Information architecture today: Meetup recording

I recently talked to the Preely meetup about what information architecture looks like (to me at least) at the moment, how it differs from the way it was written about in books from the past (including mine), and what skills are needed for it. Of course, I don't see everything that's happening around the world, so it's just based on what I see people talking about at conferences and what I'm asked about frequently. If you're wondering what information architecture is and what skills you might need, check it out.

Transcript below the video (edited transcript, that is. Turns out I say 'like', 'kind of' and 'you know' far too much, often all together)...

Transcript

Tina: First of all good morning to you all I'm Tina Øvad and I will be your host today and today we have a Donna Spencer that will share her knowledge within information architecture and again we have a series of people so many people joining and it's just great to see that there's still  people who want to join these online meetups we have a lot of zoom fatigue we know and we are looking into having some of the meetups physically as well. I hope all of you have had a lovely morning so far and i guess it's good evening to you Donna.

Donna: it's good evening to me yes

Tina: yeah that's great, so Donna is a product designer at MakerX and she has extensive experience in user experience, service design, workshops facilitation and information architecture and you've also written five UX related books and was the founder of UX Australia and ran it for nine years. And to be honest I'm also one of the fan girls so look I'm a bit starstruck today to be honest.

Donna: and then i go make faces at the cat!

Tina: I love that because then I was just like oh she's human that's nice. So the agenda for today is that when I'm done with all the practicalities then I will start the talk you can ask questions in the Q&A section here and we will go through them at the end of the session. And today we actually have a bit of more time that usual to have this Q&A section so i hope that you are up for asking all your questions related to IA and navigation design and so on to Donna. We are recording everything and we'll share the recording during the next week together with Donna's slides. And then of course for those of you who do not know Preely - Preely is a self-service platform for unmoderated remote user testing and also user panel hosting so feel free to reach out to us if you have any questions or maybe if you would like a demo of the different services that we provide. So I actually think that was what I have so Donna I think just take it away.

Donna: okay I will grab some slides. Here we go, share it, there we go that worked okay. Nice to meet you all virtually I mean we're not meeting but I feel like every time I see a bunch of names on a screen that counts as meeting you all somewhat. And so what i'm going to do today is just talk a little bit about what information architecture is like today and the things that we might need to know and do for it. But i won't give away the punchline too much just yet. So as as Tina said I am working as a product designer right now. I've done a lot of design work. I've had a fair emphasis on information architecture in my career but I don't do it, like I don't do it as a job anymore because yeah some of the stuff I'll talk about - as things change like work just disappeared. IA projects in themselves disappeared so instead I kind of  use the skills I have on all kinds of different projects. And the two information architecture related books I wrote one is about card sorting and one was called a practical guide to information architecture. The others are about presenting design and facilitation and web writing so still UX related but not so much information architecture.

What I thought I would start with is what information architecture used to look like so that we can do a bit of a compare and contrast to what it's like today and not just to reflect on the past but because this actually really matters for a very specific reason.

So information architecture and, I'll talk about 15 years ago. At the time when most of these foundational information architecture books were either in development or written - I think some of these are 10 years old some are 15 to 20. But at the time these were written and were the canon of information architecture as a field we were primarily working on fairly heavy content sites so content-based things. I'll say things because it could also be intranets or document management, content management kind of things as well. They were not necessarily static but there was often a big static element. Like the things we worked on, and this is before web 2.0, were things where like we would write a bunch of content and we would present that content to people.

And so information architecture then was about figuring out ways of organizing the content, of structuring it of helping people understand. You know what they needed to do so it was a very information retrieval tasks at the time. And so what we were often working on then was making hierarchical content structures, working on things like taxonomies for larger more dynamic sites. We spent a lot of time thinking about labelling or what we called things and what went in navigation bars, and how we how we labelled ideas and concepts and sections.

And through that period it was a really practical discipline. All of these books, and actually any other information architecture books through this period were also very practical. And that was fantastic - like that is literally exactly how it should be because it was a really pragmatic field with a bunch of people who were new. We were figuring it out from scratch. We didn't have a big field that we came into. We were figuring out what navigation meant how was the best way to present large amounts of information, how do we get people to write for the web, so the books were really, really, really practical. And honestly, I'm never going to criticize a book for being practical.

But what they didn't have, mine included, was theory. I've got some - I've definitely got some theory in my book - I've got a really big chunk of theory around categorization and how it works in our brain, and cognition so some really really lovely theory there. But they don't have a strong theoretical base.

So when we move through time and a theme changes as it will, and I'll explain where I think information architecture has changed, these books don't necessarily help you. You definitely can pick them up and get things out of them. There are definitely principles and good stuff in them but you can't necessarily pick them up now and apply them to a project. So that means that it's hard to use an old text for a new problem even though within them there are definitely some some good principles. So that's kind of what I think has happened.

I said that information architecture at the time was very much about content tasks. Since then things have changed. I'm going to talk to you a little about what I think has changed and what that means, and how it's changed and you'll see some patterns here.

IA now and even for the last kind of five years - I would say is quite a different beast to what I described there with you know fairly static content and us really helping people with content tasks.

One of the places that i see information architecture still existing is in this world of headless CMSs. So in a headless CMS, the content management system holds all the content inside in a big bucket and it's separated from the display. So then the content can be displayed in theory in a number of places. In order to make your content work in a headless CMS you really really really need to be able to think about the structure of it. And you really need to think about it in abstract outside of its visual presentation. In the old days we would design things like navigation and page layouts and often designers would do those first and then the content kind of had to fit in. This way we're really thinking theoretically about structure and then having content  displayed. That's just literally a different way of, a different style of project and different way of working. I've not worked on a lot of headless cms projects. I do see some uh conversation that they're overused - that not all projects need to be as headless as they are, but there is an advantage sometimes in really thinking about structure. So that's one way that information architecture is different today.

Similarly I'm starting to see people again, after this gap starting to talk about structured content. There was a a really good structured content conference only last week. And here we're both you know thinking about kind of that headless thing of being able to figure out how your content is structured and then later on figure out how it's displayed but also very much about reuse - not writing a similar thing in 10 places but writing at once and having it displayed in different parts of a content set, often in ways that you - you don't actually draw the pages out and like draw what the content looks like on it you have to think about it abstractly and then figure out how it goes. I'm screwing up my face because it's a really different way of thinking. It requires quite a different set of skills to do this in abstract. I'll talk about last what kinds of skills do we need to work in this kind of new style information architecture.

Some of us are working with very large collections of content- so that might be really large catalog sites for e-commerce, really large catalog sites for things like you know museum collections, really large content sites for marketing. When things are huge again you don't go and design every single page. You have to design in abstract and you have to design from a structured model. That's quite different to creating hierarchical navigation to display content. It's much more about the deep structure.

You can hear the similarities between that can't you? What all of these mean is that instead of designing every page we're designing a lot of templates. I think the time when I realized this had changed so much is - I was trying to teach IA and I was talking about how the way that we have to think is that we figure out these structures so that it goes into templates and somebody said I don't know what you mean by templates. I'm like oh - so templates. You might be doing that as well or you might be like yeah move on I know this. But can you think about  a supermarket website or something like etsy or any kind of ecommerce. Again we don't design every page, we design the product page, we design the marketing page and then the content that we have has to fit into that structure. So instead of designing pages we're designing templates and when we're designing templates we need to have our content structured in a way that we can slot it in.

Recipes are a really good example of this. If you think about a recipe - apart from thinking about how recipe bloggers you know write this much boring content before you get to the recipe - but a recipe is a beautiful structured thing. It has a title, a time, a list of ingredients, a method, equipment that you need, a cuisine, a bunch of things like that. It's a beautiful structured thing and so you can design a template for a recipe and the content structure and pour the two things together.

That's what we're working on more often - the templates for for the content to fall into. What this means is that we're working at really two quite different levels of structure. We're both thinking about like the macro structure - so I've got a site full of recipes- how am I going to help people narrow down and find the ones they want, how is the overall navigation going to go, how do the you know search and filters and narrowing down work? At the same time as working with that kind of template and page little structure so we're working at two ends both at once. We probably were always doing that but it's an interesting thing.

I wrote this on here - not quite sure how I fit this one in but content marketing and search engine optimization have basically killed our ability to find good quality content. We used to think a lot about how to help people find what they need. I probably should just add google to this as well. We used to design as if people came to our website and then we helped them from there. Now people are starting at google this whole content marketing that's happened over the last 10 15 years means that google prioritizes new content over old, prioritizes long over short (long isn't necessarily better). It's really like it's really hard for us as designers to help people find the things that they need. We are just awash in crappy crappy crappy content and I know this because i teach user experience design at a postgraduate level and I'm always trying to go and find really good quality articles and I can't find them anymore. I need to find a really good persona template that isn't a rubbish one, a really good journey map article that isn't a rubbish one. All I can find is the rubbish ones so how my students need to actually learn anything... I don't know how IA this is but it certainly uh affects findability and how it presents information to people and how we help them learn.

Another thing that we need to do here is automate relationships. So if you think about - I was talking about recipes before we might have a recipe site and we've got our lovely structured recipe and a template and then we're going to give people other things to move forward in case this recipe isn't the one they wanted - we're going to maybe show them some related recipes. We're not going to go through every recipe in our database and say well this sourdough recipe is like these three other recipes. We're going to have an algorithm that says how to figure out this so we might say if it's a bread recipe show one bread recipe, one sandwich fillings recipe and one soup recipe or something like that. Instead of doing things again manually and figuring it out. You're not going to do an e-commerce store with clothing and say well for this top actually this belt and this scarf goes really well with it you're going to need to do that in an automatic way and that automatic way means the content all needs to have attributes on it so that a piece of mathematics can figure out what to show. You can hear this is really quite different to what I talked about before and what is covered in those books.

Where we used to spend a lot of time thinking about how we would display content what sequence it would be in, what things came first because they were important, what things came last because they were less important, what things go in the middle. Now algorithms are deciding both what we see and in what sequence it's displayed. That changes dramatically how we how we curate how we how we consume content. The things we see - you can think about this you know in terms of anything like Facebook, twitter, LinkedIn but also these kind of regular sites like a recipe site or clothing site. The algorithms are deciding what you're going to see and it may be based on a bunch of other stuff that a site knows about you as well but there's not so much a deliberate hand creation of great stuff - we're getting an algorithmic view of what the computer thinks we should see. It may not be bad, but it might be.

Similarly we're increasingly seeing decisions being made by artificial intelligence - I always think about this one in terms of things like you know things like bank loans - I applied for a loan last year um to buy the household I'm in now and I had to give lots and lots and lots and lots of information and at the time I was self-employed and in Australia banks are, it's really difficult to get a loan when you're self-employed - when I'm employed I just give them my payslip and they're like sure we know how much money you earn. When you're self-employed you provide all of this other content and then you know here's a piece of algorithm piece of ai somewhere going no without considering you know circumstances or understanding where I really am it's just like no. The consequence of this is that where humans you know used to have some discretion around considering real context when they were making decisions and also um really considering who they were excluding from decision making when when a computer does it and our AI or you know regular algorithm does it then there's less consideration about who is being stopped from participating in this in this economy or participating in this service or being allowed to to access something. Without careful work we may not even know that there are a whole bunch of people being excluded from services. So again it's a different world. I don't want to be negative except that my face goes... Anyway i finally got a loan but i had to get a lot of human intervention and i had to get a lot of hand-holding from an individual who did a lot of work for me to get my loan. I was already paying off an equivalent loan there was no doubt that i could pay it - I just fell into this hole of algorithms in the bank that just meant the computer said no. Anyway it was a very stressful time.

So what's common to these kind of projects and what I think information architecture looks like today - you've heard me say a bunch of these already. A lot about a lot of like how to do that work is about understanding concepts, ideas and the relationships between ideas and also the relationships between content. Really understanding in an abstract way what goes with what and being able tomark that on the contents so that we can use it in a in a  templated automated way. All of that work really needs an understanding of the relationship between ideas and content. In all of that there is still a lot of structure, there's a lot of figuring out how things are set up and related and there's a lot of organizing still especially given that a lot of the what i was talking about there are large amounts of content.

The work is all about communicating with people and helping them understand and helping them make good decision. There's so many ways that that can can can play out in projects but that's still a fundamental piece of all those project types. As I said as I was going through that a lot of those are built on rules and algorithms not necessarily built by hand.

So what then are the uncommon themes - the things that are different in that big set of IA kinds? You can probably see that they're really different project types. Working with a team who's doing something you know with AI is really different to working with a team who are working in e-commerce. A e-commerce work is really different to working in government trying to help people make decisions. They're all really different and that means that where the books that i mentioned at first - you can pick those up and the practical nature of them meant you could say okay we're we're building a website I'm going to start at the beginning of this book and I'm going to read it and it'll tell me pretty much how to do things. The style of project that I just talked about  - they're quite different. There's not a lot of consistency so it's really hard to say do things like this.

I said uh before i was trying to teach information architecture last year and I've been teaching information architecture since probably 2002. I'm at the point where I can't teach it anymore because all of those project types are so different. I don't know what I'm teaching anymore. I was really in a mess trying to figure out what I was telling people because I was using my old teaching methods. I know that I wasn't successful like I used to be when I could teach in a  how-to way.

There's then also not a really consistent entry point to this kind of work so  in the the IA yesterday days you could start on a small website and then you could then you could get a new project and work on a bigger website and then you might work on a bigger one or a different kind of style or in a different field and there was a consistency to them. If you tried now I don't know even know how you would start to tackle a project like the ones I mentioned if you were you know close to a beginner. There's no easy way of saying well this is what a like this is what a small project looks like it's like research where you can say okay here's a straightforward study learn about that and then you kind of build it up. That's a problem and in a lot of those - other people on the team aren't necessarily going to know that information architecture skills - those structuring and organizing and labeling and you know marking content with attributes. They're not necessarily going to know that that's what you need to do in the projects so and certainly this is why i think a lot of the large sites we use now especially e-commerce sites are really really really hard to use. Because there's probably nobody in a team who thinks about helping people understand, make decisions, filter through content.

So it might mean that you already have a set of good IA skills and you have to communicate to other people - what I have to do communicate to other people exactly what the skills are and how they affect the content and how they play out 1in a project. I was working on a government project most of 2000 and it was definitely a project that needed information architecture and they knew it so they hired me for it but I still spent a lot of my time trying to not use jargon to explain these ideas around you know categorization structure and metadata on content and principles and rules-based design and giving them lots of examples. I spent more time explaining what the work was than actually doing the work and that's me with you know a ton of experience so this is all hard. If you're working on a project that you feel like is one of these and you're feeling lost that is exactly why it's not you it's a world that doesn't see the structural elements and it doesn't see the underneath things. It only sees the surface. I don't do visual design so people are like what do you do as a designer if you don't make things look good?

The things that I think are the skills um needed to do those kind of projects - absolutely positively research skills are still relevant. We still need to understand what people already understand where they are are they beginners in a in a topic are they experts, what kind of experts are they, what do they know, where are they now? That's one of the really important parts - helping people get from a place where they don't know something to a place where they do know something - understanding where they're starting from. And also you know what they need to understand. I don't know that regular user research really focuses here - regular user research I see a lot of focus on tasks and not so much focus on learning and understanding people so this is still a really important thing for us - researching to understand people's information needs are really important.

The other thing that's still really important is really understanding how people think - really understanding what their mental model is of a content area, of a task they need to do, how to approach something. This is still crucial  so those skills are things that many of us have and are still um amazing and important.

The other skills are these real structural ones. Concept modeling - what concept modelling is understanding a topic, a domain that you're working in and being able to  describe it in abstract and describe the relationships between things; and communicate that to other people. This is still a really important skill and this is the thing I like most about a lot of analysis and information architecture work is figuring out how the world works for a particular topic.

We also will still need to do content modeling so that's different to concept where the concept is- what are the ideas in this topic. What are the pieces that we're producing, sharing communicating with and how do they relate to each other. This is macro and micro so how does the content work out at a really high level. What are the main chunks of it how are we going to design main navigation and the main pieces of of a domain but also that micro level content like the really nitty gritty things on a page like you can see this in a recipe for - how how long does it take to prepare how long does it take to cook - they're really little micro pieces of content

It may not be that a person has that has a skill in that really broad modeling and they're really detailed - that might be two sets of brains because it doesn't always fall naturally in one head.

We need to understand how categories work - this is the part I said is the theoretical part in my information architecture book all about how categories work in our brain how we think about how categories work and how they actually work um how people develop their own folk taxonomies - their own ways of working and thinking about the world. So this is all still important.

Classification - so exactly what what are we going to use to bundle up a set of content, how we're going to classify it, how we're going to do that kind of detail. Again with recipes what are the actual lists of cuisines, what are the actual lists of cooking types. How we're going to classify those.

So these skills are all what we need to do that work but you'll also kind of see these are these are fairly conceptual kind of skills. Where the research one is pretty practical, these are it's going to be weird to say to people I'm really good at this. We know what taxonomy development also fits into working with big content sets.

So that's where I think we are right now. I would say if you asked 10 information architects where we are today we would all have a different lens on it and we'd all have different lenses about where we got and why we got there and that's okay. I'm not trying to present the definitive view - this is just my view of how things have changed and what we might need to do

So I think we're up to questions.

Tina: thank you so much Donna for an interesting talk and again all of you behind the screen you can ask questions in the q&a section. We have some coming in but I would actually like to start the bowl because i was quite affected with the stuff you said because I've been so frustrated with a lot of in general when when I meet sales people because they want to do something else than I want to do as a ux designer right. I was so struck by you say that the longer text they want that and that's not always good because sometimes it's shorter and I'm quite famous for simplicity in general and also in text and so on so everything I write does not perform that good in google and I'm so annoyed by google for that.

So my question is actually do you think that information architecture is dying due to optimization

Donna: I don't say dying because the skills are still needed. I think it reduced in popularity for a bunch of reasons. I'm missed a thing on my IA today slide the other thing that happened is we went from doing kind of content-based websites to applications so we started turning everything into software - both making apps for devices but also literally. When you do a content-based website you can create that with very little technology and a bunch of writers. But we kept we kept throwing software people at our websites and our projects and software people want to build things of course so we end up with all these really application heavy ways of working including where we like drive people through decisions instead of letting them kind of puddle around in content and making decisions themselves. So that's also a reason why information architecture became less popular but also like the the people in it moved on to other things like user experience started getting more popular and we all love the talking to people and understanding users side. A bunch of people went more into strategy so the middle just fell out of it as well. So that's part of why I think it disappeared.

I think the google thing and these nasty weird long terrible content - this is all seo and marketing driven. Everybody wants attention on their site and one of the ways that people were told to do it was to make more content and now the like one of the things I've been uh seeing um in the media in the last couple of weeks is like much more about AI writing content. I like simplicity. I want a short sharp article. I do not want to - and like people laugh at the recipe bloggers - it's a joke about like the long rubbish content before you get the recipe- but it works because people are going from google and starting from google not knowing what websites to start with.

Tina: Yeah exactly exactly we have some questions coming. We have one from shannon and she asked how does the AI decision making relate to IA. I understand the point about it affecting people's ability to participate in eg the economy but not understanding how that relates to IA. Do you mean it affects that content what content they are able to see or find online?

Donna: Yeah I said that and I didn't explain the connection very well. So the way I think it relates to IA is when we have a bunch of algorithms making decisions and decisions on either like my loan application kind of decision or like LinkedIn and showing how the algorithm shows content, and we don't have people involved in who care about users, who understand how to help people understand, who understand the consequences of categorizing people - when we don't have these people involved in projects we end up with um systems that are hostile to humans. That's my bridge - those of us with with skills of understanding how categories are made, how people are categorized, how these decisions get into systems - so if I'm working with an AI team I'm going to keep injecting humanness into the process and hope that the outcome is more humane so that's that's where I was going but I really didn't explain it because I explained it quite quickly.

Tina: Shannon i hope that is helpful. Sheldon is asking regarding goal setting for design projects  - a lot of time goals are tied to business value eg increasing sales by 10 percent unfortunately this is not based solely on design but on many business functions. How can we define better goals for design projects.

Donna: I think personally there aren't such things as design projects. Designers work within often product teams and should be contributing to decision-making in those product teams. So in a good well-functioning team the whole team will be coming up with that percentage not just a product manager. It doesn't mean that many or any teams are actually well functioning - that means that those of us are doing research like structural design like these are all involved in setting the goals and we set them based on the skills of a whole team including the skills of the people who are talking to the users, not just a product owner. If a product or a product owner just says you know increase by 10 then your job is - okay well let's figure out what that means what does 10 percent mean, which 10 percent, which users so then you start really unpacking that. But yeah I never think about things as design projects - design skills work within a product usually

Tina: That's difficult right because also in the organization I worked in like what does that mean for for the product and like what should we do of course it's easy if there's issues with the product. I've worked a lot with product with connectivity issues right so of course we could say okay we need to have better connectivity there or we need to have the other interface that people understand it but going sales that's also that's not just us right so it's a whole team effort.

So we have uh Martin asking how would you describe the difference between IA and navigation design if you see any or maybe more correctly where do they differentiate and where do they overlap.

Donna: Yeah so I did a lot of hand waving. When I'm thinking about IA I'm thinking about the structures, I'm thinking about the content and how the content relationships work and the templates and the things that we might need to display. If I think about IA I'm thinking about all of those structural elements and the relationships and then navigation design is how do we represent this  via a screen  usually - because it could be we could be writing a book - but how do we represent this via a screen in a way that we can help people do the things they need. There are tons of places where we do navigation design we've got our normal kind of things we might have like navigation bars, menus, we've also got filters, we've got related links, we've got links in page, we've got things like tags on content, we've got search results and the ways of you know of navigating by search. There's probably some other things - we've got landing pages and cards and all those all of those are navigation design. They need to be built off an information architectures o you can't design navigation or page layouts or links or templates before you've figured out the structure of the content. You can't pour content into a page. You need to figure out the structures first.

Tina: Yeah i think that makes sense. [someone] is asking how do you say see IA in relation to Web 3 - will things improve according to IA or will it be worse.

Donna: Web 3 is a big thing. It's not just one thing. I'm working in web 3 at the moment so at least I know how to answer this. I'm working in the  blockchain and nft part of web3. I'm not working in metaverse. I'm not working in decentralized finance part of web3. You think about all the project types I had - web 3 has tons of different kinds of projects so when I'm working on a NFT we're still going to come up with navigation we're still going to come up with metadata we're still going to help people to find content based on attributes - it's exactly the same for that for that kind of element. I don't have much in metaverse at all so I don't know what the IA bridges are there yet at all um. New things like web3 sometimes need exactly the same as the old things particularly um they particularly need you know exactly the the skills we already have with user research they still need all of the skills that we still have with conceptual design, with strategic design and with interface design. Some of the projects will need more information architecture than others.. I don't think it will necessarily improve. I don't think there's anything inherent to web3 to either the decentralized part of the inspectable blockchain part of it to the easy financial transaction part of it - I don't think any of that is necessarily going to be better because of IA.

Tina: Shannon is asking what is the difference between categories and classification.

Donna: Yeah sorry i said that one quickly as well and even when I said it I thought I need to explain this better. So categories are theoretical. A classification is a way that we implement categories in a practical sense so we might have a classification of cats and the classification of cats may have fluffy cats, baby cats, cats with short, legs cats with no tails - that might be our cat classification. Similarly recipes we might have our classification around cuisines, types of cooking. So they're the practical side of the categories that are like the concepts in our heads. Sorry about those I know that also like these are things that may not translate super well.

It is a bit fluffy right because this is all this is it's all fluffy because we all think in a different way and this is like uh one of the really interesting things for me working with people from around the world is I get to understand how you all think in different ways um according to your  cultural background, where you grew up how your family thinks. And that's why it's fluffy because the way our brains do things and the categories we create are built on our experiences. So yeah it's really floppy.

Tina: think we might have time for the last question here. Sheldon is asking how do you approach designing interfaces for comparison for mobile screens.

Donna: so first you think is this the best place for this task? Do people really want to do a complex comparison on a mobile screen and if the answer is yes then you need to dig a level deeper you need to go okay people want to do a comparison on a small screen whether it's because they're commuting whether it's the only device they have access to - okay yes we we definitely want to do a complex comparison on a small screen what do we need to know? What are we actually comparing so this is a very much an IA question. Which elements are the things that we need to to compare between so you might, once you've answered these questions, you might be down to well actually I just need to know location and price- I don't need every other attribute to be on this screen so then the smallness of the screen becomes less of a problem because you've solved the more detailed problem of what are we actually trying to do here and you've also asked the is this actually the best place for this task. If it's a complex comparison - you know in a kind of work setting where people have access to a large screen then they might actually prefer to do that where they've got space. Really understanding like you might just need to show two things. So it's not about the layout and it's not about can I fit skinny things into columns - it's what do I need to do here.

Tina: So time is up but thank you so much for very interesting. Donna I definitely believe that like I'm quite wiser going from here and it's also just nice to hear that you're struggling with some of the same issues that we are as well. So I'll just wrap up so - we are recording everything as I said in the beginning so all of you of course will receive a recording of the talk afterwards and also Donna's slides.