Purchasing Pixels

598

December 6, 2003

One of the features of the PACA International Conference in New Orleans was a panel of
six art buyers (3 from New York, 1 from Chicago and 2 from New Orleans) who explained how
they work and what they need from stock photo archives. Some of their comments were
predictable, but a few offered insights to possible new trends.

Print Catalogs

When asked if they liked to receive print catalogs the general consensus was, "We don't
use the books. Stop sending them." It was clear than everyone on the panel is doing
their research online. They acknowledged that, "the books are sometimes inspirational",
but they seldom, if ever, use them when they are looking for a specific image for a
project. The process of using the books is simply too slow and they continually
emphasized that it is important for them to be able to find the image they need quickly.

They also made the point that while they use Getty and Corbis a lot they don't always
find what they want on these sites and they are actively seeking other resources. When
they are looking for specialized imagery they definitely go to sites that they know
specialize in the particular subject matter. For image suppliers looking for ways to
market to the picture buying community they suggested SHORT e-mails, not too frequent,
that quickly focus on the agency's specialty. Use hot links for additional information
rather than trying to pack too much into the e-mail. They tend to add addresses of such
sites to their address book. (One of the implications of this statement is that agencies
that have a few strong specialties might be advised to create separate web addresses for
each specialty, and market each web address very specifically.) Direct mail promotions
were also mentioned, but the panel seemed to agree that e-mail may be more effective.

Finding small stock agents and the specific collections seemed to be a real dilemma for
them. Rod LeMeier, a New Orleans Graphic Designer, said he recently spent a full day
just looking for new stock photo sources. Sabrina Olivia, who has worked as an art buyer
at several top New York agencies for a full range of clients from pharmaceuticals to
beverages said, "a lot of my art directors don't want to see the 'standard image'. They
want the creative image that is different. There is still a calling for the happy family,
but the art directors want something more, something different." These art buyers are
constantly looking for something they haven't seen before. One of the problems in going
to the same one-stop shops frequently is that buyers get the feeling they are seeing the same
images over and over again.

They made the point that because clients expect things so fast in today's environment
creatives must rely on the Internet to provide that speed. The result is that it has becomes
almost impossible to review anything that is not on the Internet.

Do They Want Research Help?

Most indicated that they prefer to do their own research online and tend not to ask an
agency or an outside researcher for help. They complained that when they ask for research
often the agency people don't find what they consider the "best pictures" on the agency's
site. For this reason they find it more productive to do their own research. They
indicated that sometimes when they are very busy and there are many images to go through
they may ask for research to narrow the selection, but this method of operation is
certainly not their preference.

They tend to use the search option that lets them download 100 thumbnails at a time and
have no problem going through 1,000 or more images if that is what the search provides.
They would like to see a better system that effectively prioritizes the most relevant
images and brings them to the top of the search. However, to accomplish this, not only
is extensive (appropriate) keywording required so the buyer can effectively narrow
searches, but somehow it would be necessary to understand what each buyer considers
appropriate (which will vary from buyer to buyer). Given the complexity that this entails
it doesn't seem likely that any supplier will fully achieve this goal anytime in the near
future.

They like systems like Getty's for choosing specific meanings when words have more than
one meaning (like turkey). They also like the search engines that allow them to do a
search within a search (for the purpose of narrowing) the group of images. This is
particularly useful when the return from the basic search is several thousand images.
And they like the sites that offer keyword lists for every image so they can pull up
similar images when they click on a word within the list.

Better Keywording

They emphasized the need for better keywording, particularly with relations to the
specifics of the image. Esin Goknar, an editor at Conde Nast Traveler, pointed out that
the weakness of many sites is that their keywording is too broad, generic and concept
oriented. Such sites often ignores the specific keywords which is what she and other
editorial focused
picture researchers need. Everyone on the panel seemed to recognize that "more keywords
are better". While they expect all keywords to deliver "relevant" images, they are fully
comfortable with using multiple words to zero in on what they are looking for. It
appeared that they seldom use single "general" words when searching unless they cannot
find anything with the specifics and need to back off into more general terms.

Karen Broderick, a freelance book cover researcher from New York, complained that she can
spend hours searching online because the keywording is not succinct enough. "If anything
could be improved time wise it is keywording so I don't have to go through 1,000 images
to find what I need," she said. While indicating that all sites need improvement she said
Getty's keywording is better than Corbis and Photonic is terrible, even though she uses
Photonica a lot looking for concept images for book covers.

One key observation after listening to this discussion was that these people are much
more sophisticated in using online search and its full capabilities than many stock
agents and photographers believe. Some agents have felt that they need to use very basic,
generic terms, because art buyers looking for images online wouldn't know how to use more
sophisticated search techniques. That may have been true in the United States two or
three years ago, but if this group was representative of all art buyers, it certainly
isn't true today.

Several problems arise when the focus is on limiting the number of keywords. First, you
tend to get too many hits when the focus is on broad category terms. In addition, if
words that will help the buyer focus in on the specifics of the image are not attached,
there is no easy way for the buyer to eliminate a lot of irrelevant images from the
search -- and it is these irrelevant images that bug the buyers. Finally if the specific
unique characteristics of the image are not included (particularly true in specific place
locations -- Is the picture of Tokyo, or a specific place in Tokyo?) there is no way for
buyers who are looking for the Shinjiku district to find that image. If the six buyers
on this panel are representative, this lack of adequate keywording is clearly a major
problem.

Googling Images

Some sellers argue that if the buyer can't find what they want with a very narrow search
they will simply use the broader term like "Tokyo" and go through all the images until
they find what they want. One of the problems here is that the average art buyer may not
know whether a picture was taken in Shinjiku, or not, if they are not told.

But more worrying for a lot of companies that have developed big databases is another
search method that Esin Goknar says is now being used by art directors to help them find
exactly what they need. She calls it "Googling Images" and says the term is now being
used (in New York, at least) as a verb. She says that, "In the last few searches we are
using the images we find by Googling more often than the material we find on Corbis or
Getty. Using keyword search takes too long for us to go through the returns these companies
offer in order to fine the right image. Time is absolutely essential."

As best I can determine, the way Googling works is that you go to Google and enter the
specific thing you are looking for plus the term "photo", as in "Shinjiku Japan photo".
The images you get using this technique are not in any of the major databases because
Google has no way of getting at the keywords embedded in databases. Instead, the images
returned tend to be from photographer web sites who specializes in the subject matter being requested,
or from local associations, local businesses, or government organizations that have an interest
in the subject. From the few searches I have done to try to understand this system, I
believe buyers tend to get a lot of crap, but they also also tend to get images that are right
ON TOPIC.

The next problem for the buyer is to figure out how to negotiate rights. On some of these
sites it is difficult to figure out who to contact, and how to get a hi-res file. Since
many of these people are not in the business of licensing rights the negotiations could
be difficult. On the other hand some of the organizations may be happy to give away
their images in order to get a little publicity.

When I raised this issue with Ms. Goknar, she acknowledged that clearing rights for these
Googled images was sometimes difficult. She also repeatedly made the point that in the
editorial business speed is of the essence and that's why they are using the internet.
But she emphasized that where the speed is most important is in knowing that the right
image exists and where to find it. Google may help the buyer find an image "on subject"
quicker that in some of the large databases where a lot of the specifics are not covered.
And maybe, once the buyer has found the image, they have a little more flexibility in
completing negotiations and getting delivery. All this is not totally clear. But it is
important to keep in mind that for Conde Nast Googling is working.

If "Googling Images" does take hold as a strategy for finding images it raises a lot of
interesting issues for the industry. It may offer a way for specialists with small files
to participate more directly and cut out a lot of the middlemen. For years I have advised
photographers not to spend money setting up their own web sites because it would be
almost impossible for them to let a significant number of customers know their web sites
exist. That advice may now be totally outdated.

If the photographer does set up his own site, all the keywords must be transparent to
search engines like Google. I think that means they must be listed as meta-data in the
header of the site, or the header of specific images if Google can search specific image
files. It is also important to structure the site so that only the images that meet the
criteria of a specific request appear when those keywords are entered. Thus, a site based
on basic HTML code may be more effective for this type of search than some of the more
elaborate databases that are available. This whole issue needs to be explored in a lot
more detail, but it may be particularly useful for photographers who photograph
personalities or locations that can be very specifically identified. Photographers need
to learn how to set up their sites to optimize their access through Google.

At the same time, Googling may be worrying for many stock agencies. Since photographers
must now to do more and more of the work to get their images seen -- and give up greater
and greater percentages to their agents -- it may be more practical for many of them to
market a greater portion of their work directly. Also, if a segment of the market can
find the images they need without going to the traditional photo archives or portals it
may mean that the total number of images required from these traditional sources will get
smaller and smaller.

Number of Images They Will Review

The number of images that a search produces did not seem to be a big inhibitor for these
art buyers. They all agreed that they liked to use sites that allowed them to bring up
100 thumbnails, or more, at a time. They also seemed to have no problem in going through
1,000 or more thumbnails to find the right image. While they need to find the right image
quickly with fast connections reviewing 1,000 or more thumbnails didn't seem to be a
problem. Some indicated that they might spend a couple days searching
for the right image when they had the time.

One useful insight here is that many photographers seem to feel that it is important for
their images to be among the first 20 to 50 in any search result because the art
directors would not have the patience to go much deeper. This does not seem to be the
case. Although the buyers did say that if they went to a site that only allowed them to
download 12 thunbnails at a time they would not go through nearly as many pages as they
would if they can return 100 images at a time.

When they get a search result of several thousand images they would like for the most
relevant images come up first. Currently that often doesn't happen. (One problem here is
that the image that one art director considers most relevant may not be what the next art
director, using the same search terms, will feel in most relevant. Improvements can
probably be made in bringing up relevant images first, but this may be a problem that has
no good solution.)

One-Stop Shopping

They all emphasized that "one-stop shopping" isn't all that important and they are more
than willing to use several different sites if the sites offer images that are on-target
and are something different from what they have seen in the past. They seem to be
constantly looking for new sources that will show them something different. One of their
complaints about RF is they feel the same images are on every site and they keep seeing
them over and over again.

Karen Broderick said, "I'm not going to Corbis or Getty for the one-stop shopping idea.
I go to Corbis for the Bettmann Archives. Do I like it that Corbis bought Bettmann. No.
Not at all. They have butchered the archive. Most of their collection is sitting in a
warehouse somewhere. They haven't bothered to scan it all. I go to Getty because they
have Tony Stone and Hulton Archives. These are agencies I went to before they were
consolidated. Now I have to go to Getty to get to those agencies, but the consolidation
has nothing to do with why I go to them. I preferred it when they were smaller and the
people working there actually knew what was in their archive."

Providing Multiple Versions of Ads

One of the buyer's complaints was that some companies charge them (up to $150) for every
hi-res download even though some of those images are not use in the final ad. The people
preparing ads said they are often required to present five or six versions of an ad
(comps) to the client before the one to be used is finally selected. They need hi-res
files for each version, but only want to pay for the one that actually goes into
production. Since often the images for the various comps come from several different
sources, it is possible that some companies will deliver images for these comps and
receive no compensation.

Pricing Flexibility and Payment

As might be expected there are some problems with price. They want unique images and
they don't want other customers using the same unique image they use in an ad, but they
don't want to pay the high prices for exclusivity.

They want bulk rates based on the number of images they use in a year. While many of the
larger companies already supply these kind of discounts to their larger customers it
would seem from the panel's comments that many of the smaller to mid-sized buyers have
not been able to negotiate volume discounts.

They like fixed prices when the discounts work for them, but they don't like them when
the fixed price works against them on a particular project. Karen told of one publisher
who had a preferred pricing arrangement with an agency. The agency had offered a $300
price for images for each one used on a book cover. (A very low price in my opinion.)
Then the publisher designed a cover that had 11 photographs on it. Ten of them were
postage stamp size and most of them were archival. The agency wanted $300 per picture
according to their agreement. (I'm pretty sure the agency established its discounted
price schedule -- not only on cover pictures, but on 1/4 pages and all other uses as well --
based on the volume of images this publisher had purchased the previous year.) The
publisher balked at paying this price. In this case the agency wouldn't budge from its
already low price. So, in the end, the publisher went to another agency that charged them
$50 per photo for the postage size pictures. (P.S. Given that these were archival
pictures it doesn't take a lot of imagination to figure out which two agencies might have
been involved.)

Cheryl Bersano, a graphic designer and art buyer for major Chicago agencies, made the
point that art buyers must, "constantly walk the line between the aesthetic and the
business worlds. Buyers in any capacity have to answer to the account people. There is a
perception -- I don't care how much production value went into the image -- that if it
hits that channel and becomes a stock image we can wheel and deal and that it should come
to us more cost effectively and quicker than an assignment photograph. Clients often
say, 'We don't have a lot of money, let's use stock'."

One person made the point that they don't like to deal with all the paperwork and
negotiations agencies throw at them. "We're a huge advertising agency and you should
trust us that you will get paid at the end of the day. Any hurdle that slows us down will
lead to our using someone else, unfortunately." Needless to say many in audience were not
sympathetic to this attitude given the number of times they have been burned by large
organizations when they didn't nail down all the rights and paperwork before the sale.
But, it is an attitude we have to live with.

These are only a few of the highlights of the wide ranging two hour discussion moderated
by Sharon Dodge. With the success of this "Buyers Panel" and the one in Miami last year,
buyers panels are likely to become a fixture in future PACA Conferences, but they are
only a small part of the valuable information and networking that is exchanged in these
conferences. Archive executives that didn't attend may get a hint of what they missed
from this article, and want to consider attending the PACA's 9th Annual International
Conference that will be in Las Vegas in November 2004.

The other major international information and networking event that everyone in the
business needs to attend is the CEPIC International Congress that will be held next year
from June 10 to June 13, 2004 in Copenhagen. For more information go to their web site
at www.cepic.org .

Finally, the other person on the panel who was not quoted above
was Dave Roberts, a graphic designer from New Orleans, a teacher at Tulane University,
and an AIGA board member. While I have not used any of his specific quotes in this
article his contribution to the overall discussion was very valuable, as was everything
all the panel members had to say.