In Part I of this two part series, I left you saying how inaccurate the CTR’s Google was give are. Now I want you to totally forget what I said! In this post I’m going to take you step-by-step through extracting and changing poor performing descriptions.

Step I:

Start by going into webmaster tools and download your keywords file, don’t forget to set the date to the last week if you want accurate results. While we’re talking about dates, the larger the time period the more detailed your results (you loose a lot of long tail data by setting the time to less than a week as the number reverts to “<10”); however if you’re link building around specific keywords this will affect the CTR’s you’re getting (as you’ll be ranking higher) so try to get a balance.

Step II:

The list will generally be quite long, but don’t worry; we’ll soon cut it down. Extract all keywords with your brand name (including misspellings and derivations) and put them into a separate tab. Branded keywords are tricky as they are affected by outside factors, such as trust and brand presence (see Part 1’s diagram explaining click factors) as well as being a specific type of search. If you have even an averagely strong brand you would expect a stronger CTR on those terms.

Step III:

The next step is to get rid of any non ranking or long tail results. It’s very difficult to see how well long tail pages are performing, and even harder to try and create an enticing description copy for all searchers. Don’t worry about loosing the non ranking pages, focus on link building/SEO to improve rankings so you have a decent data set to work from. Filter results that are <10, and filter any rankings above 10.

Step IV:

Now we need to make an array to see what our expected click throughs are, create a horizontal table with a column from 1-10 with 0.1 steps in between, mirroring Google’s Avg. click through scale. We have a rough idea of what CTR’s people should be getting  from SERP data a while ago, yes I know it’s old and yes the SERPS’s will distort the results but this is about finding results quickly, here are the estimations :

Ranking Number 1 receives 42.1 percent of click throughs.
Ranking Number 2 receives 11.9 percent of click throughs.
Ranking Number 3 receives 8.5 percent of click throughs.
Ranking Number 4 receives 6.1 percent of click throughs.
Ranking Number 5 receives 4.9 percent of click throughs.
Ranking Number 6 receives 4.1 percent of click throughs.
Ranking Number 7 receives 3.4 percent of click throughs.
Ranking Number 8 receives 3.0 percent of click throughs.
Ranking Number 9 receives 2.8 percent of click throughs.
Ranking Number 10 receives 3.0 percent of click throughs.


Add these to the whole numbers and work out the rough intermediate values. Now Vlookup the CTR we’re expected based on your average ranking position. Add a new column on the original table which works out the difference in click between the estimated and actual value, remove any results where the actual CTR is higher and sort by the differential column. You now have your worst performing keywords, here’s an example of what you should have:

Meta Description analysis

Repeat the process for those branded terms. I usually take the top 20 in a month and look at them closer. Let’s take the worst performing keyword from the above example and look closer; go back into webmaster tools and click on the poor performing keyword for a full break down of performance:

Detailed keyword analysis

Step V

This keyword clearly has issues, even when we’ve had a placement on the #1 ranking it’s not achieving at all. Looking at the pages, thee pages have had the lions share of impressions, and are all worth investigating further. The next step would be to look at the SERP for this keyword, which as it happens turned out to be filled with Verticals possibly explain the poor result, and investigate the 3 poor performing pages (one of which had no description).

What makes a great description?

So we’ve identified the poor performers, what makes a great description?

The factors I listed in the previous post are going to affect the CTR but the two most easily changeable ones are related to the content within the description:

Description CTR factors

There are two sides to a good description, getting copy that is relevant to the searchers and text that pulls them in.

Be Relevant to the Searcher

Relevancy itself consists of several different factors; firstly you want to include the keyword the searcher is using. This may seem obvious and simple, buy you have to remember you’ll probably be ranking for 100’s of different keywords for just one page, try to pick up recurring keywords and add them in. The data you have mined above should point to descriptions that are missing certain keywords.

Keyword placement is also important; people probably won’t read the whole description but just scan it. PPC experts recommend an “arrowhead” or “zig-zag” formation for top priority keywords as they will be seem more, use description preview tools to test before you update.

Adding synonyms of keywords will help make your description read better without loosing the relevancy that you want. Finally don’t forget about your long tail content, go into Google Analytics and see what keywords your landing page is currently catching, are there underlying themes you can play to?

An example is a camcorder page I was working on, it was getting lots of hits from keywords with colors in them. I simply added a line about our large selection of different colors such as black, white and blue and we saw a nice CTR and traffic increase.

Pull Them in

Think about the Pull of your description, if you were a searcher would you click on your own add? See what competitors are doing as well as the PPC ads, what kind of language are they going for and try to get some ideas.

Trying to show the value of your page is key, what’s the key selling point about that page that’s going to make people interested, is it a great price point or sale (adding in £ and %’s off is a nice way of attracting attention, but beware to update descriptions if prices change), maybe it’s because you regularly update your blog or maybe because your trialling your software for free.

People like to feel secure on the web, especially older audiences if you’re not a well known brand you may have issues with people not trusting you. Tell them how safe they are with you, that you offer full protection your orders, that people can email you with questions whenever they want.

Add a sense of urgency to encourage people to click. This is a classic sales tactic but one which should used carefully, only do this when it’s valid. It’s not worth being dishonest to your customers for a measly 1% click through increase.

Keep Testing

One thing to bear in mind is that (funnily enough) people are all searching for something different, weight your relevancy/pull to suit the area your talking to. If you’re a blogger it’s all about relevancy, people want information, is your blog relevant to them? e-Commerce is all about pull, why you should you shop with me over my competitor.

Most of all it’s about tweaking, remember just because your achieving the average click through doesn’t mean you can’t do better. Keep testing on your high impression keywords and see the traffic come in.