Product and design trends that will shape 2018

What product themes and technology are most likely to define the year ahead?

Before kicking off our own goals and resolutions in earnest, I sat down with VP of Product Paul Adams and Director of Product Design Emmet Connolly to chat about where our industry is headed in 2018. We cover everything from trends like voice UI, augmented reality and AI, as well as the new roles software companies will need to hire for and whether product building has truly become commoditized.

Some of these predictions might hit; others, well, we’ve been wrong before. But if you’re curious how our 2017 predictions turned out, check out our year-in-review conversation.

If you enjoy the conversation, check out more episodes of our podcast. You can subscribe to it on iTunes or grab the RSS feed in your player of choice. Below is a lightly edited transcript of the conversation.


Des Traynor: First off, it’s January 2018. Do either of you plan your years? Did you have your set of goals or New Year’s resolutions? Do you share them with people? How do you do it?

Paul Adams: What I’m about to say is super corny, so you’ll have to forgive me. I’m a huge believer in goals, and yeah, I write goals before the year starts. In mid-December 2017, I started thinking about 2018 and wrote down some draft goals. In my heavy experience, when I write things down, they tend to happen, and when I don’t, they tend not to. Years ago my wife and I wrote down life goals. We wanted to live by the sea, get a dog, buy a certain type of house and live in a certain type of neighborhood. Lo and behold, because we had a list of things to aim for, they happened.

Des: How often do you revisit your goals?

Paul: It depends. Every Monday morning, I don’t go to the office, I go to a coffee shop on my own. I’ve had the weekend to kind of decompress, and I think about all the stuff in front of me. What’s the most impactful thing I can do this week? I’ll transform that into a set of goals for the week.

Des: What about you, Emmet? Do you observe the discipline of goals?

Emmet Connolly: I certainly do on a daily and weekly basis. In regards to the year, it was less goals and more of a concrete set of intentions. I don’t think at the year level I’m quite ready to put a specific target that I can hit or not hit at that scale.

I did an exercise at the end of last year with my team, where we put together all of the things we can improve and came out with five themes. We were aware going into the holiday break of what to mull over while in that big decompression period. Now, we know what it is that we want to tackle, so it’s a matter of really breaking that stuff down.

I think it’s just good for everyone to come into the new year feeling aligned and knowing what we’re all going to try and solve together. It’s the same thing as doing your goals on a Friday. You don’t come wandering into the office on Monday morning and wonder what you’ll be up to that week. It would be lunchtime on Monday before you figure that out.

Des: Say something big goes down on Tuesday, like some huge product news. Does that change your goals, or can you push it to next week?

Paul: You shouldn’t be totally dogmatic. If something happens, it happens. You need to reprioritize. These things are aspirational and not a set of tasks. A goal isn’t something like present at the All Hands. The All Hands is happening, and you’re going to show up because it’s your job. They’re aspirational things – what would a great week look like for me? It’s if I get X done and Y done, that’ll be a fantastic week, and sometimes that doesn’t happen because I get dragged into other stuff.

Emmet: I always nerd out a bit on the personal productivity side of this, everything from “Getting Things Done” by David Allen, to Inbox Zero and bullet journal. For me, an important one is staying on track and not getting distracted by whatever’s flowing into your inbox on a daily basis. The other thing is to have a reliable system. We all process so much information on a daily basis, that if I didn’t have a bucket that I could put a lot of that inbound stuff into, that I could then later come to and prioritize on a daily and weekly basis, I’d just forget everything I need to do, and I wouldn’t be able to cope.

Product roles that will rise in importance

Des: I’m guessing one of your goals for 2018 is hiring. What are the new roles, or the areas you think software companies will find themselves looking to hire in, that maybe wouldn’t have occurred to them before? What’s new here?

We’re folding content strategy into almost every aspect of how we design products.

Emmet: In my team, the product design team, we’re moving into the stage where we can fold specialists into what we’re doing. We’ve had lots of generalist product designers, and now we’re having things like UI design roles opening up, allowing people to just specialize, and go really deep. Most people are naturally T-shaped, but we’re explicitly hiring for that now in some ways.

The other thing we put a lot of work into, especially in the latter half of 2017, was content strategy and making that a real integrated part of the product design team. For us before, and at most companies that I’ve come across, it’s treated as more of a service. You could call on the content strategy folks if you have something that fits them. Now, we’re really folding them into almost every aspect of how we design products, to the point we’re actually toying with calling it content design instead. We’ll embed them in product teams once we reach the mass that allows us to do that. To get there, we’ll have to share content designers across different teams.

Paul: The content design versus content strategist title is interesting. One of the big areas that we work in and have learned a lot about here, is machine learning and bots, because we developed our Operator bot last year. It made me think that there’s other new jobs there too. Content strategy isn’t a new job, but there’s new types of skills that are even more important in other contexts.

One of them is the content that you have machines learn over – these big data sets and what’s in them. It could be text, bodies of content, conversations, etc. You hope that you’ve written an algorithm or some code that can run over this data and learn and get in these feedback loops, and yet, machines are contextless. They’re not people and don’t understand where things came from.

An important job will be to have someone, a human being in some kind of management role, oversee all this stuff. They’ll look at the output of all these machine learning systems and course-correct them. Maybe there’s an analogy to factories here. When the factory line became automated, there were still people in the factory line, making sure everything’s working and the door isn’t on the bonnet of the car.

Projecting the ceiling for Voice UI

Des: Let’s talk about voice UI. We did a podcast at the end of 2016, where we all predicted in some way, shape or form, that voice would be huge in 2017. We were definitely wrong if the criteria was huge, but it definitely didn’t flop. Amazon’s product had a good Christmas, while Siri hasn’t gotten much better. Certainly Ok Google is still the same Ok Google. What do you think we’re waiting for here?

Emmet: More than it being some silver bullet thing that we’re waiting for to arrive, why are we waiting for something? Why do we have this crazy expectation that everything is going to be a mad, world-changing technology once it matures to a certain level, which is certain to happen within the next 12 months?

Maybe there’s a certain class of new technologies and products that won’t necessarily change the world in the way the smartphone has, despite the fact that we all might look forward to and be excited at the prospect of that. Maybe we should be more comfortable with the idea that there’s a natural limit.

That said, I don’t think voice has hit a natural limit, or that it’s for timers and music and nothing else. But it might be the case that it won’t be the next UI. In all probability, not every technology is going to be the next UI. Some are going to be handy little additions to our everyday lives; some may be world-changing new technologies. You can have a guess at what those might be, but it’s a lot harder to say when they’re actually going to land or if it’s going to be a change that takes decades, as in retrospect, the internet took. Or a year, as smartphones took.

Des: I’ve long said that, there’s a very clear use case for voice, which just isn’t that big. It’s eyes busy, hands busy, I’m driving or I’m cooking or whatever, and I just want to take a simple task and execute it.

You mentioned, very briefly and very naturally, timers and music. What’s unique about those two flows that make them amenable to voice?

To remember and utter the incantation has a tiny bit more friction than pulling out the phone.

Emmet: Generally speaking, the failure rate of, “Des, I didn’t understand what you said when you said set a timer for three minutes” is pretty rare. The voice recognition has gotten really good, and there are really simple tasks that you are confident it can complete. But other things are a little more complicated, even to the extent of ordering a cab with an Echo. I spent ages switching on the integration and trying to remember the specific phrasing needed.

For a lot of those things, to remember and utter the incantation has a tiny bit more friction than pulling out the phone. That’s the difference, at least in my personal use.

Des: There’s something about the diversity of the command. Most people say, “I wanna set a timer for five minutes” very similarly. On the other hand there’s many different ways to say “I would like to travel by public transport to this building.”

Paul: As Emmet spoke, the thing I thought of was AirPods, which I bought last year. I went through the same path as almost everyone, which was some version of “You look like an idiot if you wear those things. They’ve got Bluetooth headsets written all over them. Get headsets. I’m not getting them.”

Eventually, I bought them in the summer, and lo and behold, they’re a life-changer. It’s the friction thing you mentioned. Suddenly, you don’t have wires and it’s just easy. Now I can’t think of going back to wearing clunky, wired headphones.

I thought of AirPods here for two reasons: one, some cost-benefit thing switched and I think voice needs those types of things. Two, obviously AirPods are a means in which you control voice UI. So, are little pieces of the ecosystem being built at the moment? I can imagine a world where I have AirPods in all the time, they’ve reached critical mass whereby I feel less self-conscious about wearing them, and I can talk to them and do simple things. I wonder if it’s this year or next year, where we look back and say, “They were the three or four connect things that had to happen.”

Des: We obviously have timers and playing music nailed, but I still wonder if people will, for instance, ever listen to their email. Or, if you can’t look at your phone, and this is all that’s going, you might listen to Twitter. But it’s just not that great in a robotic voice.

One genuine ingredient here will be better computer voices, for lack of a better word. On a five year time horizon, can we record our own voices, so that when I send you a text message and you press play, it’s actually read in my voice? That would totally unlock this new use case. Yes, I do realize that’s also known as sending an audio file, but the idea is that it’s multimodal. You can choose to read it, or you can choose to listen to it.

The massive potential of augmented reality

Des: When it comes to augmented reality, Magic Leap is the mystical, “one day maybe they’re going to show up and change all our lives” example. But in general, there are augmented reality offerings of sorts on the market. Does this actually matter, or is it just still going to be a weird, niche case this year?

Paul: This one is clear to me. With AR, I think there’s a future state, which isn’t that far away, where we effectively would all have augmented reality in all parts of our everyday lives and we won’t be able to look back. Like in the same way that today, it’s hard to imagine a world where we didn’t have a smartphone, didn’t have the internet in your pocket and didn’t have instant connectivity to almost all the world’s information and all the people in your life. You could never go back to that world.

After years of hype, Magic Leap revealed its AR headset in the final days of 2017.

Kind of like AirPods, I think once someone figures out the hardware, and once your experience is augmented, people won’t go back. I can think of all sorts of use cases. Maybe you’re going running and there are literally crowds. You could be running around an empty stadium, but in augmented reality it’s full and there’s cheering. You’re playing soccer and suddenly there’s other famous footballers you’re playing with.

Emmet: It strikes me that talking about things like watches and wireless headphones, they’re limited enough in terms of the bandwidth of media they can deliver to you, either tiny screen or audio only, that they’re certainly useful additions and they’ll enable you to do maybe more of certain things or some small new things.

For example, I also got AirPods and I listen to podcasts way more because you can just pop it in and it starts playing and you’re good to go. The friction is removed. There’s lots of little watch use cases like that too. But, the promise of AR is it literally has the potential to be all-consuming in terms of the number of things that it can augment your reality with. The number of use cases probably vastly outstrips those other things I just talked about.

It also seems that there’s just going to be a tipping point, if you believe in the inexorable crawl of progress around technology, that we’ll reach at some, where it’ll become better and good enough. VR seems like it’s closer to that tipping point, but of course, you’re not going to wear a VR headset all day. But for whatever limited adoption that VR might achieve compared to AR, it’ll almost certainly happen sooner.

Des: I learned a phrase recently, prosthetic knowledge, which is the stuff you don’t know but you know how to find out immediately. It’s basically the knowledge that’s one order removed from your actual brain. Thanks to our phones, we all have this insane amount of prosthetic knowledge. I can imagine a world where you meet somebody for the first time, but you know a lot more about them because they too are wearing an AR headset.

In general, if you have two choices, you can view the world the way we all do today or you can view the world with all of this extra information immediately mapped onto it – whether it’s personal information relating to humans or you look at a restaurant and you can see it’s TripAdvisor rating beside it – that will be a one-way switch. I don’t know how far away that might be, but I can’t imagine people frequently taking off those glasses or whatever the hardware might be.

Paul: That’s how I feel about it. Commerce is a huge one. The things I mentioned earlier are a little bit out there. They’re really rich-media oriented. But imagine if you could filter this stuff, whereby, as you’re walking down the street, there’s notifications effectively from the stores you like, saying, “The thing you have in your shopping cart on our website is in the store.” Small utilities will make people flip.

Emmet: I just want an ad blocker for the real world.

Drawing a line between AI and bots

Des: AI and machine-learning has crossed something where we’re no longer talking about it in a future tense. Obviously, we have our own horse in the race as well, Operator, but do you think 2018 will be the year where it’s totally acceptable to go to a business and talk to a bot?

Paul: We need to make a distinction between bots and AI. Bots were in the hype cycle in 2017, and they came out of it throughout the year. Now we’re in a place where it has matured, in some sense, and in 2018 you will see a lot of useful bots. AI is still, in my opinion, in buzzword land. People say AI and don’t actually know what they mean. Other companies say they have AI, and they don’t at all – they have simple programming systems.

We need to make a distinction between bots and AI.

For most businesses, I think the AI theme for 2018 will be to learn what it actually is and is not. We’ll learn to distinguish between companies that are selling you up the river and the companies who are real.

Emmet: And learn how to apply AI, right? The naïve view is to sprinkle some of that magic AI dust on top of your product and things will get way better. It’s actually really challenging to build products and solve real problems with this stuff. You have to learn bit about how it actually works in order to learn the difference between two slightly different approaches to the same problem. One could be a three-week project using off-the-shelf technologies, and the other could be a nine-month research project with no promise of success at the end of it. Figuring out how to build with this is a whole other thing.

Des: Emmet, outside of all these tectonic shifts in technology, more practically, what are the big trends that you think will change product and product design in 2018?

Emmet: One thing in the product design corner is tooling. There’s a lot of healthy competition going on between tools like Figma, Framer, Sketch and InVision. Adobe has their thing. There’s a lot of competition to become the tool of choice for design teams everywhere. On top of that, building on the topic of AI, a lot of these companies are looking ahead at opportunities to fold some of that in.

Honestly, if you look at how a lot of design work gets done, it’s still an incredibly manual process. We’re building design systems to make it slightly less manual, where you’ve got some pre-build components that you can work from, but it’s still extremely labor intensive for such a high tech thing.

Airbnb interestingly built a simple system recently where you can sketch a little wireframe on paper, snap a photo of it, and using image recognition, it would actually compile and build a working app based on that sketch. That is a promise of an AI-augmented workflow, although that’s probably oversimplifying it. But those designs will need to be supervised a great deal, and tweaked at the end.

The commoditization of product

Des: Paul, what’s your take on what are the trends that are affecting product in general?

Paul: What’s been on my mind is that you can consider product-making to be somewhat commoditized. What I mean by that is, if you look at your industry, there’s likely five to ten competitors, and it could be 100 competitors. There could be a 1,000 to-do apps, for example.

Messaging apps are just one of many examples where near-infinite competition exists.

It’s very rare to not have competitors that are good. They’re good products, they’re built well, and they’re designed well. Our industry has matured to a point where there are genuine competitors in your space, that are good, so the basics are somewhat commoditized. As a product maker, you then have to ask, what fight are you fighting? What’s the frontier? Is it better workflow design? Better interaction design? Something else?

Des: And that “something else” could be a better brand, a better platform or even a better connection with customers. In a sense, you have to design futures assuming that if it works, everyone’s going to have it. And if it doesn’t work, then no one’s going to have it, including you.

Paul: Facebook copied a lot of their competitors last year, and a lot of people criticized them for it. I actually think, first of all, they did the right thing for their business. Second of all, they were simply doing what everyone else is doing. Everyone’s copying everyone. They just got called out for it.

There’s no such thing as a new idea, it’s simply a new combination of two existing ideas.

People think that copying other companies and taking what they’ve done and taking it for yourself, in your own way with tweaks and so on, is bad. I don’t think it’s bad. It’s fair game, and it’s how the world has always worked. One of my favorite ideas that has dominated a lot of my career is that there’s no such thing as a new idea, it’s simply a new combination of two existing ideas. In that regard, this isn’t bad. People should just adopt it, embrace it, move on, and understand where the actual fight is.

Emmet: There was a lot of pearl-clutching going on when, for example, Instagram came out with their Stories. That was clearly, at least initially, directly the same as what Snapchat had been doing.

But, as someone who really enjoyed Instagram a lot this year – when there were probably a lot of parts of the web and products that didn’t give you the warm and fuzzies to open – it made the product better. They subsequently iterated like hell on top of that flat rip and made that feature a lot better. I’ve actually spoken to some people who were working on some of that stuff and they said, “Look, the way we view it is that this was a new type of media. Fair play to Snapchat, they were the ones to discover it, but it exists, it’s out there in the world, and it’s there for everyone to use.”

Look at how desktop computers evolved and smartphones evolved, especially between Android and iOS. Over the course of a decade, they borrowed wholesale from each other. That rising tide made everything better. In some sense, it’s part of competition as well, right?

Des: If this is a “new style” of product development, and basically, all product will become commoditized and we can replace Paul as VP of Product with one that’s far cheaper, then how do companies prepare for it?

Paul: This is new to us in the technology industry or the internet industry, but it’s not new. All cars have wheels, they’re all round, they’re all on axles. All toasters look the same, and all fridges look the same. This isn’t a new idea.

There’s some curve of invention for everything, in every industry, and it reaches a point where it’s mature. For example, to-do apps work the way to-do apps work because there’s not much invention left there. I think that’s true for a huge amount of software. Google Docs versus Microsoft Word, where are we at? Are either of them about to radically rethink document creation? Probably not.

First accept that you will be copied, and then two, adapt your plans. Think about the fact that you will be copied as you’re building and designing the thing.