AGENCE FRANCE PRESSE v. GOOGLE INC.

Filing 60

RULE 26b4 STATEMENT. (Attachments: # 1 Supplement Part 2# 2 Supplement Part 3)(Elgin, Jennifer)

AGENCE FRANCE PRESSE v. GOOGLE INC.

Doc. 60

Case 1:05-cv-00546-GK
b r ä s taks [ c o l l o q . ] n. details of i m m e d i a t e practical importance

Document 60

Filed 01/29/2007

Page 1 of 7

REPORT FOR WILEY REIN & FIELDING

Brass Tacks Design
1014 Colonial Avenue Norfolk, Virginia 23507 www.brasstacksdesign.com

A Critical Analysis of AFP vs. Google
Prepared by Alan Jacobson, President, Brass Tacks Design, LLC Janaury 2007 OVERVIEW I have reviewed the 143-report produced by Mr. Philip Nesbitt for AFP. Mr. Nesbitt claims that "This report will show that the three major items in contention ­ headlines, blurbs, and images ­ as an informational package is more than enough to satisfy most users." Mr. Nesbitt offers no quantitative data, qualitative data or surveys of online news users to establish this claim. Instead, Mr. Nesbitt offers up opinion as fact, many contradictions, and research that, by his own admission, is in doubt: "Whether by survey or eyetrack, there are as many points of view as there are studies. There is a lot of contradictory information as a result of this research, many different interpretations of the data, and some outright dismissals..." That admission is remarkable in that it casts doubt on all of Mr. Nesbitt's research-based conclusions. Futhermore, Mr. Nesbitt admits that some studies contradict the research he cites, but he makes no attempt refute these contradictory reports, nor does he provide any basis for including some studies and omitting others. Mr. Nesbitt's core claim, that AFP's "headlines, blurbs, and images -- as an informational package is more than enough to satisfy most users," does not apply because Google News never takes such a package. At Google News, all headlines do not have blurbs. And headlines, blurbs, and thumbnails do not necessarily come from AFP any other single news source. So the informational package as described by Mr. Nesbitt is rarely, if ever, from AFP. Mr. Nesbitt offers contested research on the physiology of human perception of print and online news products in an attempt to prove his contention that Google News satisfies most users. But none of his "evidence" addresses user satisfaction.

Dockets.Justia.com

Case 1:05-cv-00546-GK
b r ä s taks [ c o l l o q . ] n. details of i m m e d i a t e practical importance

Document 60

Filed 01/29/2007

Page 2 of 7

There's only one way to determine whether users are satisfied: ask them. I know of no such research, nor has Mr. Nesbitt offered any. Evidently AFP elected not to pursue such research. If such research was undertaken, it would require the testing of multiple informational package schemes because users might find current formats unsatisfactory if they were offered alternatives. Research shows that consumers of online news use more than one news site. This implies that no single online source of news, including Google News, is satisfying its users. Moreover, and fundamentally, Mr. Nesbitt never defines user satisfaction, making it impossible to determine whether it has been achieved. The research findings offered by Mr. Nesbitt show where uses look, but not why they look, casting further doubt on his conclusions.

Brass Tacks Design

Case 1:05-cv-00546-GK

Document 60

Filed 01/29/2007

Page 3 of 7

PAGE-BY-PAGE ANALYSIS The evidence presented by Mr. Nesbitt does not provide meaningful answers to defined questions. Having noting some the of key weaknesses in the overview, I will now go through page by page:

Page 5: "These clients still subscribe to their news services and select,
With the growing importance of new media and online delivery of information, news service clients have adapted their print, radio, or broadcast products for delivery on the Web. These clients still subscribe to their news services and select, edit, and disseminate news and information in a Web format that has increasingly become the sole source of news and information for a growing segment of the global population. Google, Inc., under its Google News label, has created a format whereby it electronically gleans material from other sites on the Web and, without editing, displays this information on the Google News Web site.

edit, and disseminate news and information in a Web format that has increasingly become the sole source of news and information for a growing segment of the global population." According to Editor & Publisher, data released by the Gallup Organization recently shows that the rate at which people are using the Internet for news is dropping, and that use of newspapers for news is leveling off. This new data casts doubt on Mr. Nesbitt's claim of increased use of online news.

5

Page 6: This report will show that the three major items in
This report will show that the three major items in contention ­ headlines, blurbs, and images ­ as an informational package is more than enough to satisfy most users.

contention ­ headlines, blurbs, and images ­ as an informational package is more than enough to satisfy most users. Nowhere in Mr. Nesbitt's report does he provide any data or a direct connection between information packages and user satisfaction. Nor does he define his core concept of user "satisfaction." Furthermore, Eyetrack III ( which Mr. Nesbitt relied upon) did not

6

address nor measure user "satisfaction."

Preface
While there is a long, established understanding of reading mechanics in print media, the world of online media is the new "Wild West." However, the bulk of information from numerous new media studies points to new patterns readers or browsers use when accessing information on the Web.

Page 7: "However, the bulk of information from numerous new media studies points to new patterns readers or browsers use when accessing information on the Web." As this statements implies, there is considerable information that points the other way. On Page 60 Mr. Nesbitt says, "Whether by survey or eyetrack, there are as many points of view as there are studies. There is a lot of contradictory information as a result of this research, many different interpretations of the data, and some outright dismissals..." This

7

admission casts doubt on the accuracy of these so-called patterns.

Case 1:05-cv-00546-GK

Document 60

Filed 01/29/2007

Page 4 of 7

Page 8: "Extrapolating information from a number of studies, it
The major point of entry in print media is most often through an image, but some studies of Web pages show that 25 to 35 percent of users enter through a picture; the rest enter through text. However, many of these studies use samples where images are lacking and the predominant material on the page or site is text. Not surprisingly, the text is short, to capture the user's attention. Extrapolating information from a number of studies, it becomes clear that the combination of short text, e.g., "keyword" -laden headlines and a paragraph of a single thought (one to two sentences), coupled with an image, has a higher draw than either alone.

becomes clear that the combination of short text, e.g., "keyword" -laden headlines and a paragraph of a single thought (one to two sentences), coupled with an image, has a higher draw than either alone." Mr. Nesbitt is doing just that ­ extrapolating ­ the very thing Eyetrack managers said they could not do. Any content vehicle (image, headline, blurb or combination) could be shown to outdraw another if the content was relevant, compelling, interesting and useful to the user. It's not the

8

form that creates interest; it's the content that draws attention regardless of format. Another driver of interest is size. Large headlines outdraw small ones, even those packaged with images. Page 9: "Many of these basic responses start with our ability to read and visually comprehend in print. As we have moved into the digital age, these responses are not weakened but actually become stronger." Mr. Nesbitt offers no substantiation or empirical evidence for this claim that visual comprehension has intensified in the digital age. If Mr. Nesbitt is correct, then why do people print out stories rather than read them online?

Perspective
To understand how the use, by Google, of AFP generated material is identical to AFP use, it is necessary to go beyond the lack of physical or verbal changes in the material by Google and to understand how readers/viewers respond to text and images. What they see, how they see it, and what action they take is based on a number of fundamental responses on the part of the reader/viewer.

Many of these basic responses start with our ability to read and visually comprehend in print. As we have moved into the digital age, these responses are not weakened but actually become stronger.
9

Page 10: The addition of visual elements to a page in large numbers is
The addition of visual elements to a page in large numbers is a relatively recent occurrence. The majority of news and information was delivered to the reader through text. In newspapers. as the major source of news and information in the 1900s, the importance of the text was paramount. The cost of engravings was prohibitive and the results were seen as superfluous.
Baltimore Sun, October 2, 1853

a relatively recent occurrence. The cost of engravings was prohibitive and the results were seen as superfluous." According to Charles Swedlund (History of Photography, 1974), "In 1897 a method was developed for printing halftones on high-speed presses, and by the turn of the century people had become accustomed to seeing photographs in their daily newspapers." Furthermore, Mr. Nesbitt offers no substantiation for his claim that images were ever seen as superfluous. It is just as likely to conclude that the number of published images has increased as the process of publishing them has become easier and cheaper, rather than images being seen as superfluous in the past. Page 11: "Even as late as the mid-1960s, newspapers weren't

10

Even as late as the mid1960s, newspapers weren't convinced that pictures (halftones) were that important. However, with the growing influence of television, some newspapers saw images as the means to boost their competitive value in the marketplace, the advantage of immediacy having earlier eroded with the mass marketing of radio.
Baltimore Sun, August 5, 1965

convinced that pictures (halftones) were that important." Photos from the Kennedy assassination and the subsequent shooting of Lee Harvey Oswald were splashed across newspapers in 1963. Clearly, photos were important to newspapers in the early 1960s.

11

Case 1:05-cv-00546-GK
We are seeing a convergence of presentation style in both print and online. The Washington Examiner is just one example of a print newspaper that has moved to headline, images and a sentence as a "menu" on the front page.

Document 60

Filed 01/29/2007

Page 5 of 7

Pages 12 and 14: The Washington Examiner is just one example of a print newspaper that has moved to headline, images and a sentence as a "menu" on the front page. There has even been an attempt to recreate the web presence in print. The Washington Examiner is not an attempt to recreate the web presence in print. The Washington Examiner is a tabloid-format newspaper, and like most tabs, it puts headlines and images on its front page. But it does
The Examiner, December 1, 2006 14

not use 9-point text on its front as do all broadsheet newspapers. Tabloid newspapers have used this means of presentation for generations; examples include The New York Daily News, the New York Post, Newsday, the Philadelphia Daily News, and the Rocky Mountain News. Page 19: On the web we often play the various elements at the same

The process of page or screen design is one of manipulation, whereby the individual placing elements on a platform can create areas of visual impact, visual pathways through the page or screen, and a strong hierarchical structure, if desired.

level of importance, eliminating as much visual bias as possible. Elements are not placed on the same level of importance by choice or to eliminate visual bias. Rather, this democratization of content is a limitation of current online content-management systems (CMS), and the limited resources news organizations bring to bear upon production of online news. This phenomenon has spawned the expression "shovelware," for software that "shovels" news from print to online without

A printed newspaper will require a stronger hierarchy than a web page in most cases. On the web we often play the various elements at the same level of importance, eliminating as much visual bias as possible.

19

reprocessing.

Page 20: All things being equal, on a given platform, the Visual
All things being equal, on a given platform, the Visual Pyramid applies.

Pyramid applies: 1. Images 2. Graphics & Illustrations 3. Headlines 4. Large & Unusual Type 5. White Space 6. Color 7. Rules & Borders 8. Text All things are never equal: No two stories, headlines, photographs, images, etc. are ever equal in news value, importance, meaning, usefulness, relevance or interest. And no two people are likely to rate any news content the same way. Any proposition based on "all things being equal" cannot be supported if the thing being considered is news content. Furthermore, Google News does not take most of these "Visual Pyramid" elements from AFP.

The visual pyramid
1. Images 2. Graphics & Illustrations 3. Headlines 4. Large & Unusual Type 5. White Space 6. Color 7. Rules & Borders 8. Text
20

Case 1:05-cv-00546-GK
In this example, the reader is drawn to the largest visual point on the page, the image. From this point, the reader's eye tends to move on to the next page. We will have led the reader away from the rest of the material at the top of the page.

Document 60

Filed 01/29/2007

Page 6 of 7

Page 21: In this example, the reader is drawn to the largest visual point on the page, the image. The reader is drawn to the image because it is the only element worthy of the reader's attention. The reader is not given a viable alternative. This is not a example from the real word, so it has no application in any discussion of the perception of news in any form.

The designer has manipulated the viewer to a specific point on the page.

21

Page 22: In this example, we have manipulated the reader to move
In this example, we have manipulated the reader to move straight to the center of the page. By lifting the image, we have created a new, different area of visual impact.

straight to the center of the page. The bigger manipulation is the attempt to pass off this example as a realistic representation of a news page.

22

Pages 23-24: There is only one way to make text number one on the Visual Pyramid (by eliminating all other elements). There are several other ways to make text number one on the Visual Pyramid: Make it larger than other elements; arrange it in a visually provocative shape; surround it with oceans of white space; use words that have significant meaning to the reader. For instance, if the words said "free, ""sex" or "money" (or all three), they could attract more attention than other visual elements.
24

Pages 25-26: The minute we add a picture, even the size of a postage stamp, the visual balance changes, and the image becomes number one. Mr. Nesbitt contends that he can change the visual balance by adding an image, even one as small as a postage stamp. However, it is not the image that changes the visual balance, but the introduction of a contrasting element. The visual balance could just as easily been changed by the introduction of one sentence in bold type.
26

Case 1:05-cv-00546-GK

Document 60

Filed 01/29/2007

Page 7 of 7

Pages 27-28: Changing the visual dynamic is easy. As seen in the next example, the picture has dropped to second place on the Visual Pyramid, the headline has become number one. On page 20, Mr. Nesbitt claims a visual pyramid. On this page he upsets that pyramid. A pyramid which claims to represent a hierarchy is of little use if the hierarchy is fluid.

28

Pages 29-30: The small line of type serving as a headline is of equal weight as the body of the text. The small line of type serving as a headline comes first, and is likely to be read first, so it is not equal to the body of the text. Furthermore, the text can be more meaningful than the headline depending what it says and its meaning to the user.

30

Pages 31-32: With the addition of a stronger headline in the next example, the user is drawn to the strongest visual element and quickly understands the meaning of the content. We cannot say with certainty that the user understands the meaning of the content unless we can read the user's mind.

32

Acuity
Size doesn't matter...that much. We are faced daily with seeing, understanding at-a-glance, and reading the content of elements around us that are small in size and rich in information.

Page 37: Size doesn't matter...that much. We are faced daily with seeing, understanding at-a-glance, and reading the content of elements around us that are small in size and rich in information. Mr. Nesbitt's statement is general and does not speak specifically to the issues of this case. However, on page 28, Mr. Nesbitt demonstrates that "A very large Headline" attracts more attention than an image, and more so than the smaller text surrounding it. Thus proving that size does matter...a lot.

37