NHS Choices recently published their 2010 annual report which includes summary data on their web site performance and quantifies benefits that resulted. ( £44M saved in visits to GP`s they claim )

The report covers three areas: Web Site Data, Market Share data and an off line user survey. An example of  the Web Site Data for September is shown ( these stats are published monthly) . The site, which uses tracking from Webtrends rather than Google Analytics, is grouped into two main areas Service Usage and Engagement With The Site. The first covers standard stuff such as Visits, Visitors, Average Time on Site and Pages Visited. More interestingly there is data on the top search terms and referring sites plus market share data. The Market Share data comes from an independent source – Hitwise. They make a good attempt to quantify Engagement, a key metric for information sites such as this. The engagement metrics they choose are New User Accounts and the number of Comments posted. All in all it looks like a decent effort to report on the effect of the site. Some information on costs would be nice.
As they are spending public money it is right that they publish data such as this. Should all charities who are receiving the public`s money publish similar data ?  Summary Press release here and the full report here

Thanks to those who came along  to see this presentation at the CharityComms event in London in Nov 2010. These are the slides I used -  with a few small amendments. You can access two other presentations that I have made on Introduction to charity Search Engine Marketing and Introduction to Google Grant AdWords.

There is a more complete explanation of how to identify poor performing pages ( slide 6)at fundraising.co.uk bit.ly/d9YvWN.

Demographic information about users of a specific web site can be difficult to find without expensive data analysis and research. But thankfully our friends at Google have provided a simple and easy way to find this information for a site. The site can be your own or that of a competitor. To access the information simply go to DoubleClick AdPlanner, as it is now called. On the Research tab you will see  information displayed for the default site wikipedia.org. Now enter the url of the site you are interested in and voila some interesting and useful information.

IMAGE PART 1 (Click for larger version)

Note the default  setting in the View Data for: drop down menu is United States. For sites that primarily serve UK markets select United Kingdom. As you can see in addition to data on Gender, Age, Education and Income Traffic Stats for the site are also provided. The section at the top marked Content Categories is potentially useful as it tells you how Google classifies the site. Note the site can be in multiple categories. This information can be used with Google Insight. For example you could check what have been the top search queries for all sites in a specific Category. It can also be used for better Benchmarking in Google Analytics. Simply go to Visitors > Benchmarking > in GA. Select Open Category List and identify the Category you want to be compared to.

AdPlanner provides further information on the same tab such as Sites Also Visited, Audience Interests and Keywords Searched for. Combining all this information will give a detailed picture of who uses this site and their interests.

IMAGE PART 2 (Click for larger version)

So how good is the demographic data and where does it come from. Well this is what Google says about it: DoubleClick Ad Planner demographics are generated through demographic inference algorithms that combine third-party demographic data with Google sample data. The third-party demographic data is licensed from an industry-accepted consumer research panel operated according to industry best practices by a full-service research firm. The firm employs a variety of panel recruiting techniques to ensure the quality of their panel. Given that AdPlanner is a vital part of google`s business I reckon the data should be pretty good. Google would have too much to lose if customers, many large companies spending big with Google found that the data was not at least reasonably accurate.  One thing to note is that I have found that data may not be available for smaller sites.

Twitter and Facebook have almost doubled the amount of traffic they sent to 6 UK charity web sites over the last 6 months. Visits identified as coming from all social media sources, have grown faster than the overall growth in web site visits while Facebook and Twitter are sending a greater share of the overall social media traffic. However for the sites that I reviewed for this survey social media is not a big source of visitor traffic. In the last 6 months approximately 1% of their total web traffic has come via social media compared to .5% in the previous 6 months.

Facebook and Twitter are most popular with UK charity visitors.

The share of traffic from social media sources varies among the surveyed sites from a (very) high 6% to less than .5%. Twitter and Facebook are the largest contributors to social media visits and are the fastest growing of the social media sources. Comparing the last 6 months with the previous 6 months Facebook sent 65% more Visitors while Twitter sent 80%. If you are comparing the absolute number of visits remember that Twitter is reckoned to be under counted in Google Analytics by a factor of 4 or 5.

This analysis is similar to the social media analysis I did back in March 2010. This time I included Visits from over 34 social social media sources and ensured that the analysis was comparable across all 6 sites for both periods. Three of the sites had visitors in the 10/20k per month range with three more having over 100k visits per month.

The data comes from Google Analytics using the Advanced Segments capability. You can do this test on your own web site by using Google Analytics Advanced Segments. You can easily create you own version of the Advanced Segment or drop me a line and I will send you the one I used with its specified social media sites. Once you receive it simply log into your GA account and paste the url into your browser. Voila !  you can check how your site compares to the ones I did.

Click for Larger Image

You can track how many visitors a page on your site receives as  a result of your Tweets by Tagging any Links that you put into your Tweets. Bit.ly, currently the most popular URL shortner, can be used with Google link tagging. Simply add tagging parameters such as Source and Medium to the end of the url, shorten in Bit.ly and post to your Tweet. Now when you look in Google Analytics > Traffic Sources you will see a separate Medium ` Twitter` and provided you have tagged your links they will appear listed for each Tweet you have sent out. Now you know how many visits your links have brought to your site.

Incidentally if your site uses the AddtoAny bookmarking service ( as this site does) then you can see full details of what pages have been bookmarked by looking in GA >Content > Event Tracking.

Combining the use of Advanced Segments, Link Tagging and Event Tracking will give you a good idea of how your Facebook and Twitter activity are affecting your site.

Adding negative Keywords can really help your performance in AdWords. Doing a search for `university places` on google.co.uk as expected throws up some ads for Uni places. However my eye was caught by the one advertiser, Google, who is promoting a different kind of `places` in this case an ad for Google Local Listing. Clearly someone has put a keyword(s) on Broad Match and has not been regularly checking the search terms that it actually triggers on. For occasional users of AdWords the idea and practice of using negatives can be difficult to grasp. This is a pretty good example where the use of a negative keyword ie `university`would dramatically lower the number of unwanted impressions resulting in a better CTR and probably lower CPC.  Mind you this is likely not a real cost for whoever placed this ad.

Google Places ad showing in the wrong place CLICK FOR LARGER IMAGE

Recently the UK government announced that over 1000 government web sites have been closed and another 615 are scheduled to close. A lot of these sites had a close relationship with the charity sector and no doubt had backlinks to various charity sector sites. It is unlikely that any of the closed sites sites have any permanent redirects (301`s) in place. This means that anyone who visits them will get a `Site Not Found ` message. At best visitors will be redirected to another government web site. Either way all the referral traffic (see Google Analytics > Traffic Sources > Referring Sites ) and the google ranking that they passed to your site via these links will be lost. This will have a real impact on traffic to many charity sites as Google gives great weight to .gov links when deciding who to rank for a particular search query.

What to do ?  Go to your Webmaster Tools account and look in Your site on the web > Links to your site. Download these links to Excel and find the ones with .gov extensions. Then very politely contact the organisation and ask them to put a 301 redirect on the pages that you appeared on. Then find the new government site ( or indeed any government site) that deals with your charity or condition and ask them for a backlink to you site.

UK Government COI report

Charities and nfp`s looking for benchmark data on their web site performance will find the the recently released report on 46 government sites useful. The data is for 2009/10 and covers 46 sites with total monthly visits of 32 Million and a total cost of £127M. For each site there is a breakdown of Cost and Usage. Across the 46 sites the overall cost per visit is £ 0.22 with a Bounce Rate of 37.5%. The Average Time on Site is around 4/6 minutes (although this excludes all Bounce visits) and average monthly Unique visits per site of 690k each month. As with all data real care needs to be taken in any comparison. To get a more accurate picture I picked 10 sites where the costs are detailed and the sites have a mid range value of monthly visitors. For these 10 sites the average cost per visit was £0.07 over a total of 52M annual visits. This data is particularly useful as it is from government web sites. These more closely resemble the function and purpose of charity sites rather than commercial sites where most data comes from.

Picking one or two sites from the list that have Costs or Visits similar to yours and comparing their data to your site could be useful. For the most accurate comparison charity sites should I reckon omit data relating directly to Fundraising traffic as this differs greatly from cause related traffic and is not comparable to the data from these government sites.

The methodology used to gather the data is clearly laid out and could be a useful basis for doing an evaluation of your site. In addition there is a lot of good advice in the COI site section Measuring Web Site Quality. The section Delivering Web site objectives is useful in pointing the way towards more detailed measurable achievements giving examples such as 1. Reduced calls due to better site info 2. registrations for events and services 3.  measurable user participation, etc. Mind you the COI site itself could do with seo friendly urls. Better advice about getting your web site seen by search engines would be useful too.

Demonstrating that a web site is worth the investment is not easy but hard data is the best way to do it. Showing that your site compares well using this data and that it compares well to its peers using the info provided by Google in GA will help. Proving that you can reach more people AND lower costs elsewhere are pretty powerful arguments as budgets come under scrutiny.

Google Insights for Search is useful for identifying new search trends and potential keywords for charity adwords or seo.  It is particularly useful in finding new or rising search trends. AnnMarie Hill of the Google Grants Team has recently written about it. A couple of examples show how it can be used.

Comparing two search terms. CLICK FOR LARGER IMAGE

First, for the United Kingdom only we can compare how frequently in Web Search the terms ` volunteer` and `volunteering` were used over the Last 12 months. We can quickly see that the search term `volunteer` is almost twice as popular as `volunteering` and that January is the most popular time for searches. More data about the nature of these searches can be got by examining the searches in each Category ( shown in the red box). Note that data from the filter `All sub-regions` for the UK has very questionable accuracy and should not be used without data from other sources.

We can now look in more detail and see what the Top Searches associated with `volunteer` were.

Google Insight rising searches CLICK FOR LARGER IMAGE

Rising Searches (shown on the right of the image above)  is particularly interesting as it shows that searches for `vodafone volunteer` is classed as Breakout which means that searches have risen by more than 5000% ! `festival volunteer ` and ` volunteer week` have also been increasing in popularity over the last 90 days.

Exploring All Categories can be useful. For charities and nfp`s in particular the Categories: Lifestyles, Society and Health give interesting insight into what searchers are looking for.

Whether you have a Grant or regular AdWords account Google has provided a clear way to improve the results you get from your AdWords spend. Quality Score (QS) has been specifically designed by Google to encourage ads more relevant to the searcher and potentially cheaper for the advertiser. Improve your QS and get more clicks for less money.

possible Quality Scores ?

The cost per click (cpc) that the advertiser pays for a keyword is directly related to its QS. Craig Danuloff has crunched a lot of numbers to show the effect of QS on the cpc. Briefly, a QS of 7 means that you are paying the standard price for each click. For a QS below 7 you are penalised eg a QS of 4 means you will be charged a 75% Premium. However a Quality Score of 1.0 means that you get a discount of 30%. So what are the factors that Google takes into account when calculating the Quality Score for each Keyword/Ad combination. For the Search Network Google lists 8 factors. However the most significant is the Click Through Rate or CTR – specifically the CTR for the Exact Match version of the keyword.  Siddharth Shah has crunched some different numbers and his show the correlation between CTR and Quality Score. The higher the CTR the higher the QS. So, a keyword CTR in the range of 2 or 3% will get you a QS of 7 but you will need a *much* higher CTR to get the magic of QS 8, 9 or 10 discounts. Looking at my own keywords with a QS of 10 ( not as many as I would like sadly )  all of them have an Exact Match CTR of over 30%. So no surprise, Google does not easily give away AdWords discounts. Typically you will only get that kind of CTR on very specialise keywords or Brand Terms.

One way to improve your CTR for a keyword is to lower the number of Impressions that do not receive a click. Identifying possible Negative keywords that will help you is an ongoing task that pays dividends. One way to find search terms that on Broad Match trigger your ad is to look in the Search Query Report. For some accounts this now shows Impressions that get no Clicks. Another way to is to enter the keyword as a search term and then check the terms shown in Google Suggest. Also some terms shown in Related Searches under More Search tools can give suitable negatives.

Note that Google makes public a linear QS scale of 1 to 10. However it is almost certain than internally they use a log scale such as the one used for Page Rank where one number is vastly greater than the previous one eg 2 is 100 times greater than 1 and so on. This means QS will in reality cover a very wide range. So when you reach the magic 7 keep improving the CTR as it is very likely that Google has a more refined internal scale and even though you cannot get above 7 your ad position and cpc may will continue to improve.

Google have again updated the search query data that they are providing in Webmaster Tools. The data can be found in WMT under the tab `Your Site on the Web`> `Top Search Queries`. ( The posting below titled `Webmaster Tools data for Marketing` gives more info on WMT layout) . This update shows many more Search Queries ( 3/7,000 in some of my accounts) with a Google Analytics style graph, a Search Box to find a specific Query term plus date menus to isolate data for a specific period. eg  How did we rank do for a specific keyword/phrase for that race/event/cause last week or month ? However the most interesting thing, which is only briefly mentioned in the Google blog posting, is the `% Clickthrough` column. Some bloggers have mentioned that Google is looking to add more weight in the search results algorithm to those pages that are popular with users. This would be consistent with the approach used in Google AdWords where a discount in applied to the cost of each Click for those keywords that have a high Quality Score and hence CTR. ( It is worth noting that the Impression and Click data shown in WMT do not match the data you will find in your Google Analytics account. My analysis shows that the number of clicks shown in GA is consistently higher than the WMT by around 20 %. )

So what to do: 1. Look at the list of Queries and see if there are any that you would like a page on your site to rank higher for.  If you see a suitable Query with a high number of Impressions take a closer look. If your highest ranked page for that term is below Position 1 or 2,  and consequently getting few Clicks, you should look to optimise that page for this Query. To do that look at on-page factors, particularly the Page Title, then look to get back links, ideally from external sites, with the Query as the anchor text.

2.  Check the snippet which appears in the Google SERPS results for the top terms that you want to be found for.  Does the snippet  ( and the Page Title) accurately describe what you are offering and if so does it encourage visitors to your page ?