Lots of crazy stuff has happened in PPC during 2012.  Even though that’s been true pretty much every year since its launch, we at PPC Hero felt that it was a big enough year to justify its very own series.  That’s why for this month’s series we’re taking on the Biggest Thing in My PPC Year.  Whether it’s tools, product releases or changes to our way of thinking, we’ll be covering the biggest thing that’s changed our work lives/brains this year.

I’m still fairly new to PPC, so when others talk about the biggest thing that’s impacted their year, my answer might be a little overly broad:

“Well this year, I learned PPC.”

So when others mention their intense love of Pivot Tables, the differences between In-House and Agency PPC, Ad Rotation Settings, Flexible Reach, and Product Listing Ads, I would have to say “yes” to pretty much all of that.

However, when I was starting out this year, wrapping my head around the proper mindset of managing a PPC account and the mounds of data involved took more time and effort than anything.

So here are the top three tips I can give you to help those of you still new to PPC understand your account history.

Free Your Mind.

1) Lesson #1:  Don’t Panic on the Day-to-Day Results.

When I initially started, I was intensely focused on winning every day in the PPC battle.  It’s certainly achievable depending on how many accounts you manage, but as you grow in account management and your accounts expand, the number of moving parts in your PPC account grows in complexity.  In such a system, you’re bound to have good and bad days.  My first piece of advice to you is this:

One day of poor performance shouldn’t send you in to a panic.  Conversely, one stellar day shouldn’t be enough for you to believe all of your problems have been solved.  There are data points to learn from each:  did some variable change to bring about this performance, such as average position, or max. CPC?  Learn from the good and bad days, but leave the stress for performance trends, rather than performance outliers.

This brings me to my second point:

2) Lesson #2:  Take the Long View.

It took me a lot of time to learn that in PPC, your performance is best measured (depending on account volume) over the span of weeks, months, or even years.  As I said, a single bad day shouldn’t be enough to stress about poor performance, and a singularly great day shouldn’t cause you to sing from the rooftops.  It’s about trends, and measuring the statistical significance of your daily performance versus the historical lifetime of the account.

Oftentimes, taking the long view of an account’s history, coupled with the changes and initiatives launched over the course of that time frame can help you understand the “why” behind your account’s performance.

This leads to my final point:

3) Lesson #3:  Look for Inflection Points and Statistically Significant Changes. 

This may be best explained through a small case study.

I was recently tasked with diving in to an account to determine why overall profits were down for this particular client, as their only concern is the bottom line.  On initial examination of the account, it’s made of three distinct campaign types:

  • Search
  • Display and Managed Placements
  • Display Campaign Optimizer

Now, overall conversion volume has remained steady, as has CPA.  But why has their overall profit fallen?  Here’s a screenshot of the account starting from December 2011 until August 2012:

From December 2011 – August 2012.

For reference, the dips in performance are due to the account being paused while system issues were worked out – they’re irrelevant to this discussion.

However, you’ll note the first jump in late January/early February, followed by a peak in April, then a slow decline through the end of August.  When you’re diagnosing overall account performance, these are the areas to look at.  In this particular case, an examination of the campaign change history tells us that the DCO campaigns were launched on this date – leading to the jump in conversions.

So that’s the first data point to take away here.  What’s the other?  Here are some screenshots of their Search Campaigns over the same time frame:

Campaign 1:

From December 2011 – August 2012. Note the sharp drop in early February.

This country-based search campaign sees a significant drop off in traffic in late January/early February.  This is due to several general ad groups, both broad and exact match variants, being paused.

Campaign 2:

From December 2011 – August 2012.

Conversely, this campaign (a worldwide campaign) sees a corresponding jump in conversions on the same date, followed by a precipitous drop in April upon reactivation.  This is due to the exclusion of several countries – including those from Campaign 1.  So the Search volume from the Broad terms excluded in Campaign 1 were then being funneled to the World Search campaign.  Once Campaign 1’s country was finally excluded, that traffic disappeared.

Campaign 3:

From December 2011 – August 2012.

Campaign 3 was another geotarget campaign created in concert with the above geographic exclusions.  They split up several countries in their World campaign according to best practices.  However, upon reactivation in May, their conversion traffic takes a nose dive.

Essentially, what we would discover in this case is that by pausing the broader ad groups, they had severely harmed their Search traffic across all Search campaigns.  This led us to examine the overall value of leads from each Campaign type, where it was discovered that Search leads were over five times more likely to convert from lead to sale compared to the DCO and Content campaigns.

That’s a long-winded way of saying “don’t sweat the small stuff”, “take the long view”, and “look for big swings in performance to guide your analysis”, but hopefully those three tips will go a long way for you burgeoning PPC managers out there come to terms with the day-to-day realities of account management.  Maybe you’ll manage to keep more of your hair than I did, too.

What tips do you have for understanding your account?  How do you look at your data?  Let us know in the comments – and as always, thanks for reading!