1. Look beyond the conversion
It's not all about conversions. I often use the 200 clicks rule as a general rule of thumb to decide whether a keyword, ad, or search query is performing, with the theory that if no conversions have come from a keyword, ad, or search query after 200 clicks, then you can be reasonably confident that that keyword, ad, or search query might be under-performing, and is worth investigating.
However, focusing entirely on conversions ignores other important metrics of value. What about visitor engagement? What about views of the contact page? What about repeat visits? PPC marketing spoils us with conversion tracking and ecommerce tracking, and it can be easy to get stuck in a mindset of considering just one metric (conversions) and ignoring everything else.
However, if you get stuck in this mindset of simply trying to maximize conversions, you will lose sight of the bigger picture, and mis-align your campaigns with what is truly creating value. If all marketing was based on trackable conversions, all TV commercials would be paused. Newspapers and magazines would go out of business. There wouldn't be a superbowl.
This 'conversion tunnel vision' created by the invention of conversion tracking is probably one reason why Google decided to not allow advertisers to track the direct response of their YouTube video campaigns in Google Analytics, by stripping all utm_ tracking parameters from YouTube referals. Although annoying, it makes sense " if we focus only on conversions (as we have got extremely used to with Google AdWords), then we will fail to see the bigger picture. We will run a YouTube video campaign, realise the ROI from website clicks was poor, and pause the video campaign, ignoring all other value that may have been created.
Look beyond the conversion when carrying out analysis of your Google AdWords campaigns. Create additional goals to measure softer levels of visitor engagement, such as views to the contact page, views of multiple key pages, and high browsing times and page view counts. Avinash Kaushik (the analytics evangelist for Google) suggests looking at how many people have made at least 5 visits to your website in the last 7 days as a powerful and reliable measure of visitor engagement. Looking beyond the click and conversion can be extremely insightful, and can help to more closely align your AdWords campaign optimisations to maximize overall value.
2. Consider conversion rate by ad position
Not all ad positions were created equal. They may have been, back in August 2009, if we are to believe Google's research suggesting that conversion rates are very similar regardless of whether your ad is showing in 1st or 10th position on Google.
However, since the advent of ad extensions, and more specifically sitelink ad extensions, ads which are displayed in the top positions on Google now have the ability to significantly stand out from the competition and have an additional 100 characters in which to impress potential customers. If adding sitelink extensions better communicates the benefits of your business to searchers, and better engages visitors, conversion rates may therefore be higher with ads which are shown with sitelink extensions.
What this means is that a click from the search query 'melbourne luxury hotels' in 1st position (with sitelink extensions) is not necessarily the same as a click from the same search query 'melbourne luxury hotels' in 4th position (without sitelink extensions). It's no longer a case of trying to maximize clicks for a particular keyword " sometimes fewer clicks from that keyword may be more beneficial, if those clicks came from ads which also displayed sitelink extensions.
To figure out whether an AdWords strategy which places greater emphasis on sitelink visibility (at the expense of click volume) is more beneficial for your campaigns than a strategy which aims to maximize clicks from that keyword (at the expense of showing sitelink ad extensions), find out which of 4 scenarios your campaigns belong to how to find your optimum CPC bid.
3. Consider the compound effect of CTR
Low CTR is a fantastic insight into something much larger. A keyword which has received 1,000 impressions and only 1 click (CTR = 0.1%), might look rather innocent at first. But if that keyword was shown in a high average position on Google, perhaps there is a reason why 999 people out of 1,000 chose to ignore your ad?
Filter your keywords to highlight those which have received 1) a high number of impressions, 2) a high average position, and 3) a CTR of less than 0.3%. Next, look at the search queries being matched to those keywords. Even though you may only have a handful of search queries from your 1 or 2 clicks that keyword received, think about all the searches which didn't receive clicks. Adding a simple negative keyword may only prevent a handful of irrelevant clicks, but that negative keyword may translate into preventing 1,000s of unwanted impressions.
Considering this compound effect of low CTR keywords can be very powerful. Adding negative keywords to filter out a handful of irrelevant clicks from an irrelevant search may only have a marginal effect on click volume, but will have a huge effect on filtering out unwanted impressions and increasing your overall CTR.
Very low CTR is often a sign that something is wrong. Since we can't see the search queries which don't receive clicks, use low CTR as your window of insight to highlight potential keywords which may be matching to irrelevant themes you may have previously overlooked.
4. Remember that past is no guarantee of the future
Unless you are deliberately asking for it, lightning is unlikely to strike you twice. Similarly, an obscure 10 word misspelt long-tail search query such as 'looking for chaep kenell for big dog who likes sunshine' which generated a conversion is quite unlikely to convert again from a different person. Just because a keyword performed well in the past is not necessarily a guarantee that it will perform well again in the future.
Of course, looking at which of your keywords, ads, and search themes perform better than others, and optimising campaigns accordingly, can provide huge benefits, and has become the foundation of PPC optimization since it's invention. However, what's important when carrying out your PPC analysis and optimisations is to remember significance. How big is your data set? How many conversions are you basing your decision on? If you are pulling back bids on the keyword 'cheap hotels with spa melbourne' due to receiving no conversions, and pushing your keyword 'cheap hotels melbourne with spa' due to receiving one conversion, they you might want to think again. Are those two keywords actually that different? Is there really a reason why such similar keywords would convert very differently in the long run?
As you carry out keyword optimisations, be careful when making changes to keywords which received only one conversion. One conversion is usually not significant, so try to move up a level (ad group / campaign) to see if you can draw more reliable insights. Search query theme analysis can also be incredibly powerful to compare themes within your search queries, and highlight macro insights from your micro search query data. Instead of comparing the keyword 'cheap hotels with spa melbourne' to the keyword 'cheap hotels melbourne with spa', try comparing searches which contained the word 'spa' with searches which contained the word 'balcony'. No more data is needed, but your findings are suddenly considerably more reliable and powerful.