![[info]](http://l-stat.livejournal.com/img/userinfo.gif?v=92.2){kind=small}
I always have a hard time describing what the field of "technical communication" really means... so if you read through all this, you should get a pretty good idea.
download the PDF version here. (now with working links!)
OUTLINE
I. Escaping Flatland
1. Show comparisons
2. Show causality
3. Show multivariate data
4. Integrate all elements
5. Document everything and tell people about it (detail = cred)
6. Stand or fall on the quality of your substance
7. Show supporting information adjacent in space, not temporally (in time)
8. Use small multiples
9. The root of all evil: the cherry-picking of data
II. Presentations & PowerPoint
1. ALWAYS give your audience a hand out BEFORE your presentation
2. Start off your presentation with a summary of no more than 200 words
3. Practice, practice, practice
4. Always show up early to your presentation
5. NEVER APOLOGIZE
6. When you give your talk, TALK
7. Use humor and hyperbole sparingly
8. Refrain from using masculine pronouns
9. Make sure that you take questions at the end, but NEVER condescend
10. Be enthusiastic about whatever it is you're presenting
11. And last but not least... always finish early
III. More on Tufte's gripes with PowerPoint
1. PowerPoint's got attitude
2. PP does funky things with language
3. Look at where it came from
4. A better metaphor
5. PGP... one of the many things to learn from teachers
IV. Tips for Improving your Visuals
1. Shoot for a 45 degree angle
2. What you leave out is just as important as what you put in
3. The universal nature of design principles
4. Down with "pitch" culture
5. Paper usage ain't going away
6. Projects cannot be totally managed on a computer screen
7. "Clutter" is just a word which means "failure of design"
8. If you want to clarify something, never take data away, instead, add detail
9. Don't include zero if you don't need it
10. Tell us something we already know
11. Sparklines hold a lot of data
12. Bad UI often shows the underlying hierarchy of the team responsible for the product
13. With innovation, usually the first time is the charm
14. Users can handle a lot of information, if it's presented well spatially (rather than nested)
15. The two worst words in the web site design vernacular are "Skip Intro"
16. The web is also a tough place to design for
17. Be more self-conscious of the media around you
18. The ideal display is 90% content
I. Escaping Flatland
The main goal with information design is to escape what Tufte refers to as "flatland." This term is used to describe the 2D surfaces upon which we are forced to display data from a 3D world. Since we're missing a dimension, we have to make the best of the two we have and break out from 2D displays like paper and computer screens by using physical models (that is, until Princess Leia hologram technology becomes widespread in usage).
The following points are guidelines from Tufte to ensure that your escape from flatland is successful...
1. Show comparisons - When depicting any kind of data, you have to give the viewer something to compare the data to. Let's say I wanna show off my new camera. If one of my goals is to brag about how small it is, I'll need to take a photo of it next to something everyone knows the size of, like a hand or a pack of chewing gum. This gives the viewer a much more informed frame of reference.
2. Show causality - If you're trying to make a point with a presentation or a graphic, you have to show what caused something to happen. If you plot out the number of shoes your factory makes in a month, and there's a big dip right in the middle, there better be something in your presentation to explain why the data looks the way it does.
3. Show multivariate data - This is where the now ubiquitous Napoleon's March map comes into play...
The idea here being that the more variables you can get into your information graphic, the better (given that the variables are relevant). While the Napoleon March example is certainly a masterpiece of design, I don't think this principle is entirely practical for the majority of information graphics people create day to day. I'd posit that as soon as you start to increase the number of variables, the amount of time it takes the viewer to understand the graphic increases. That may be ok for something like the Napoleon march where people are more forgiving time-wise to figure out what all is going on, but in a quick status meeting, people just need to see the data that's relevant with a format that is quick to decipher.
Tufte definitely has a preference over cleverly designed info graphics to commonplace (but still effective) ones. Of course, those are the ones that are more interesting to look at, but you won't get much specific advise out of him regarding how to make your mundane bug chart graphic more snazzier. (I know because my manager and I asked him... his response: "increase resolution ten-fold and use sparklines." (I'll talk about both of those things later...)
4. Integrate all elements - If you're going to give a presentation and there's a picture of a dog on the screen, you better well be talking about something dog-related. Same goes for a report or anything that mixes text and visuals. The two should work in concert with one another. If your visual is distracting the audience from the message you want to convey, ditch it entirely or pick something else.
of course there are always exceptions :)
5. Document everything and tell people about it (detail = cred) - This is one of Tufte's favorite points to talk about. When presenting any kind of information, be as thorough and forthcoming as possible with where and how you collected your data. Sure, your dataset can be incomplete, but as long as you lay it all out there, your butt will be covered. Not only that, but spending more time to talk about your methodology and sources gives your message much more credibility, which is especially key if you're trying to pitch something or present something controversial.
6. Stand or fall on the quality of your substance - No matter how much you dress up your presentation or document, ultimately, its success will be determined by the "meat" contained within. Tufte emphasizes over and over that if you're faced with a choice between prettying up a document versus beefing up the content, always choose the latter. This is one of the primary points he uses against the popularity of PowerPoint. An inordinate amount of time is spent in PP on making things look pretty, that often the content is thrown together as an afterthought.
7. Show supporting information adjacent in space, not temporally (in time) - This is one of Tufte's main points for backing up the usage of sparklines. Here's what one looks like:
To us mere design mortals who aren't using sparklines just yet (give it a try here), I think the main takeaway is that you want to put that pie chart showing the demographic data in your report right next to the paragraph where you talk about demographic data. If your audience has to flip back to an appendix or even onto a second page, you've lost an opportunity to forge a tight relationship between your visual and textual data.
While I think sparklines are a great tool for communicating large amounts of data in an efficient way, I think it will be awhile before they go totally mainstream. Some newspapers, magazines, and other Tufte-files swear by them, but there's still a ways to go before we see little Jimmy using them in his 7th grade science report. Perhaps when sparkline creation technology becomes more integrated into standard office-suite products, then the idea can really take off.
8. Use small multiples - In your information graphics, Tufte emphasizes that the more detail, the better. If at all possible, show your data in both space and time. Here's an example using sunsets...
The power of this graphic comes from all the data used to generate it. With your own graphics, avoid the urge to throw stuff out, but instead, find creative ways to put more in.
You also want to try and minimize the "figuring out" time of your design. While this example certainly looks pretty, it takes a bit of time to understand the X and Y axis to figure out what's being depicted. Aesthetic decisions aside, the designer could've included labels for the individual axis, thus minimizing the viewer's effort to understand the message (aka "shortening articulatory distance").
9. The root of all evil: the cherry-picking of data - Back to a point from before, the usage of small multiples automatically raises the "inherent credibility" of your message, while also deflecting one of Tufte's most hated things in the entire world... the "cherry-picking of data". Tufte claims that cherry-picking is the single largest threat to getting at the truth of a matter. If you've got some sort of weird anomaly or outliers in your data, do not throw those babies out. Rather, include them, and then give some reasons as to why they may be there. If you discard them entirely, you risk losing your reputation and any chance of getting outside expertise which could help explain or correct the anomaly.
Some signs to look for to detect if cherry-picking is going on:
- Does the data look too good to be true?
- Is the data error or problem free?
II. Presentations & PowerPoint
What follows are some of Tufte's tips for beefing up your presentation skills, as well as some of his qualms against PowerPoint.
There is much that I disagree with Tufte on regarding the latter (a lot of which I didn't know I disagreed with until I read Jean-Luc Doumont's excellent paper, "The Cognitive Style of PowerPoint: Slides are Not All Evil"). I'll throw in my two cents as I go through each of Tufte's tips (and beefs with PP).
1. ALWAYS give your audience a hand out BEFORE your presentation - This is perhaps one of Tufte's most controversial claims, but it's one that I agree with him on. Like the guy in Memento, we're all cursed with having horrible short-term memories. I'll go into a meeting, 80 things will be discussed, I'll be lucky to remember two or three. If you have a presentation to make, you better have some key points you want your audience to leave with. Why not put them down on a sheet of paper that people can follow along with and take notes on while you're talking, as well as an artifact they can have to take with them.
One of the main advantages Tufte espouses of paper handouts over slides is that slides, by their very nature, force the audience into the speaker's "cognitive style" (aka "how they think about stuff"). With a piece of paper, the presenter will have organized the information in a certain way, but the audience is still free to glance around on it as they choose. Because of the strict linearity enforced by PP, the presenter becomes the information dispersal overlord who dictates to the audience when and how to access the info.
Where Tufte misses the mark on this is to pin the blame on the parceling out of information strictly onto the shoulders of PowerPoint. Teachers have been using sheets of paper to slowly reveal the info on their transparencies for years. Even scrolls do that by their very design.
A common argument against the pre-talk handed-out handout goes something like this: "Well, if I give people a handout with all my points on it, they won't listen to what I have to say." That may very well be the case, and if it is, you probably shouldn't be giving a presentation on whatever it is you were going to talk about. The handout shouldn't be a substitute for you and your exposition. It should serve as an outline... a memory jogger... a notepad... nothing more. YOU are still the key component in conveying your message by using illustrative examples and providing the right context around your key points.
If you are going to use slides, make sure that your audience can walk away with your message in the absence of slides or a handout. The reverse is also true. If you're going to go to the trouble of providing a visual element to your talk, the supporting materials should stand on their own enough so that the main message you want people to leave with is there, even if you're not.
2. Start off your presentation with a summary of no more than 200 words - You should start off any presentation with a statement of the problem or issue at hand. Next explain why anyone should care (why is this thing relevant?). Lastly, present a brief version of your solution. If you do those three things, you're guaranteed to have a much more interested audience.
3. Practice, practice, practice - It doesn't really need to be said any more than it already has, but Tufte is also a huge advocate of rehearsing presentations. He recommends video-taping yourself and shutting off the picture so that you only concentrate on what (and how) you're saying things. You'll hear all sorts of things you never knew you did before. Crutch words such as "um," "like," and "basically" will give you that nails on chalkboard effect that your audience would've had to suffer through. It's unavoidable to be nervous and you'll screw up more with the tape rolling, but I bet you'll be a lot more relaxed if you do this when it comes time to actually get up in front of a critical audience.
4. Always show up early to your presentation - Those moments where you're setting up your stuff and greeting people as they come in are golden. New connections might be made, respect for you as a presenter goes up, and people are more likely to listen to what you have to say as opposed to walking in late and fumbling around with the projector.
Now here comes my #1 favorite tip of Tufte's for giving presentations.
5. NEVER APOLOGIZE - I cannot count the number of presentations I've seen where the very first sentence out of the presenter's mouth was "I'm sorry...". While it's well and good to be a humble person in life, the intro to your presentation need not include an apology. Even if you're late, the projector broke, the handouts all turned pink, whatever... it's fine to acknowledge those things briefly if you have to, but get off of that stuff as fast as possible and onto those three key things in your summary (problem, relevance, solution).
6. When you give your talk, TALK - Don't be one of those automatons reading notes off of a slide or a piece of paper. One would hope everyone in your audience can read, so when you do your thing, relax and just talk about your subject, using your notes as only a road map to make sure you cover everything you wanted to cover. (even that can be avoided if you practice enough)
Another big gripe against PowerPoint is on this very subject, a phenomenon called "synch and launch" that David Farkas describes in his Tufte-PP rebuttal, "Understanding and Using PowerPoint."
"Synch and launch" refers to the presenter focusing the audience's attention onto a particular bullet point, explaining the point at greater length than what's on the slide (aka "glossing"), and then moving on to the next bullet point.
You want to make sure that your points (bulleted or not) are presented in a way that doesn't call too much attention to this phenomenon. If you're too casual and skip some of the bullets on a slide for example, that will annoy the audience. If you go through each one with military precision, it makes the presentation seem rigid and annoyingly predictable. You want to make sure whatever's on your slides or handouts mesh well with what you're saying, but make sure to avoid either extreme.
7. Use humor and hyperbole sparingly - These things are ok to use in presentations, especially when employed to reinforce a point (just make sure not to over-do it). I think simple common sense works best on this one.
8. Refrain from using masculine pronouns - Even if your audience is 100% male... it's 2005 and I agree with Tufte, any grammar Nazi's that wanna fight over the usage of genderless plural pronouns like "they" need to seriously get a life. While it may not be grammatically correct, it's much better than strange variations on "he/she," "s(he)," or the overly dramatic use of just "she."
9. Make sure that you take questions at the end, but NEVER condescend - Any audience member asking you questions is most likely doing so because of a personal concern or issue they themselves are facing so the last thing you want to do is make them feel like they're stupid. If the question is too difficult to tackle or highly individualized, see if the person can talk to you after the meeting, just please don't say "let's take this offline." It's just as easy (and less jargony) to say something like "let's discuss after Q&A."
10. Be enthusiastic about whatever it is you're presenting - because of you're not, you better believe your audience isn't going to be either. If it's something dry like how many pencils your pencil factory shipped this month and everyone already knows what you're going to say about the subject, perhaps a presentation isn't what's needed (or maybe it's time to look for employment elsewhere).
Instead of a boring presentation, a well written status email may serve just as well. These are the sections I've found to be helpful for such mails...
- Highlights - what's going well on the project
- Main issues - what are the current challenges (never say lowlights or anything else that will demoralize your team)
- Stats that show how on or off track the project is going (use sparklines for bonus points)
- Relevant dates, milestones
- Short, one line status for individual team members
III. More on Tufte's gripes with PowerPoint
By now, we've all heard a ton about the evils of using PowerPoint... whether it be from the guy in the cubicle next to you, from Tufte himself, or from David Byrne of the Talking Heads...
Here are some more of Tufte's main gripes (and my thoughts on them)...
1. PowerPoint's got attitude - The origin of Tufte's PowerPoint hatred comes from what he describes as its "attitude." All interfaces project an attitude and the UI characteristics dictate what that attitude will be. In PowerPoint, Tufte lists these main characteristics as:
- foreshortening
- low-resolution display
- breaking up of information into small pieces
- fluff (reliance on clichés)
- preoccupation with layout
- commercialism
All of these, he argues, are hostile to scientific, rational, sound thinking. This means that even if what you're saying is good, if it's coming through PowerPoint, it automatically carries over the program's attitude along with your message.
I agree mostly with this analysis, but I wouldn't say any of this "attitude" factor is so great that people should stop using the program entirely.
For me personally, I'd rather use paper handouts in conjunction with PowerPoint, acting as a simple slide projector for images (which is what Tufte advocates)... not because I think the program is evil, but because I think most people are sick and tired of looking at bullet point-laden PowerPoint slides.
You automatically set yourself apart from almost every other presenter out there if you come at your presentation from an angle which pushes PP to the backseat rather than the front. If the goal is for your audience to listen and ultimately understand what you're trying to say, this tactic gets you more attention from the audience, before you've even said a word.
2. PP does funky things with language - "Accelerate the introduction of new products" is a perfectly good newspeak-esque phrase in a typical PowerPoint slide, but elsewhere, it's a sentence without a subject.
From my experience in large corporate cultures, that way of speaking seems to trickle down into real-life conversations where people are rewarded for extreme brevity. Yeah, we understand that you get 200 emails a day and you like your communications to be as easily digestible as possible. But when your primary way of communicating is done in this style, not only does it truncate language, it also serves to truncate thought. (note to self, go read 1984 again).
While you want to keep your points presented on slides or handouts brief, make sure they convey complete thoughts that don't sound too hokey spoken out loud. Most (non-robot) people want to hear a good narrative, so give it to them.
3. Look at where it came from - Tufte is also wise to look at the origins from which PowerPoint emerged. The primary two factions that had a hand in its birth are programmers and marketers (cue ominous music).
If you look at what each of these two groups are good at, you'll find a strong correlation with the aforementioned PowerPoint attitude. Tufte states that programmers are good at sequential hierarchy and marketers are good at advocacy, exaggerated claims, and speed. Marry these two lovebirds and it's easy to see how PowerPoint came to be.
So what Tufte wants to know is why on Earth would anyone want to use a tool riddled with such a history as the de facto standard for presenting substantive data (and this is where he's gone off the rails).
Again, referring to the awesome Doumont rebuttal article, Tufte's huge mistake in his 28 page condemnation of PowerPoint is that these slide decks exist in a vacuum. He presents the worst slides in a giant deck from Boeing regarding the Challenger disaster as the examples he tears apart. He doesn't give enough credit to what the presenters said verbally in addition to the slides, and he certainly doesn't talk about how these decks are supporting material that go along (not to supplant) an actual thorough report.
Yes, of course, if any substantial, scientific report is to be given only in PowerPoint (or any slideware for that matter), it will have some huge setbacks for sure. But as a means by which to give a presentation, as long as the real report is out there (using Tufte's escaping flatland principals), then I see no reason not to use well designed, thoughtful slides.
4. A better metaphor - So if Tufte doesn't think programmers and marketers are the leads we should be following, then who should we look to?
TEACHERSThis was perhaps the most inspirational part in Tufte's whole class for me. I agree with him that 99% of books out there on being a good presenter are garbage. You don't need catchy phrases and one-liners or self-affirming BS to raise your confidence.
If you think back to some of the best teachers you've had in your life, think about what made them so great. Was it how jazzy their PowerPoint backgrounds were? How much jargon and buzzwords they could throw into a lesson? Of course not. It was most likely due to their passion and mastery of the material they were trying to communicate.
5. PGP... one of the many things to learn from teachers - Tufte suggests a common teaching method referred to as PGP (particular general particular) when trying to teach something. What this means is that you want to pick a specific example of a concept to illustrate your point, then talk about the general principals behind why or whatever that particular thing may be, then finish off with another particular to reinforce the general principal.
Examples are great aids in learning about something. If you went with just a GP method, I bet it would take much longer for your audience to get their heads around the general theory in the absence of something concrete to latch on to.
related links:
- next time you need to make a snazzy PowerPoint presentation, head on over to eBibleTeacher.com for inspiration
- Ian Parker's New Yorker piece, "Absolute PowerPoint" (pdf)
- Taking Saudi Out of Arabia
- Norvig's "The Gettysburg PowerPoint Presentation"
- Daniel Radosh:
IV. Tips for Improving your Visuals
Here are some other kitchen sinky tidbits from Tufte's class that I found helpful...
1. Shoot for a 45 degree angle - If you're going to graph something, aim for having your line ramp up (or down) at a 45 degree angle. You can adjust the units in your X and Y axis to make this happen. This may seem a bit arbitrary, but the underlying principal is that you want to choose an aspect ratio so that your graph looks lumpy. If your graphic is stretched out too wide or too tall, it will be harder to make comparisons and to see where anomalies occur in the data.
2. What you leave out is just as important as what you put in - When looking at Napoleon's March, you'll notice that Napoleon's name is no where to be found in the graphic or its accompanying text. The same goes for Newton's name on the Principia's title page. The point being, what you leave out of informational material is just as important as what you put in it. The creator of the March graphic wanted the focus to be on the army, not on Napoleon himself, so this design decision reinforces that goal.
3. The universal nature of design principles - Designs principals are universal and independent of language since they are tied to basic human abilities and limitations. If you keep that in mind, that should lead to better designs.
4. Down with "pitch" culture - Another huge evil in Tufte's universe is the act of "pitching" (y'know, what advertisers and marketing people do). Pitching to Tufte embodies all that is wrong with information design and is the source for much bad decision making. It also doesn't help that the main "pitchers" of the world use PowerPoint as their main tool for communicating. The mantra here is "pitching out corrupts within." This basically boils down to lying to your audience whomever they may be... and no one likes that.
5. Paper usage ain't going away - Tufte says here that:"In the last 10 years, every country in the western world has increased its per capita use of paper. For some cognitive tasks, paper outperforms computers, a case made in the book by Abigail Sellen and Richard Harper, The Myth of the Paperless Office."
One of the big misconceptions since computers really took off in the late 90s is that computers have decreased the amount of paper we use per capita each year. Nothing could be further from the truth as it continues to go up due to the convenience of printing technology. I dunno about where you work, but the companies I've been to have their printers overflowing with specs, mapquests, to-do lists, contracts, receipts, etc. While I personally have drastically reduced the amount of paper I receive by handling all my finances online, I still manage to print a lot of stuff out, exactly for the reasons Tufte lists for why paper is still needed.
6. Projects cannot be totally managed on a computer screen - They are too complex and low resolution to house all the data associated with them. Instead, Tufte recommends the use of wall charts, which I hear companies like Boeing make ample use of. (if anyone has some good examples of such things, I'd love to hear about them).
7. "Clutter" is just a word which means "failure of design" - Something that is cluttered doesn't contain a relationship between the amount of information displayed (detail) and the ease by which the data can be read. The problem isn't there being too much data... it's just that the amount at the resolution/layout shown is at the wrong scale. This brings us to perhaps Tufte's most famous suggestion for improving any information display... "increase your resolution by ten-fold." What he's really saying here is that if you've got a ton of data, don't try and cram all of it into a tiny space, grab a bigger piece of paper, use a poster board, bucketize the information, do something to get it a scale that is human (not computer) friendly.
8. If you want to clarify something, never take data away, instead, add detail - By detail, more often than not, this equates to providing increased context around what it is you're trying to say. For instance, if I'm trying to teach someone a new word, the best thing I can do is put it into a sentence which illustrates how the word is normally used. That addition of detail leads to greater ease of understanding.
9. Don't include zero if you don't need it - When it comes to illustrating financial data, it is not necessary to include the zero point in your graphic. Many times the inclusion of zero totally screws up the scale of the other data present in the graph, so it's best to just leave it off and keep everything at scale (upwards or downwards), making sure that it's obvious that zero is not included.
10. Tell us something we already know - If you're presenting something and want to add some umph to your point, make your data confirm something that was previously known. If your new demographic pie chart confirms what the U.S. census data already says, then whatever other points you make off that data will be underscored.
11. Sparklines hold a lot of data - The average sparkline includes 14,000 points of data. You could easily have a well-designed document showing stats across an entire season of NFL football that contained one million points of data, all on one 8 by 11 piece of paper.
12. Bad UI often shows the underlying hierarchy of the team responsible for the product -If you look at a product with a poor interface, 9 times out of 10 I'm willing to bet the problem is that the different components of the design are disjointed from on another. These disjunctions often indicate the underlying hierarchy that is contained within the organization that made the wretched thing. In Tufte's words, "The major problem of bad user experiences is that they segregate information by the mode of production." If I can tell which part of a product was designed by a developer, which by the localization expert, and which by marketing, then it's safe to say that the design is a piece of crap. (insert standard Don Norman diatribe about why the iPod is so great)
13. With innovation, usually the first time is the charm - The example Tufte uses here is that of the original Graphical User Interface designed at Xerox PARC (not at Microsoft or Apple). Tufte claims that many times, the first version of something is the truest to the ideal for how the thing should ultimately be. Case in point, the Xerox designed GUI had no concept of "file types" or applications. Everything in the Xerox OS was a document... just like a piece of paper. You could open a document and draw in it, make notes on it, put a spreadsheet in it, use it to write code, and so on. You just opened a new document and off you went. Along with documents, the GUI had a printer icon and a trash can. That's it.
There was no overt knowledge of what OS you were using, what version you were on, which applications you had running. Tufte claims that all of these things evolved over time as the evil marketing forces got their hands on the technology and began to see the OS and apps as marketing opportunities that could be used to exert control (and dollars) from users. A favorite joke amongst many HCI folk sprang up at this point in his talk... the one about how only two industries use the term "user"... pharmaceutical companies and the software industry.
Compare the Xerox GUI to today's modern operating system where the average computer has over 100 applications and file types that users have to get their heads around to use. Both Tufte and I agree that this is one of the major failures of the modern computing world.
Why should my mom care that she needs to use Illustrator and not Photoshop to make a pretty flower to include inside a greeting card for her friend, which she'll then have to export out as a 256 color BMP, that can then be imported into some sort of home publishing software, that then uses some other kind of proprietary file format. I believe that this lack of of flat interoperability is holding progress back and is the single biggest example for how far we have to go until we hit what I call the "user experience promiseland."
To help illustrate, click on the graphic below which will take you to a version of it with overlying notes that explain what all this means. Hover your mouse over the different parts and hopefully what I'm trying to say will get across :)
14. Users can handle a lot of information, if it's presented well spatially (rather than nested) - Tufte predicts that the UIs of the future will move towards flat interfaces (not nested hierarchies) that promote spatial looking versus stacking.
The illustrative example he uses in this case is that of news websites. Most of these have over 300 links on each page that are laid out in a spatially hierarchical manner, rather than a nested one (underneath other links). Users of these sites aren't going to them looking for pretty designs, they're going there for the content.
Another luminary in the field of UI and user experience is Steve Krug, the author of "Don't Make me Think" which espouses a stripped down approach to usability. At a recent talk, Krug stated that the best sites and applications are the ones that do one or two things "really well" and the example he gives is the Google homepage. Don Norman, on the other hand, argues that the Google homepage is a giant missed opportunity which I think Tufte may side with him on. While I don't think anyone wants to see the functionality of the Google homepage turned into a massively convoluted portal a la MSN.com, there are many points of interaction between the Google homepage and their other properties like Maps, Gmail, Blogsearch, etc that could be beneficially surfaced here.
15. The two worst words in the web site design vernacular are "Skip Intro" - If you've ever been to a website with Flash animation intro then you know what he's talking about. Tufte calls this "the long tunnel of entry," another evil concoction coming from a pitch-cultured organization which thinks it can milk a few more dollars out of consumers by forcing them to sit through a pitch before interacting with the main website. What "Skip Intro" is essentially asking users to do is become interface designers themselves. Instead of allowing the site's design to facilitate a more natural navigation, the Flash intro creates an unnecessary barrier between the user and the content. A well designed website, according to Tufte, should be self-effacing, letting the content take center stage instead of the window dressing.
16. The web is also a tough place to design for - If you look at most of the great artists, they tend to think big, they're very expressive and lavish with their creativity. The web makes that difficult because of its relatively small display and rigid HTML-rooted structure. While all forms of expression have some set of boundaries, the web in particular is fraught with them. This requires a special breed of designer, one that can handle strict constraints, but also have the creativity to do something interesting within the established parameters.
17. Be more self-conscious of the media around you - When making presentation decisions, consider the pros and cons of the mediums you have at hand. If you're looking at the web, it's strong suits are its ability to display dynamic/changing visuals that are interactive. Paper however has the advantage of being highly portable, less buggy, and also a much more permanent artifact (compared to a file sitting on a hard drive).
18. The ideal display is 90% content - Lastly, one of Tufte's more controversial statements is that the average information display is only 18% content, the rest is just empty space or UI devoted to non-content things like ads, OS elements, etc. In Tufte's ideal UI-land, 90% of all screens should be content. Now if you stop and think about that, even with a bit of blank space in between all that content, that is A LOT of content... sort of like this post. I wonder if I've made Tufte proud? :)
wow, so if you've made it all the way to the end of this, I applaud you.
I'd love to get other people's questions, opinions, and feedback on anything I've written about here.
October 25 2005, 14:29:54 UTC 6 years ago
Just some thoughts
Man how long did this take you, I'm always amazed by your gargantuan posts going from topic to topic (like musicians and games including tons of links). I always wonder if you're taking notes during the day so you can get back later and put together a post like that.
October 25 2005, 17:40:14 UTC 6 years ago
There is definitely a balance that has to be struck between overwhelming people, but still providing a wealth of information (something that I'm constantly trying to figure out).
One of Tufte's arguments behind the 90% statement is that humans, through evolution, are highly adapted to deal with lots of complexity. If you take a look at the brain, it's amazing what kind of filters we have going on so, for instance, if you step into a forest, you don't immediately shut down from examining the complexity in the leaves, trees, etc. (i think I've actually heard of people with a disorder where this isn't the case).
He takes this mental ability and applies it to UIs like in big news websites that are filled with content. If we can handle not freaking out at the sight of a tree, he claims the same holds true for a well designed page that employs spatial rather than stacked hierarchy.
This obviously has to have a limit though. As evidenced by this post, sure there's a lot of info here, but I doubt anyone aside from my friends that are into design would read the whole thing because of this intimidating, all-at-once format.
I could've parceled this out into several smaller posts, which would've attracted more eyeballs, but in this case, time was unfortunately a limiting factor.
It took me two nights to write this, but it's something I've been developing in notes and in my head for awhile. I use this great hierarchical notepad program for my notes called Keynote. I believe there is Mac software out there that is comparable (and probably even better than the PC one).
thanks for the feedback!
October 25 2005, 17:51:31 UTC 6 years ago
The thing with that Wall Street Journal page is I found it completely unusable and confusing. There was just too much on it that I didn't know what to do. I meant I knew that I should probably start at the top and read to the bottom, but in terms of being able to use it to find things I was interested in it was difficult. If I spent enough time with it I'm sure it'd make sense but that sort of thing I just find very offputting. I know I'm on the minimal extreme but I think someone is more likely to wade through that page because they really want to. I think there's a way to represent that same amount of information in the same basic oganizational structure but to make it a little more readable. So I guess that solution would be somewhere inbetween.
October 25 2005, 15:42:41 UTC 6 years ago
Tufte lost a little cred for me when he started on the "skip intro" bit -- obviously they're bad, but people quit doing those a long time ago. I don't think he's much of an effort when it comes to screen/web design, and should probably stay out of it until he's up to speed.
October 25 2005, 16:04:52 UTC 6 years ago
October 25 2005, 16:11:43 UTC 6 years ago
October 25 2005, 16:32:30 UTC 6 years ago
October 25 2005, 17:42:13 UTC 6 years ago
I think the "skip intro" lesson extends beyond to just that one example... it's message is pretty much, "keep as much crap out of the users way as possible."
October 25 2005, 17:56:50 UTC 6 years ago
October 25 2005, 18:01:41 UTC 6 years ago
I think one thing to remember about a Tufte class... while learning is a big part of it, we also all want to feel like we're in some sort of secret good designers club. By pointing out something like that, it's an obvious way to get the crowd laughing and nodding in agreement... but yeah, anyone who didn't know that those suck before Tufte needs a lot more help.
October 25 2005, 16:01:52 UTC 6 years ago
October 25 2005, 17:43:01 UTC 6 years ago
October 25 2005, 16:31:40 UTC 6 years ago
Excellent
Re: IV.4I completely agree that we need to step away from pitching. I work with primarily B2B clients and my biggest push this past year has been to stop thinking of your website as another sales person. Instead think of it as a resource center for potential customers. If you become recognized as the leader in your space, however obscure or minuscule, and portray technical credibility, then customers will naturally turn to you for your solution. I urge my clients to write articles, make charts, even write blogs about what they do. Over time, people will naturally start linking to them and sharing the website with colleagues (as I have done with your PDF) because of the resourceful nature, and not because of any sales pitching. In this example, your paper is not selling us anything in particular, but what you have sold is the fact that you are insightful and apt at providing a solution for problems in your specific space.
That being said, I think you did a great job, yet I still need to digest further. You covered a lot in a relatively small space. The text invites others to explore deeper, which is what I plan to do.
-rolo
October 25 2005, 17:45:24 UTC 6 years ago
Re: Excellent
stop thinking of your website as another sales person. Instead think of it as a resource center for potential customers.amen. I just wish more organization would take this message to heart!
and Thanks for the props, hope it comes in handy.
October 25 2005, 18:57:08 UTC 6 years ago
One thing I would add is to make sure you know your stuff inside and out before presenting it to an audience. What I do now is try and come up with all possible questions that can be asked about the topic and make sure I can answer them w/out doing the "taking it offline" BS.
November 5 2005, 21:13:43 UTC 6 years ago
Great stuff! So much here to respond to.
Tufte's B.S. about PowerPoint drives me up the wall. "PowerPoint was created by engineers and marketers, therefore, BAD!" Uh, yeah, so was most other software. *rolls eyes*
I don't agree with the "don't include zero" bit. Any graph that doesn't should feature a small inset showing the same data with a zero-grounded axis, so you're reminded how small the effect being shown actually is compared to the absolute magnitude of the data.
The "present data in a flat spatial arrangement" comment is interesting. I feel that it's unfair for me to comment without hearing the argument in greater detail, but I'll say this: my immediate question is: What's the navigation model?! Having recent seen another pan & zoom document prototype made me wonder if there are any decent implementations of the idea. (In my mind, the ones that make sense are the ones where the spatial arrangement reflects an underlying strcture/hierarchy.)