Less than .5% of web site traffic was driven by social networking sites based on analysis of 6 UK charity web sites for the period 1st Oct 2009 to 28th Feb 2010. The sites examined had monthly Visits between 20k and 800k.

Most popular Social Sites

Social Sites such as Facebook and Twitter are now major sources of traffic for many web sites and for some web sites rival Google in their ability to drive traffic.

However for these UK charities the social  networking sites are not big drivers of traffic. The analysis showed that Facebook and Twitter accounted for over 95% of visits from the social sites, although Stumbleupon had some success. While the data says that Twitter brought about 33% of the visits that Facebook brought in reality Twitter probably brought more as many believe that Google Analytics undercounts Twitter visits by a factor of 5 or more.
Social Networking sites have an important role to play in reaching specific audiences but for now it is best to tag each tweet and check the stats for the specific landing page to really measure the impact on visits.

Back in November Google introduced Ad Sitelinks which allows you to put 4 additional links to your site into your adwords ad.

Google RNIB AdWords Sitelinks

Ad  Sitelinks give you more screen real estate in a prime location and early indications are that it results in more clicks. However to take advantage of it you need to have a good AdWords Account CTR. My experience is that you need a CTR of around 3% spread across all your ad Campaigns to be offered the Ad Sitelinks option. You can check if you qualify by logging onto your adwords account, go to Campaign > Settings > Networks, device and extensions. Here, if eligible, you can enter a brief description and the associated url for the appropriate pages on your site.

Location of Ad Sitelinks settings in AdWords

Google deals with the question `What ads qualify ? and they specifically say ” Ad Sitelinks are most likely to trigger on unique brand terms”. My experience has been that this is the case on all accounts that I have done this for.

Now on your main brand term you can occupy a large amount of google SERPS screen real estate – your AdSitelinks, regular site links and (in many cases) a site specific Search Box plus Local Listings. Anyone looking for you under your brand term should be able to find you.

Two keys metrics in Google Analytics that are very widely used, particularly by occasional users, are Traffic Sources and Bounce Rate. Both are useful for example when applied to individual pages but they need to be used with an understanding of their limitations. Direct Traffic is defined as ” people who clicked a bookmark to come to your site or typed your site URL into their browser “.  While that is true Direct Traffic can also include any visitor that the system can not identify the source of. Examples would be visitors who came from a 301 redirect, from a Campaign eg banner, email where the campaign links were  not appended or from a visitor who is using the Private Browsing feature of Explorer 8.  In order to ensure that the correct marketing channel gets credit for bringing visitors Google Analytics stores for 6 months the Traffic Source the visitor first came from ( in the _utmz cookie). So if a visitor finds a page using Organic search ALL visits subsequently made by this visitor ( actually all visits made from the same browser on the same machine ), EVEN if she comes from a bookmark that she made after her first visit, will register as coming from Organic.

Bounce Rate is defined as  ” …the number of visitors who enter the site at a page and leave within the specified timeout period without viewing another page, divided by the total number of visitors who entered the site at that page “. So it does not take into account the number of Page Views by Visitors who did not land on that page. For example a page may have 5 direct visits and 4 leave immediately ie a Bounce Rate of 80%. However that page may also have 200 Page Views from other visitors to the site. The average time on page may be long. That piece of data gives a very different `spin` on  how that page is performing.

The latest Google Analytics update has one new feature that is extremely useful for information sites – Engagement Goals. For charity and nfp sites I have found that the traditional url based goals do not give a clear feedback if users find the site useful. The new goals, which are very easy to implement, are well explained here by Justin. Putting in an initial set of  Goals covering aspects of site engagement such as Time on Site and Page Views shows over time, by keyword, etc how interesting and useful visitors find your site. Before entering Goals it is useful to check the site average for the two main metrics and then take advantage of the `less than` feature in Goals. So for example if the average number of page views is  4.76 per visit you could set  the first Page View Goal at 3 and identify those who are below average. Note that while Goal names and parameters can be changed the data relating to the original goal will remain. So when you set up the Goal initially and specify a threshold of say 3 pages and later decide to up that to 5 pages viewed to achieve the Goal then the Goals achieved will be based on 3 pages for dates prior to the change.

Google Analytics Goals layout (click for larger image)

Sadly you can only include Event Tracking in Goals by using the `_track Pageview function`. This is for most users too complicated to implement – you have to involve the web dev or IT guys and we all know how busy those guys are. If you can track events eg downloads, sign-ups, feed subscriptions, phone call requests, etc then you have two great sets of Goals that you can measure progress with. By Source, Keywords, etc you can see first how Engagement is improving and then how visitors are acting on that engagement.

Goals achieved by Traffic Medium (click for larger image)

Remember that Goals are cumulative so that on the extreme right (above) the Goal Conversion Rate is the sum of all Goals achieved. If, as here, you have Goals of 2, 4 and 6 Pages Viewed then a Visitor who Views 6 pages will have achieved 3 Goals.

Implementing the new Goals is really easy and is a great step in understanding user behaviour.

The presentation on Search Engine Marketing that I did for UK charities and nfps back in June 2008 has now been viewed over 1,000 times. It covers the basics including products such as Google AdWords and Google Analytics. Hopefully people have found it useful. It feels like`good karma`- I have now given a  little back having been a major beneficiary of all that great data and info on SEM available to all at the click of a mouse.

Since I gave that presentation I have been working on eMarketing and Search Engine Marketing challenges with some UK charities – including YouthNet, RNIB, maggie`s and of course Breast Cancer Care. Top thing that I have learned so far ? that web dev teams are really busy making those lovely features that marketing want. Meanwhile marketing groups have web monitoring tools that would really benefit from dev team input. But everyone is too busy to sit and review the data. The result is sub-optimal measurement and understanding of the web site performance.

One of the most common questions from pretty much any organisation that uses Google Analytics (GA) is `what Key Performance Indicators (KPI) should we be monitoring ?`. Of course the simple answer is to measure what is important to you. This points us in the right direction but gives no clear answer. For me the biggest challenge when talking about GA is the lack of engagement from marketing, fundraising, management, etc with web stats. Now, if the finance dept show data dealing with budgets or wages then there is noooo problem in getting attention to and understanding of numbers. So we have to conclude that the reason that few pay real attention to GA stats is that they feel the data does not impact them and/or they have no control over it. So the 3 KPIs I have chosen are designed to show that GA data can tell you something important and you can impact the customer experience:

1.`Nobody likes my pages`- bounce rate: this is the % of site visitors who entered and left the same page without going anywhere else. For the biggest impact go to GA> Content > Top Landing Pages and over on the right you will see the Bounce Rate. Below 30% is excellent, 30- 50% is not bad. Anything above 80%  is a bit embarrassing for the person or group who created the page. However do check the `Time on Page`as some pages such as Sign-Up, Register can legitimately have high Bounce Rates.

2. `People do not like our stuff`- engagement: most nfp sites are designed to provide information. What should be happening is more visits consuming more info ie staying longer. Look in Visitors > Visitor Loyalty at both Length of Visit and Depth of Visit.  Over 70% of people looking at just one page and staying less than 10 secs is not a good sign. To isolate this data for a group of pages or a section of the site either create a specific Profile or use Advanced Segments.

3. `Yeah, but our visit number are growing` – share of search: the number of visits is rising and everyone is happy but what if someone else in your market is gaining visitors 3/4/5  times faster than you ? You have a problem and you do not even know about it !  Compete will give you info on how you are doing relative to your competitors. Google Trends will show you what people are looking for. Google Trends for Websites gives you data on your competitors. Finally look at the Keyword Tool and see how many searches per month there are for your main keywords in Exact Match. You should be getting between 20% and 50% of that traffic if you are in top spot and if you are not in top spot for your top brand or cause related terms then your audience is going somewhere else.

ah. But what about the stuff that everyone is talking about eg Conversion Rate,  Donations / Sales per visit, etc?  For most of us the first tasks are to show that GA data is relevant and individuals can influence it directly. Once you have won those battles *then* you can start to build KPIs that matter to people and that they will pay attention to.

Google`s new search engine produces, as expected, a range of results when tested with top keywords from UK charities. Announced last week the beta version of Google`s search engine, codenamed caffeine, with the same interface as the existing search is available for testing now.

To see how it may affect UK charities I compared the search results for the current Google and `caffeine` using the nifty comparison tool put together by sembience ( nice example of link bait !). For four different sites that attract between 50k and 250k visits per month I looked at their top 10 keywords. From them I excluded the `brand` terms eg organisation name and terms clearly not related to the organisations mission – leaving about 5/6 keywords per organisation. For one organisation `caffeine` put all 5 terms lower than current search results. Two organisations had a majority of terms producing lower results and one had no net loss or gain.

Before it becomes Google`s new search engine no doubt `caffeine` will have a lot of modifications but it is well worth checking your top terms now and seeing if you are on track to lose or gain visitors when it happens.

Google recently updated Webmaster Tools making it easiser for  marketing and business types to find great data to improve site performance.  The good stuff is under the tab Your site on the web.  Top Search Queries tells you the number of times your site appeared in Google search results ranked in order plus the percentage of those top queries represented by each query. So to begin with you are looking for the larger percentages. Next, to the right of the query term, is the Position that your highest ranked page appeared for that query. So if you are ranked low for a query term that has a large percentage of impressions that represents a good opportunity for more traffic for that term if you can get a better ranking.

Webmaster Tools Top Search Queries

Further to the right is the section headed Clickthrough. Here you can see similar data but  now for those who clicked through to your site.  First  look at the search query terms.  Are there any here that you did  not expect ? If you are gettings clicks from a term that you do not rank highly for and want more of that traffic then get working on that. Note that you can select the specific date range and region that you are particularly interested in.

Further down Your Site on the web is the Keywords menu. This shows ” the most common keywords Google found when it crawling your site “. This is great data when planning the next steps for your site.

The AdWords Impression Share Report tells you the percentage of times your ads were actually shown compared to the total number of times your ads could have been shown. Google introduced it some time ago as a share of market/ share of voice measurement. While it is not an easy report to fully understand it does give useful information which allows for more targeted and successful Campaigns.

Impression Share Report

The Impression Share Report can be created in the AdWords Report centre. The report shows the Impression Share(IS) for the Campaign and the percentage lost due to either the budget being too low, Lost IS (Budget) or lost due to poor ranking, Lost IS (Rank). For most Campaigns the largest loss has to do with Rank.  As Rank is CPC X Quality Score the most cost effective way to improve is to improve Quality Score.

Given that not all of the keywords in the Campaign will have a great Quality Score and your Maxm CPC will not always be high enough it is very unlikely that you will get 100% Impression Share (IS).  One reason for low IS could be that the keywords are not showing up on broad match. The Exact match (IS) figures helps here as they show how the Campaign would perform IF all keywords were Exact Match. Increasing your bid will also improve your Ranking too. However in chasing greater Impression Share remember that 30% or 40% Impression Share of a large `market` is better than a 100% IS of a small market. So do not be afraid to keep less than optimal keywords in the Campaign.

One drawback of this report is that it is only available at the Account and Campaign level where there is often too much data to make sense of it all. If this report was available at the ad group level I think it would be more widely used.

Building an efficient keyword list is critical to the success of an AdWords Campaign – whether using a Google Grant or a regular paid-for AdWords account. The best and easiest way to start a Campaign is to begin with small groups of closely related keywords/ phrases covering a single idea in each AdGroup. Initially they should be all Broad Match. With Google this means extended Broad Match ie your ad will be triggered by synonoms or keyphrases related to your initial keyphrase.  After a week or so of the Campaign you should check how each of the main keywords are performing.

New Search Query Tool in AdWords

The new AdWords interface allows you to very easily access this information, by AdGroup or individual keyword, via the Search Query Performance Report.  Say you had an AdGroup relating to working part time and you had  included the keyphrase `part time`. After running the Campaign for 7/10 days you check the Search Query Report and see that for the keyphrase `part time` 25 clicks were registered. Of this 25 the report shows that 5 were as a result of Exact Match,  10 via Phrase Match and 10 via Broad Match. Review the keywords marked as Phrase or Broad Match and see if any of the search terms that triggered the ad were used 2 or more times. If so add these terms into the AdGroup as Broad Match. Doing this regularly will exand your keyword list with more high performing keywords. Note that if any of the Phrase or Broad match terms are not related to your ad eg `electrical timer`, `kitchen timer`you should put `electrical` and `kitchen` as negative keywords to prevent your ad from showing for these unrelated terms. Keywords that regularly achieve less than a 1% CTR should be looked at very closely with a view to either deleting them or moving them to a more appropriate AdGroup or Campaign.