August 3, 2012
As sad as it may be to whomever Google’s version of Thomas Jefferson is, all leads are not created equal. Account-wide CPL goals are a great way to measure overall performance, but true profitability for lead generation relies on back-end data. Only through detailed analysis over an entire sales cycle can you really get a good grip on PPC’s effects. Lead generation doesn’t have the immediate satisfaction of seeing revenue numbers tick upward in Google Analytics, but if your client is using a program to attribute sales to their original source, you can streamline your PPC efforts to increase both total revenue and Return on Ad Spend.
As long as you have a pretty solid grasp on your sales cycle, you can attribute leads all the way back to the keyword that started a prospect’s long, beautiful journey from internet browser to beloved client. Let’s take a look at one scenario to better understand just what paid search is doing for them.
Our client has specific leads goals on a month-by-month basis. They rely on six different marketing streams to generate those leads, and PPC has a lion’s share of that (over 60% of all leads are supposed to come directly from our efforts). I know that a good month for PPC almost assuredly makes for a good month for the marketing efforts of this company.
But what really makes a good month for PPC? Identifying which specific search engines, networks and campaigns generate actual, tangible revenue for our client is essential to properly allocate our resources.
Let’s start with Bing (which might be one of the first times in Internet history that someone started with Bing. You’re welcome, Bill Gates.). Our client has a rocky history with the engine. In November of 2011, we heard from the client’s sales team that bad leads were coming into the system. Real names, real phone numbers, only the names and the phone numbers didn’t match up and the people at the end of the other line hadn’t put the info into our client’s website. After some investigation, we had them opted out of search partners in Bing and that stopped the bad leads coming in.
Although traffic has been down since opting out of search partners in Bing, leads are still coming in. And the great news is that Bing has a tremendous ROAS for our client. Their sales team is happy with the quality of leads, our marketing contacts are happy with the consistent (albeit low volume) flow of prospects, and the customers at the other end are happy because they’re finding the type of product that they wanted to find when they first entered a query into Bing.
Although Bing is the more profitable venue for this client, the lower market share compared to Google means that there just isn’t enough traffic there to sustain the lead generation quota for PPC. Thankfully, Google also has a very positive ROAS (it’s about half of Bing’s, which is still totally awesome).
Why such a disparity in ROAS from engine to engine? The Google Display Network. And that seems like an indictment, but I assure you that it isn’t. I personally love the Google Display Network and fully recognize that without it we wouldn’t be the primary lead generator for this client. Of our Google spend, 74.5% comes from the Display Network. It spends a lot of money, but there is absolutely revenue on the back end. The ROAS is less than half of what it is in Google search, but that is revenue that we wouldn’t be able to get otherwise. And the ROAS is positive enough to ensure profits for the client.
Over 50% of the total revenue generated from PPC over the last year came from the Google Display Network. The actual closed sales cost more to get than they would have in regular search, but those leads couldn’t have been found by traditional keyword targeting. Knowing that search volume is going to be limited at some point is key to accepting the lower ROAS in Display. It all comes down to what tolerance levels can be allowed for profitability.
Evaluating the Display Network on a campaign-by-campaign basis also illuminates what marketing-types mean when they say “fill the funnel.” Some of the campaigns are positive, but only slightly. Even overlooking the fact that these campaigns are relatively immature and may not have the full sales cycle’s worth of history behind them, these campaigns still prove profitable on the back-end. Because we use remarketing so extensively, any clicks in these campaigns serve as an introduction to our brand that remarketing can later capitalize on. And our Remarketing campaign actually has a higher ROAS than our search campaigns.
Now that we have a steady baseline to base performance off of we have been able to calibrate our targeting accordingly. Here are the general results we’ve found, using Google search as the baseline ROAS. These are just for one account, so they’re just for demonstration purposes. But in your own accounts it’s worth running this analysis to see where you should put your money.
I enjoy looking at adjusted ROAS using a steady baseline for comparison. In the below chart Google Non-Branded Search is considered the average and everything else adjusts up or down from there.
Branded search is the thing a PPC’er has the least control over, so identifying areas where you can wring even more profitable leads out is key to growing your accounts wisely. (And it should be noted that Bing’s branded numbers are off the charts. It’s a pretty small sample size, so that figure most likely isn’t sustainable.) Even though the Display Network has the lowest adjusted ROAS, it also has the most spending potential by far. And the more we spend there, the bigger audience we have for remarketing which just continues to grow ROAS.
How do these figures stack up with what you’ve seen in your own accounts? Has the Display Network had a positive ROAS for you?