Aaron Weyenberg AMA

Thanks Jen and tell them I say hello!

Hi and thanks, Jesse.

Pretty much everything the product team does here is in service of the content and mission. We are always pushing and pushing to expand the content around the world. But in terms of the brand, no. The design and product team stays out of the way of the content. The brand is centered around ideas.

TED.com is about 7 years old now. The content has outgrown its ability to scale with it. So we are going through a redesign of right now, and we established early on that the tenets of the TED brand and design ethos were right as they should be -- and that a redesign of the site shouldn't change any of that. You'll see some new features in the upcoming design, and the site will look quite different than it does today. And we think that's ok. But you'll see the content (ideas) is still the star, and everything else is in service of it.

I think a successful design of TED will make people think "Wow, that is a great talk and I want to share it" and not "Wow what a great design".

So while I'd say yes we push for bigger things, the brand isn't really one of them. Ways to amplify the ideas therein is what we are always looking to push.

Hi. Good question because I think there are a lot of ways to answer that.

Here's my take: I would avoid falling into the comfortable zone of thinking of UX as deliverables (a set of wireframes, designs, process flows, etc). That's just part of it. UX continues indefinitely in the form of changing, testing, improving, iterating and measuring. Also do not think of UX as design/UI. Any component that affects a user's emotional response to using a product, feature or service is the domain of UX.

If I was starting out, I'd spend some of my time gaining a sense of how people think and what influences their expectations about things they interact with. This can really be powerful in shaping a product to align with how our brains work (quirks and all). I'd also enter it with humility. Let data inform your decisions. Over time you'll know when it's ok to zag when the data says zig.

Good luck!

Thanks Mark. 

Well, there are so many so I'll give you a few.

Ben Schneier talks about security and how our notion of it just doesn't match reality: http://www.ted.com/talks/bruce_schneier.html

Bryan Stevenson will make you think again about the justice system in this country: http://www.ted.com/talks/bryan_stevenson_we_need_to_talk_about_an_injustice.html

And Dan Pallotta will almost certainly change your mind about how we think of charities: http://www.ted.com/talks/dan_pallotta_the_way_we_think_about_charity_is_dead_wrong.html

Barry Schwartz's The Paradox of Choice: http://www.ted.com/talks/barry_schwartz_on_the_paradox_of_choice.html

Amy Cuddy on body language: http://www.ted.com/talks/amy_cuddy_your_body_language_shapes_who_you_are.html

And no discussion about favorite TED talks can be complete without Ken Robinson: http://www.ted.com/talks/ken_robinson_says_schools_kill_creativity.html and http://www.ted.com/talks/ken_robinson_how_to_escape_education_s_death_valley.html

SEO is actually a pretty weak area of mine. The idea I usually subscribed to is that SEO is about optimizing for ranking and inbound traffic etc and _not_ optimizing for users experience once they get there. I'm not sure those are conflicting forces necessarily. But for the sake of this discussion let's invent a new term: UXO, or user experience optimization.

So in my mind the throughline between SEO and UXO would be the content. UXO being concerned with its sensible (even enjoyable?) organization and presentation to a human being, and SEO to a machine. Does optimizing for UX have a splash effect that assists SEO too? Maybe. I don't know enough about SEO to say so.

Can you tell me a bit more about what challenges/complexities you face in this area at your org?

Not sure I can give a satisfactory answer to that. Is it just the presentation layer that is changing?

I should add, that in all our TED job postings, we end by saying "we only hire extremely nice people." There are no big egos around here and we love to keep it that way as we hire a few more folks to fill some gaps. :)

Hello Jen!

I think like a lot of opportunities anywhere there's a timing factor. TED was in need of a UX person with interaction design experience and some technical/dev skills. My whole career has been about that mixture. I've been fortunate to have the advantage of some CS and coding background mashed with communication design. I have sort of a frankenstein academic and working background that helps me understand various platforms, their strengths and limitations. Am I the best designer in the world? Nope. Would I be hired to build a CMS to power the forthcoming redesigned TED.com? Nooo. But knowing the vocab and having tinkered in these various fields enough puts me in a unique area. That's why I make the swiss army knife comparison a lot.

UX is a really broad field. And knowledge about testing, prototyping, user research, design, typography, copywriting, engineering, IA, etc, all play into it. Too much stuff for one person to have super deep knowledge and experience in, but a cocktail of these seems to be a strength.

To return to your last question, this is of course what everyone strives for. I'd say the biggest asset any UX designer can have is an affinity for empathy. Being able to predict and relate to how a user will feel when they interact with your product is what the field is really all about. I think of UX as a broad set of underpinnings that HAVE to work (starting with performant back and front-end engineering) on up to a layer of delight (language, unexpected polish without trying to be too clever). It's not too unlike Maslow's hierarchy of needs.

Learning UX, like anything else, means making mistakes, too. But that's a good thing, provided you are measuring your mistake and learning from it.

I think too many people think a "UX person" is a wireframer. They're not. That's part of architecting a product that has a good chance of giving a good experience, but that's just one constituent of the UX spectrum. UX is about predicting and knowing pain points in a process, knowing things about human behavior and expectations, and identifying and removing friction. After those needs are met, then you can get into things like delight.

Sure. For the sharing control test we were looking for clicks on those elements (since we can't measure if the share actually happened on the other service). I think for each test we are trying to nudge something up, with the ultimate goal of being increased video views.

In the case of the Surprise Me feature we were really just looking for how many users clicked the surprise me button (and on through to the 'Watch playlist' button). Since we know from other tests that we get almost 2x views/session when users are in a playlist mode of content consumption, the click through rate of the Surprise Me generator was the metric we were looking to increase.

But yeah, ultimately it is views per session and session length as well. The longer we can keep people engaged, the more likely it is they will discover a talk or an idea that resonates with them and sparks their curiosity.

We also look at bounce rates and compare that to how people arrive on the site. The assumption is that if a user clicks through FB or Twitter or some social media channel, that their interest in the topic is already piqued (and the content probably validated by a friend or other respected source) and they are more likely to watch the full talk. If they are just wandering through the site the abandonment rate is a bit higher.

One other things we are eager to measure is when people share talks. Most of our talks are pretty long (up to 18 minutes). That is a long time to keep people engaged online for any type of content. So we're interested when people share the talk (within the first minute? at the end?).

What would I love to test and change? Headlines (titles) for talks. I want to be able to publish a talk with 2 or 3 headlines, written by our editorial staff, and see in real time which one is performing the best, and then quickly ditch the underperformers and continue with the winner. Language and words can have so much impact on hooking someone into clicking through to watch a talk.

Another thing I'd love to test in real time is photography of the talks and their effect on talk viewing selection. Just about all our talks have what we call a "stage shot" of the speaker. Example:

http://images.ted.com/images/ted/4a7be98e8b1380e18ee758b3887f250683ba04e9_240x180.jpg

I'd like to be able to change these and see if that affects click through rates. We need a more sophisticated image server to do this, but we'll get there eventually.

TED is a non-profit organization. Revenue comes from a few sources including conference attendance fees, display and postroll ads and other partnerships on TED.com. The conference is expensive to run, the bandwidth we serve from TED.com and mobile apps run us a pretty big bill, and many other costs are associated with keeping TED going. After keeping the lights on and running things, surpluses (if any) are invested into the org's mission.

I can think of a few things, but the one that I think is most notable is a feature we built called Surprise Me on TED.com. You can find it about halfway down the front page. The goal was to port a feature on the iOS and Android TED apps (called "Inspire me") over to the website.

We thought this would be a pretty successful feature. It was, but not quite in the way we expected. Users engaged with the feature at a very low rate. Much lower than the native mobile apps. While that was disappointing, what we learned was that the user behavior _after_ they engaged with it showed a much higher views-per-session metric -- almost double the standard. This was a key finding for us. The reason for the higher views was the format in which the Surprise Me results were given to the user -- in playlist form (as opposed to a talk in isolation). We gave the user a cluster of related talks with the next talk in the series autoplaying.

This set the stage for the TED Playlists feature we built later. We created TED Playlists with the same intention in mind, and we experienced a similar bump in views per session. So out of a "failure" emerged something that significantly improved a key metric that serves the greater purpose of ideas worth spreading.

Yes we do.

One test we ran was placement (and size) of the sharing controls (Twitter, FB, etc) on the page that we call the Talk page (where the a talk video and comments are displayed). We actually ran 2 variations against the untreated version. One of those variations was a clear winner, even though we expected the other variation to win. So intuition can be strong, but data frees us from overconfidence and tell us what is.

Another test we ran was for a feature called Surprise Me. We tried a different placement on the front page to see if it would outperform the untreated version. It did, but we chose not to implement it for other reasons. Sometimes data can tell you what happened, but doesn't tell you if implementing the change is ultimately a good idea given other factors.

The reality is that soup-to-nuts design process here is rarely identical in any two cases.

Sometimes products or features from TED arise out of an internal idea (e.g. "Hey wouldn't this be cool?") or a need (e.g. "This is broken and no longer serves its original purpose effectively. It needs to be replaced/fixed/abandoned/whatever.") or a request from our community of users (the Open Translation Project is a great example of this). 

Yes, the Steve Jobs style product design process (people don't know what they want until you give it to them) can work, but sometimes the greatest things come from ideas from the community of people we serve. Some of the most far reaching and impactful things TED's ever initiated (e.g. OTP, TEDx) have come from wants and needs from the community we serve.

Idea to implementation

TED is a small team. Sometimes we can't build what we want internally so we have a trusted stable of talented contractors that help us. But the main way we get things done now is more iterative. We've gone from a waterfall process to a more agile one involving rapidly trying things to see if they are viable. Sometimes things "fail" (quotes intentional) and sometimes they don't. But in all cases we do our best to measure and learn and adjust.

Questions we ask ourselves

This may seem cliché, but we really do ask ourselves one core question with just about everything we do: How does this serve our mission of Ideas worth spreading?

On pitching value of change

This is tricker. Right now we are investing a lot of energy and resources into paying off technical debt and building backend services that will power the future of TED.com and other properties. This can be a challenging sell because people don't see immediate results. But it's of topmost importance. The returns come later and it takes reminding people that we can build things right or build them fast. But we don't change things that go against the grain of the mission. We don't compromise there. That is the common understanding we all have and that everything we do is held against.

... I'll come back to your other question.

Here are some sources that my colleague (@mmcwatters) and I follow:

Baymard Institute: http://baymard.com

Bjango: http://bjango.com/articles/

Brad Frost: http://bradfrostweb.com/blog/

Little Big Details: http://littlebigdetails.com

Matt Gemmell: http://mattgemmell.com

Smashing Magazine: http://www.smashingmagazine.com

Boxes and Arrows: http://boxesandarrows.com

Thanks Patrick, it is a challenging and fun gig.

For the purpose of my answer, let's define "design" as actual visual design -- typography, fonts, colors, photography, layout, contrast, grids, etc etc.

That said, I would say this to such a person: UX is a very broad field in which design is only a piece. UX is not graphic design. It isn't interaction design. It isn't wireframing or information architecture. It is a blend of everything that affects a persons emotional experience with a service or product. That means it has to be error free and fast (engineering), it has to speak the right language (copywriting) and it has to present information sensibly (IA) and many other things. I think you can come into the field of UX from any of these constituent specialties, but the goal of a UX practitioner is to become competent in these other components and to see how they fit together to create a product that people feel good using.

We have a mixture of quant and qual data to inform us. Sometimes quantitative data is all you need to make a decision. A/B tests are a great example of that, but sometimes they don't tell the whole story of why something is happening. That's when qual data can help a lot -- but that is more time consuming and expensive to gather. Depends on what it is you are trying to uncover.

Hello humans!

My business card says "UX Lead" but really I'm a swiss army knife doing various things like front-end work, prototyping, design, and testing. You can ask me anything about those things, our UX practices, TED's culture, mission, people, history, and initiatives.

I'll be less likely/able to answer questions around content, curation and editorial, but you can ask if you want anyway.

I'll be online from 10:30am to 12:30pm EST Monday September 30.

Feel free to upvote questions you'd like me to notice first. I'll do my best to answer as many as I can!