Manifest Technology Blog
| DVI Tech
| Site Map
TECHNOLOGY & SOCIETY ARTICLES
| PC Video
| Web Media
| DVD & CD
| Portable Media
| Digital Imaging
| Wireless Media
| Home Media
| Tech & Society
Technology & Society:
| Technology & Society Articles
| PC Technology References
The Art and Science of Google Ads (3/2007)
by Douglas Dixon
Business and Competition
Search Marketing Optimization
What's your business? For Doug Kerwin and Richard Sbarro, as they left their
IT positions at Merrill Lynch in 2002 to form an Internet start-up called
Metaverse Corporation, the business was "content management" --
back-end software systems for large websites.
But when the dot-com crash and tech implosion wiped out their market, what
was previously a side project transformed into their main business -- an online
art gallery and store called FulcrumGallery.com (www.fulcrumgallery.com).
After only 2 1/2 years of operation the business is now doing $4 million in
sales and supporting a staff of over 40 people at its offices in central New
But now the business is transforming again. Since most prospective customers
look for art prints through searches on Google, the company is back to building
technology -- The custom search marketing optimization software that is required
to manage the over a million Google ads that Fulcrum Gallery runs that respond
to specific searches for art, including more than 200,000 works, 10,000 artists,
and many categories.
Why the importance of Google for online stores? "We do a lot of e-mail
marketing to our own customer base," says Kerwin, founder and president,
"but for new acquisitions, search marketing is where it's at. We would
never put a billboard on I-95 saying 'Interested in art? Come to our site!' --
999 of 1,000 people would pass it by and not care, and for the one guy who cares
to respond, you would need a hundred of him before anybody buys something. The
numbers don't add up."
In comparison, search marketing is about responding to specific interests.
Kerwin, "Someone at home types into Google something they're specifically
looking for, like ' snow covered trees,' and sees our ad, which says exactly
what they're looking for about snow-covered trees, and they click. The website
doesn't take them to the home page, it takes them to exactly what they're
looking for -- you see couple hundred pieces of art of snow-covered trees."
"It's very targeted," says Kerwin, "and that's a refreshing
thing. In the software business it's direct sales, where you're interrupting
someone, you're hijacking their day when they have important things to do. With
the new search marketing, the people are looking for you, they're calling us up
because they are interested in something. That's a refreshing side to be on.
There's nothing more targeted than when they're looking for exactly what we
have, at the exact time that they're looking to make a purchase."
To see how this works, go to Google and type in a general search like "art
prints," or a more specific search like "snowy mountains art." In
addition to returning a list of the resulting hits, Google also shows
"Sponsored Links" in the right column and at the top of the page.
For a search on art topics, Google typically with ads from FulcrumGallery.com
and its competitors including Art.com (www.art.com)
and AllPosters.com (www.allposters.com).
These are Google AdWords (www.google.com/ads)
-- posted by businesses that signed up to post ads when users enter specified
keywords, and then pay a cost per click when viewers click through to the
The challenge for businesses like Fulcrum Gallery, of course, is to choose
the right keywords and manage the advertising budget to achieve a reasonable
return. This is even more complex because Google manages its AdWords program
through a process of bidding for keywords, which requires constant monitoring
"We put out a bid per click on all these keywords," says Kerwin,
"and if no one's looking you could spend your way out of control." In
addition, "when you're in that range of a million keywords under your
control you can't do it by hand." The answer for Fulcrum Gallery has been
to develop custom software to automate the process, and to interface directly to
Google's software. Says Kerwin, "We have software to automate placing new
advertising. The big thing we're working on now is software to automate bid
changes, maintaining accounts where we need to make a change to a thousand
keywords at a time."
"We have a unique history with this company," says Kerwin. "We
started out in 2001 / 2002 really as content management software company. I was
a software development manager at Merrill Lynch, and another founder, Richard
Sbarro, worked for me. We did contact management software there, and then came
up with our own ideas." At the same time, Microsoft was releasing its next
generation of programming tools, the ".NET framework," and Kerwin's
group was among the first to start developing software for it.
"We had this idea for a content management system," says Kerwin.
This was the back-end software infrastructure to simplify managing a large
website -- "a program to let you change your website without having to have
technical knowledge," says Kerwin, "to allow non-technical business
users to change their website."
And how could a small group in New Jersey compete in this niche? Says Kerwin,
"our twist on why the world deserved another content management system was
'software as services,' like Salesforce.com, using the Microsoft .NET way of
doing things." Instead of selling a software package to be purchased, with
the ongoing pain of installations, support, and upgrades, they would use the new
model of a Web-based monthly subscription service.
However, says Kerwin, "We started in 2002, which was the worst time in
the world to do such a thing. We had come out of this era of people throwing in
Porsches as signing bonuses, and all of a sudden all that got flushed down the
toilet. People were just lucky to hang on with their jobs."
"And right in that period we decided that's the time to leave our paying
jobs and start the software company," he says. "It was not a great
time. We were never able to make a real business out of content
Part of the reason that Kerwin did not account for the dot-com bust was that
being at Merrill Lynch delayed the impact: "We were at a bigger
company," he says, "and we didn't feel it as much as other people. We
were still in the mindset that if you do something cool with the technology,
people will come."
"It started as an idea," says Kerwin. "We started writing
software on the side, and doing our own competitive research. We saw that this
could probably be a business, and slowly made the decision to do it." In
addition, he says, "I guess it was a kind of a delayed reaction for me from
being at bigger company. It wasn't as exciting for me to be there."
Kerwin had started at Merrill Lynch at age 21. "This is a little
unbelievable," he says. "I had people working for me that were more
than twice my age. I had to cover that up, so nobody really knew. When I left
everyone thought I was in my thirties, even though I am just turning thirty
"It was definitely a fun ride being there, starting out at 21 at that
position. Then all of a sudden it stopped being exciting. We started seeing
people leaving. What was cool yesterday in the technology industry was all over,
saying ROI and technology in the same sentence was laughable."
"I had it good at Merrill Lynch," says Kerwin. "But I had more
or less plateaued. I could keep the job, and get incremental raises and more
responsibilities. But everything had been on such an up scale to get that point.
From then on it was not going to go up dramatically anymore." Instead, he
says, "I wanted something with more control, to move on in my personal
career to go to the next level. And also with this idea, we had some code, and
it built some momentum, and then it got to this point where we said 'let's do
"I started at its Merrill in 1999," says Kerwin, "we got this
idea in 2001, and we left to do this in mid-2002." The software was close
to ready, and they had one production installation in a beta stage. The first
customer was NWL Transformers in Bordentown. "I go way back with
them," says Kerwin. "My first job there was when I was 15 doing C
[language] programming for OS-2 [IBM's competitor to Microsoft Windows]".
The new company was called Metaverse Corporation (www.metaverse.cc).
"We raised a little seek capital," says Kerwin, "from private
investors, people I knew professionally, people who were invested in me rather
than in ideas." The company promoted versions of its software for small
businesses and enterprises in September 2002, and promoted new capabilities with
Microsoft through May 2003.
However, it was quickly clear that the idea was not working. "We saw
within six months that this was nowhere near ready to break even," says
Kerwin, "our burn rate was much too high to keep going." The company
had four software developers (including Kerwin and Sbarro) and, says Kerwin,
"two salespeople that didn't work. I learned the lesson that you need to
learn to sell your own product first before you hire salespeople. But I came
from an engineering background so I thought I should hire sales experts."
Without a deep well of venture capital, says Kerwin, "in order to keep
paying the bills, we switched to something that automatically works, to
consulting." The company was then down to three people. "We did pretty
well at consulting," he says, "but in my mind that is not so much a
business, it was very similar to having a job."
Meanwhile, there was a little side project called Fulcrum Gallery,
which was "entirely an accident," says Kerwin, as it took on a life of
its own. Fulcrum Gallery was originally designed as the demonstration site for
the content management software. (The site explains that a "fulcrum"
is defined as a hinge or point of support on which a lever pivots to transmit
force, so "we thought this aptly portrayed the role of art in society, and
used it to derive our gallery's name."
Metaverse launched the gallery site in October 2002, and used it to
demonstrate its website software to prospective clients, says Kerwin, "to
show them how the software works by modifying it." Other companies would
typically build simple faked-up demos, but, he says, "since we have to
spend a bunch of time and energy on it, we decided to make a demo site that's
kind of cool on its own, and something that we were interested in." So they
created an art gallery.
However, Fulcrum Gallery was just a content site that displayed artwork.
There was no e-commerce, says Kerwin, "we still had no idea we could sell
this." But, as the company switched to consulting, the gallery "grew a
life of its own. People would e-mail us in and ask to add their artwork to the
site. It was easy enough to do that with the software, so we started doing
"I don't know how they found it," he says. "We never did
anything to promote it. We added a few artists over the months, and then we
started getting e-mails saying 'I wanted to let you know I sold a painting, and
they say they found it on your site, so I want to say thanks.'"
"That was the first we thought that we would like to do something about
this," says Kerwin. However, the staff was all busy consulting, and nobody
had time to turn the gallery into an e-commerce site, develop new software, and
then run the business. That's where George DiLorenzo came on the scene. After
attending the same high school as Kerwin and Sbarro, DiLorenzo graduated from
the Fashion Institute of Technology in 2003, and had most recently been an art
director at Fantaseyes in Manhattan.
"So he volunteered," says Kerwin, "and in the beginning he was
on commission. It was meager beginnings: it was a demo site, it was a bizarre
accident. We said let's give it a try, with no capital, no software resources.
We put an 'Add to Cart' button on the page and then George worked with the
artists." In September 2003, Metaverse relaunched the new Fulcrum Gallery
with e-commerce for online shopping, as "an online gallery of original and
limited edition artwork, with a current inventory of original oil and acrylic
paintings, photographs, mosaics, etchings, sculptures, and digital prints,
priced from under a hundred dollars to over $50,000." Artists could exhibit
their work for free on the site, and Fulcrum collected a 30 percent commission
In the beginning, Fulcrum Gallery did not do direct sales. Instead it
brokered sales for original artists. But, says Kerwin, "it's tough to do
that, working with artists, convincing someone to buy a $1,000 oil painting
sight unseen online." The site graduated into art reproductions in 2004, he
says "which is all that we carry now. That's when this started to become a
"This is proven stuff that sells," he says, "$30 fine art
reproductions instead of $1,000 oil paintings." And even before
advertising, "people would find us through search and it started to take
off. Then we got into the Google advertising and that's what really made it take
And the business has grown, says Kerwin, "today we do over $4 million in
sales annually, up from zero two and a half years ago." Fulcrum Gallery
started out selling reproductions with some 20,000 pieces of art from two art
publishers, and today works with more than two dozen art publishers, offering
about 200,000 pieces online, from about 10,000 different artists.
In the beginning, Fulcrum Gallery operated out of Kerwin's 4000 square foot
home in Robbinsville, New Jersey. This was not a big deal at the beginning when
they were just brokering for artists. Says Kerwin, "it was just a few guys
typing on computers; We never even saw the product." But once the company
started selling reproduction prints, they needed to stock and ship the artwork.
And once they got into custom framing, they needed to build and mount the
frames. "We could have gone to outsourcing," says Kerwin, "but
when we looked at the cost savings of doing it ourselves, I couldn't bear to pay
someone else to do it." So they bought framing equipment and continued
upgrading it, so, he says, "there's no better equipment than what we have
now -- and we had that in my house."
"It was quite a scene," he says, "pushing a home-based
business to the limit. We had 10 to 11 people, and 20 foot miter saws in my
basement. We were probably shipping around a hundred pieces of art a day. The
FedEx people couldn't believe what we were doing. The framing vendors who
dropped off supplies were amazed; they couldn't believe we were doing that kind
By January 2006, says Kerwin, "we had to get out of my home. We were
supposed to move to this building in January, but this was brand-new
construction, and it was delayed. We wound up in existing space up front in this
same facility. It was actually smaller than my house." The business then
moved into its current space in September 2006.
The current building is 5,000 square feet, with both office and production
space, including the inventory of prints and shipping tubes for unframed prints,
plus framing and assembly equipment including the molding supplies, wood saws,
glass trimmers, and shipping cartons. The company also uses an additional 65,000
square foot warehouse across the street for storage. "We buy 20,000 tubes
at a shot," says Kerwin, "that takes up a lot of space."
The staff is also growing. "We need more
people in operations," says Kerwin, "including custom framers for
packaging." But the focus of growing the business is marketing and software
development. He has three full-time software developers, "not including
myself; I don't have time for software anymore." And for marketing, says
Kerwin, "four to five months ago I had no marketing, it was just me doing
all the marketing. Now I have four people, and am looking to expand that group
As a result, the company is running out of space again. "I already have
to move out," says Kerwin. He wants to say close to the Princeton area,
"so I don't lose half of my people." However, "the challenge for
us is that we need flex space, which is kind of unusual in this area. We need
office space and production space in the same location."
The content behind the site is expanding as well. Kerwin is still adding new
publishers, and adding new artists and artwork. He also plans to expand into
other product categories related to artwork. "People already on our site
looking to decorate their walls," he says, "and maybe they're looking
to decorate other ways, with a lamp and other product areas that have synergy
with what we already sell. With the kind of traffic that we already have, we
should be able to add those products."
Kerwin was recruited to Merrill Lynch by a placement firm. He was previously
with Towers Perrin in Philadelphia. "I was only there for a short
while," says Kerwin. "The offer from Merrill was more than double what
I was being paid there, so that was kind of a no-brainer."
Kerwin actually began with computers and programming at a young age. "My
first computer was a Texas Instruments TI 99," he says, "back when you
had to have a cassette tape to load programs. I graduated from that to an Apple
and then to a PC. I was working with an Apple in elementary school. I learned
how to program in the BASIC programming language when I was 10 making games, and
picked up C when I was 12. I got involved in programming early, I like the
computer; it's something I enjoy."
Kerwin got involved with programming with the advanced C language through a
bulletin board system (BBS), "the Internet before the Internet," he
says. "You would dial directly into someone's home computer, and you could
share files." He then set up a computer and software to host his own
bulletin board system. Even better, says Kerwin, by registering the software and
paying $50, the BBS software developer would send the C source code, "which
you could modify all you want."
So, he says "I ended up modifying it so much that it was an entirely
different piece of software." Then he offered his own source code for sale.
"When I was 13 or 14 years old," he says, "I had people sending
me checks to register. I made about $10,000 doing that. I remember my parents
wondered what I was doing, why all these people were sending me money."
But even with this deep involvement in software, says Kerwin,
"Ironically, when I was a senior in high school I got interested in
accounting, and somehow -- I don't entirely remember this thinking -- I thought
that would be a better career path." After high school, he had been
influenced by a youth pastor to enroll at a Baptist school, Cedarville
University in Ohio, where he majored in accounting, graduating in June 1999.
But Kerwin got back on the software track during an internship with Deloitte
& Touche in Princeton in his sophomore year of college. "All I was
doing there was software, and the people there talked me out of accounting; they
opened my eyes to the opportunity of IT. That was around 1997 when IT was just
starting to take off."
When he returned to college the following semester he switched majors to
management information systems, as combination of business and computer science.
"So I started out in computers," he says, "went to accounting,
and then came back home."
Kerwin's father also was making a similar transition to mainframe IT
technology consulting around the same time. "My dad had a career in
marketing," he says, "working for Conrail in Philadelphia for twenty
years. He was doing more and more programming, and when Conrail split apart he
found out his technology experience was more relevant than his managerial
While taking a full course load in his senior year at college, Kerwin picked
up a job as lead software developer for Appletree.com. "It was an online
dating service similar to eHarmony," he says, "but it never really got
"I talked my way into it," he says, "and I ended up hiring
five people who are in my major and knew how to do this stuff, and we built
their whole e-commerce site."
"I got the job through the university," he says. "There was a
job posting, although it was only a $10 an hour job in Columbus, Ohio doing Web
work. I was the first one to contact them about it, and transformed a $10 an
hour nothing job into $100,000 opportunity once they knew what I was capable of.
It was a mutually good situation. They got the same work at the least half of
what he would have paid otherwise, and a great situation for me."
"Part of it was good timing," he says. "Everyone was trying to
throw money at technology at that time. However, just because there's
opportunity doesn't mean everyone takes advantage of it."
"For me personally it was such a dramatic curve," says Kerwin,
"everything was just moving so quickly. You just could not ask for anything
better. I really loved it at Merrill. I had a good experience there. I did very
well there, but when it came to 2002, all that excitement seemed to plateau, and
most people were just lucky to keep their jobs at all."
"The next step had to be something bigger and more interesting. I always
wanted to make my own way, do my own thing. I'm building something that's
valuable to customers. I've created jobs in this area, and there's the
satisfaction of doing something that goes beyond just a paycheck in your
So where does Fulcrum Gallery go from here?
"This is just getting started," says Kerwin. "We've had
wonderful success so far; we've been growing between 10 and 20 percent each
month. But we are still a small business, and there is a lot more
"It's more than a local business," he says, "although in the
larger scheme of things it's small potatoes. But it has the potential to grow
and to be something very big. Our big competitor is Art.com [of Raleigh,
North Carolina], with about $150 million a year in revenue (www.art.com).
BareWalls.com [of Sainte Genevieve, Missouri] is second with $12 million
a year in sales (www.barewalls.com), and
third is us."
"That's one of the encouraging things about what we're doing. No one
handed us $3 million to make it. Art.com and AllPosters.com [of Emeryville,
California, which merged with Art.com] both have very similar histories; it's
striking how similar our histories are."
"That's kind of unique," he says, "to go from nothing to this
percentage of the market in 2 1/2 years. It's almost like this whole area of
selling artwork online was overlooked by the Amazons of the world. It was a
forgotten opportunity through that whole dot-com era; no one with public money
has really made an effort in this area. That makes some of the opportunity for
us, because it's enough to compete with an Art.com -- they are big, but they're
not an Amazon. If Amazon were to enter, they would probably buy someone. These
are custom-made products, and Amazon really doesn't get into that. They don't
even sell a lot of their products direct; somebody else fulfills the
The challenge for online businesses, then, is to buy the right keywords for
Google searches without blowing through the budget, and then convert the
resulting clicks into sales. "We are all bidding on the same search
words," says Kerwin, "There are a handful of people who do this well.
If you do a Google search for something like 'Andy Warhol prints' you'll find
out who all our competitors are."
Google sponsored ads from search for Warhol art
"It's about how you can monetize that traffic," he says,
"doing better at converting that traffic than others. We put all the work
into our software and our offering. We have our 800 number prominently displayed
on the site."
But how much can a company afford to bid? "I know I can afford to pay
more for that traffic than the small guys," says Kerwin, "because I
have a better chance of converting clicks into a sale. Maybe I can spend 40
cents on a keyword for a click, while my smaller competitor can afford to spend
only 20 cents. The better we can convert our sales, the more we can afford to do
the advertising to compete with others."
"Our offering is pretty similar to our competitors," he says.
"They still offer a few more things than we do, but I'd like to say our
online frame shop is better than theirs. It's an intensely competitive market
for keywords. You see this more and more, as more people wake up and realize
that this is a really great sales channel. Only now there are crowds of people
who are all bidding for the same thing -- there's only a fixed number of people
in the world searching for Andy Warhol prints. The trick is to come up with the
best possible offering so that you can afford to pay more for each
"It makes me choke every time I see how much we spend on Google,"
says Kerwin. "We give Google more money than anyone else: It's more than
payroll, it's more than cost of goods sold. It's outrageous; sometimes I feel
like we're in the business just to give Google money. But at the same time I
know why we continue to do this -- It's because it works. It's not spending
money putting up a billboard and crossing our fingers -- I know this converts,
as long as we keep it targeted and do it right. It's just really a matter of
So Kerwin is back to a major software development effort to manage those
million keywords. "We originally started building custom software for our
own purposes," he says, "but as we've gotten further along we see that
this software has usefulness far beyond our own purposes." Kerwin
developing the Fulcrum software with the possibility of taking it to market as a
separate search marketing optimization software product. "It's not for
everyone;" he says, "there only a few advertisers who have the scale
of keywords and feel the same pain we feel. There is no software available --
not from Google, not from anybody else -- that really addresses the unique
challenges from having so many keywords under management."
"This work with search marketing is another area we got into kind of by
accident, but it's potentially bigger than just selling artwork online. Some of
our history is going in whatever direction the wind takes us, as long as it
seems prudent. We decided to do this software because it's something that we
need for ourselves, so it's not like we're going on this wild tangent on a whim.
I know that other people in our situation will be able to use this as
"This is definitely unusual," says Kerwin. "I don't know
anyone else who is doing what we're doing here, especially with search
marketing. We try to hire our marketing staff and bring them up to speed with
this stuff, but you can't find anyone who has any experience with this, the
people just aren't around here. This business was meant to be in Silicon Valley,
but I just happened to be born in New Jersey so here we are."
Originally published in the U.S.
1 Newspaper, March 21, 2007.