Ah, 2003. Back then, the RIAA was suing music sharers, Arnold Schwarzenegger was elected Governor of California, Yahoo! was still a real competitor to Google, and web page indexes with funny names like HotBot and Altavista were flourishing.
And I was taking my first steps into the world of PPC.
At the time, paid placement in search results was just a few years old. Still, I had landed my first major client, a video editing school in Los Angeles that was looking for an increase in prospective students. Although there weren’t a lot of resources available to teach me the ropes of PPC, I took what I had learned about web usability and dove in.
I spent days researching Jakob Nielsen’s usability studies, and redesigned my client’s website to include things like a legible font size, easy forms and clear calls to action. We saw a 30% increase in form submissions in the week after the redesign. With that momentum, I was able to convince my client’s executive team to approve some budget for paid search-first with Overture and, later that year, with AdWords.
Needless to say, my expertise (and the PPC landscape) have both evolved significantly since the dinosaur days of 2003. Here’s my take on how Overture, AdWords and analytics have changed over the last 15 years.
Search engine marketing was a relatively new phenomenon in 2003, which means that best practices hadn’t yet been spelled out. On the one hand, it was a fun time because most everything was experimental. But on the other, getting reliable results in this environment created a lot of pressure. One major platform, Overture (previously known as GoTo), was attractive because it served ads in Yahoo! and MSN (now Bing). Overture also featured a bidding system that let all advertisers see the Top 4 bids. The cost per click was $0.01 above the bid under you.
In those days, you could intentionally instigate a bidding war by taunting the competition into bidding very high for a given keyword. After a while, you could “let them win,” and bid $0.01 below their bid-an action that would charge the maximum click amount to your competitor. For a couple years, it was a solid strategy to run out your competition’s budget at a very high CPC, and then adjust downward. Along with this, the top bidder always got the top spot, regardless of relevance.
So in the beginning, the paid search world was full of roguish techniques and spammy links, and success meant using all of this to your best advantage.
The Overture system changed significantly by 2005, when Yahoo! purchased and rebranded it. The company made the bidding system “blind,” meaning that you could no longer see your competitor’s bids. This eliminated the type of strategies that worked well before. In that same year, Google rolled out its Quality Score, which uses algorithms to give advantage to ads that were especially relevant to the given keyword. Gone were the days of the top bidder being guaranteed the top spot.
The other major platform at the time, Google AdWords, was originally pretty simple. I remember the interface being remarkably more intuitive than Overture, though I was bummed that we couldn’t see our competitors’ bids!
At first, the AdWords network was available on around 100,000 websites, which seemed like a large network at the time. In AdWords’ early days, there was no checkbox for excluding parked domains or other spammy sites. It also lacked a good system for discovering click fraud. Making sure that your client’s ads displayed on quality and relevant sites required constant attention-as in, sifting through long lists of URLs and disallowing them individually. In those days, I was able to recoup around 10% of my client’s budget each month simply by investigating abnormal traffic and finding fraudulent clicks.
Of course, there was no Google Analytics yet. That came later. Back then, we used a software package called Urchin, which had to be installed on the server, and was very slow and clunky compared to what’s available now. Urchin was a great product for its time, but was still rudimentary enough that I found myself combing through server log files by hand to discover that fraudulent traffic I mentioned above.
And installing it was such a pain! In those days, paid search experts had to know at least the basic Linux commands to get analytical tracking up and running for our clients. Let’s just say this wasn’t my jam. I was always afraid I’d accidentally “rm -rf /”-as in, delete everything in the root directory. It seems funny now, but command line process was always “white knuckle” work for me. It’s better left to server admins, but in the early 2000s paid search pioneers had to wear that hat, too.
Google eventually purchased Urchin and turned it into Analytics, which improved things a lot. Suddenly, we could install it with a simple copy/paste in the site’s html. This seemed like magic compared to the labor-intensive drudgery we went through before. And because the software was housed on Google servers, we didn’t have to worry about our client’s website slowing down when we ran reports, or having the historical traffic data fill up the server’s hard drive. (That was really a thing, even on pretty decent dedicated servers.)
Paid search platforms and tracking software are so much more powerful in 2018 than they were in the dinosaur days of the early 2000s. In a sense, PPC management is easier because there have been so many improvements and efficiencies added over time. On the other hand, a huge array of new features have been added that allow for very fine tuned and effective management – if you know how to use them! Additions like YouTube advertising, Shopping campaigns, Sitelinks and Audience Targeting make me glad I have those years of experience to draw from, and our clients agree. New PPC managers find a very complex landscape compared to the old days, but at least they aren’t fumbling through Linux to set up bounce rate tracking.
As a PPC veteran who lived through these awkward early stages, I can say unequivocally to today’s paid ad experts, “You kids don’t know how good you have it these days.”
Now, get off my lawn!