We always challenge our clients. Would you rather have 500 visitors to your site and have five turn into customers, or would you rather have 100 visitors to your site and have 20 turn into customers? The answer is always the same. The latter. Because if people are coming to your site and not converting, why should you even care about them? (Let’s spare the remarketing or better content debate and just keep things simple for now). The whole point of your website is to bring in qualified visitors. Although some brands do care about impressions and site traffic, most businesses, especially in the B2B world, only care about leads. That’s what is keeping the heat on.
If you’ve ever worked with Google AdWords, or a similar pay-per-click (PPC) software, you know that it can get complex really fast. It requires a lot of attention, careful planning, and constant maintenance. And of course, there are all sorts of metrics that can help you understand your performance, measure your success, and help you make important decisions to improve your program.
There are a lot of things to look at and consider: total clicks, impressions, click thru rate (CTR), conversions, total spend, average cost per click (CPC), impression share, estimated cost for first page bid, estimated cost for first position bid, quality scores, and many more. Of course, the way you manage your program and the decisions you make all depend on one thing: your goal.
Last year, we took on a new client for pay-per-click services. From past experiences, and a past agency that they had worked with, they were familiar with seeing a handful of monthly metrics that helped them judge success. This included wanting to see positive growth for clicks, a healthy click thru rate (CTR), and a cost per click (CPC) around $1/click (it is not a competitive industry). We had a conservative budget of $900/mo, or around $30/day. At the time, they were not tracking cost per converted click, but we measured the number to establish a baseline for us. Here’s the metrics we were looking at for the previous six-month time period:
August 2013 – January 2014 (6 month time period)
- Clicks: 6,254
- Impressions: 7,190,658
- CTR 0.09%
- Average CPC: $0.75
- Total Spend: $4,694.54
- Conversions: 6
- Cost per conversion: $782.42
As you can see, the program was in rough shape. We were seeing high levels of clicks and impressions (what they were trying to use to measure success) but a horrific CTR of 0.09% and a very expensive cost per converted click at $782.42. They were spending far too much to get someone to complete a form on their site.
We spoke to the client to understand their goals, and when it came down to it, all they wanted was more leads. They gave us full control over the program and trusted our judgment to mix things up. We decided to revamp the program with a different measurement for success. Instead of focusing on impressions, traffic and click thru rates, we decided to focus just on cost per converted click. This stat would dictate all of our decisions. Because the less we spend on a lead, the more money there is available to get more leads. See?
From Google, conversions happen when someone clicks on your ad and then takes an action that you’ve defined as valuable to your business. This can include a phone call, a form submission, an email click, a transaction, and more. Conversions for us were defined as a form submission, i.e. a request for quote.Cost per converted click measures how much money it takes to achieve a conversion. The less money spent to achieve a click, the lower your average cost per converted click.
Ideally, you’d want to place a value on each of your conversions so you can track your conversion value cost. This allows you to see which clicks result in larger profits. For instance, if I sell a motorcycle with a margin of $2,000 and biker gloves with a margin of $20, naturally, leads related to the motorcycle are more valuable to me. Placing values on your conversions in Google Analytics/Google AdWords can help you manage programs this way, in a much more efficient manner. However, if you do not have as advanced of a system and have not labeled your conversions with values, then the next best approach is cost-per-converted click. This is the model we used.
Since we now recognized the goal and knew the metric we would be most closely looking at, we decided to dig in and make some drastic changes.
Here were some actions taken
- First, we paused all display ads. Display ads appear on third party websites in Google’s ad network rather than on search engine results pages. These can be text-only or image-based ads. These ads were bringing in a lot of visitors at a very low cost (bringing the average CPC down), but they weren’t seeing any conversions from this ad set. None. We opted to stop the program completely and allocate those dollars to the search ads where they would be more costly clicks but much better quality.
- Next, we went in to review the search AdGroups. For the AdGroups that were bringing in the most conversions, we allocated more budget dollars. For AdGroups that were not seeing many conversions, we took budget away.
- We then examined keywords for all campaigns. For keywords that were consuming budget but leading to no conversions (or very little conversions at expensive rates), we paused. For those that were seeing conversions at healthy cost per conversion levels, we either kept bids the same, or bid more competitively (if needed).
- Finally, we conducted more thorough keyword research to add more cost-effective keywords (cheaper long tail keywords or cheaper variations of keywords with modifiers), as well as add more negative keywords that would help eliminate irrelevant clicks that would not convert and drive up our cost per conversion levels.
We let the program do its thing and paid close attention, making minor tweaks and adjustments over time, as needed. The program ran for a year. I prefer to look at data year over year because it takes seasonality out of the equation. So let’s compare the original set to what we have seen over the past six months. These stats come as a direct result of the changes above.
August 2014 – January 2015 (6 month time period)
- Clicks: 5,288
- Impressions: 709,415
- CTR 0.75%
- Average CPC: $1.04
- Total Spend: $5,478.64
- Conversions: 28
- Cost per converted click: $195.67
|Label||August 2013-January 2014||August 2014-January 2015||Change|
|Average CPC||$0.75||$1.04||39% increase|
|Total Spend||$4,694.54||$5,478.64||17% increase|
|Cost per conversion||$782.42||$195.67||75% decrease|
We saw less traffic than the first time period and far less impressions. Though we weren’t focusing directly on this statistic, we did improve CTR by 0.66% because of the strategic changes. A slight increase in spend took place since we were bidding more competitively. But the biggest improvement (and most important to us) came from our total conversions (with a 367% increase) and a dramatic improvement to cost per converted click (a 75% decrease). This put us in a much better position to utilize our budget and get more bang for our buck (more leads for our budget).
Google AdWords now generates the majority of this client’s goal completions, even more than organic traffic, which traditionally had been their biggest driver. This tells us that the quality of visitors is far better. Paid traffic is now a major component of this client’s overall online marketing program and it is in much healthier shape now that we’ve weeded out the ads and keywords that were not delivering and changed the focus to be just on ads and keywords that are working. The changes greatly improved the program and the client was satisfied (and continues to be satisfied) with the results.