in middle school. I literally put in a single prompt to chatGPT and this was what it output on the first attempt. I could have massaged it for multiple iterations, but that’s not the point here.

What chatGPT pulled out of the ether was several areas that it could have the potential to help — places I hadn’t even considered and I’ve started down a rabit hole here…

“Hey chatGPT, how should a modal window inform the user that the information that they are entering won’t be saved unless they have reviewed the terms of service”…

A modal window should inform the user that their information will not be saved unless they have reviewed and accepted the terms of service by displaying a clear message, such as “Please review and accept our terms of service before submitting your information. Your information will not be saved unless you accept the terms of service.”

This message should be prominently displayed in the modal window and should be clearly visible to the user before they enter any information.

Additionally, a checkbox or button that the user must click to indicate that they have reviewed and accept the terms of service should be included in the modal window.

What should the text of the button be?

The text of the button indicating acceptance of the terms of service should be clear and action-oriented, such as “I Accept” or “Accept Terms of Service”. It’s important that the button text makes it clear to the user that they are agreeing to the terms of service and that their information will be saved when they click the button. Additionally, it would be good practice to mention the name of the service or the company as part of the button text, such as “Accept XYZ Service’s Terms of Service” to make it clear what they are accepting.

I think some of my design documentation is starting to write itself with this… and we all know how fun it is to write design documentation.

I can’t wait to keep chatGPT close and hand and see how it might start to help me in my work. Maybe I’ll keep you all posted.

AI is finally getting weird

Artificial Intelligence has been this looming concept for longer than I’ve been alive. It does seem that in just the last few years, with a lot of talk about AI generated art and how ChatGPT is going to impact industries, it is actually starting to make an impact on us. I’m sure it was before, but it just seems much more time-of-mind now (aka making it to the front page of reddit).

Any UX designers or researchers delving into this yet?

Have any of you put ChatGPT to good use or even thought about how it might actually affect UX designers, researchers or writers? I’d love to hear a humans though on this.

in ai user-experience

That’s 6000+ days of my life dedicated to working for one company. My previously longest-held job before that was a year and a half, just for some perspective.

This story is going to be a bit lengthy and I’m not trying to write a tomb of a post about it right now — so let’s start small and hope that I’m motivated to keep writing about my experiences — because I think they’re worth sharing.

There’s going to be a good bit to cover — why I stayed at one place so long, how I made the decision to leave, what I did wrong in that process, what it was like interviewing, putting together a portfolio, and how I came to choose the company that I’m now working for.

I’ll start with the good parts of my previous role and what finally convinced me to get out in pursuit of better opportunities.

I loved my job

My job paid well, gave me the freedom to work on what I wanted to work on and to take any approach that I felt would be of value. Over that 16 years, unlike many of my coworkers, I was able to stare at the ceiling and ponder ideas, facilitate rooms of amazingly smart people to help solve problems and to spend my time working with creative people designing solutions that were unique, innovative, and most importantly — viable in old and new markets.

It was fun, oftentimes exciting, and gave me a balance between my professional and personal life.

Deep work

Siloette of a bird sitting on a branch in front of a sunset.
Photo by lensmatter via Flickr

Last year, at some point during quarantine— I was given enough breathing room in my life to realize where my strengths were and where I wanted to go in my career. I talked about the newfound skills that I discovered in a remote working environment, wrote about not wanting to give up deep, focused work just to go back into the office, and told every one of my bosses that if the opportunity came up to be completely remote going forward, that I would jump on that chance. I was hopeful that they wouldn’t have a problem with me going remote-only.

I was wrong.

My pleas for going remote-only

A boxer dog sitting int he grass looking so so sad.
Photo by cuatrok77 via Flickr

I told everyone — The creative director, the head of design, the CFO, and the CEO. The higher-ups understood, but middle management just wasn’t on board.

So instead, I was asked to start going back into the office. in July it was one day a week, in November it was two.

Instead of embracing a culture that succeeded so well in a remote-only workplace, my bosses were convinced that collaboration and innovation were better suited to be done together in rooms with print-outs and sticky notes when we had already proved the opposite.

The impasse

A close up of a wall panted pink with spackle filling in a large crack.
Photo by Sergio Sánchez via Flickr

In much of my career, I had felt like my voice mattered but, with this, there was no getting through.

I had contemplated leaving many times in the past for many different reasons, but that was the last I could take. I had to part ways and there was only one thing that they could have offered to convince me to stay — Not the pay raise, not the extra bonus, not the promise to lead larger teams — but the ability to work on my terms from my home office.

And that offer never came.

What’s next?

Spoiler alert — I am currently working at a new job that is completely remote with an amazing team, an opportunity to grow my skills, and doing work that will make an impact. I couldn’t be much happier.

One of my goals this year is to get back into writing more. Much more. I have a lot to say about how I got here.

in business user-experience