Domain Search
Customer Login
HostwebUK provide Domain Names and Website Design Services for Private Individuals, Businesses, Small Companies in Bournemouth, Poole, Christchurch, Ferndown and Wimborne.
If you would rather have a fully managed service instead of 'doing it yourself' we can provide these services for you,
simply click the 'manage my site' or 'professional help' icons!!
simply click the 'manage my site' or 'professional help' icons!!
Gathering the data
Web analytics
Before you can make any changes to your website you first need to understand how your website is being used (or not). Adding web analytics script to the areas of your website you want to track and monitor (ideally all of it) will enable you to report on:
Visits Page views Traffic sources Keywords used to find your site Bounce rates Dropout rates Conversions Average time spent on site/ pageChoosing your web analytics software
There are a lot of web analytics suppliers to choose from, some of which you have to pay a lot to use. For all but the busiest of websites you should consider adding Google Analytics to your site as standard. It is free and starting to rival many enterprise class analytics solutions.If you want to shop around a bit more other popular (and affordable) web analytics solutions include:
Yahoo Web Analytics (previously IndexTools): http://web.analytics.yahoo.com/ StatCounter: http://www.statcounter.com/ OpenTracker: http://www.opentracker.net W3Counter: http://www.w3counter.comInstalling Google Analytics (GA)
If you already have an account with Google you can use GA from that account. If not, signing up is free and very quick. Google provides a series of comprehensive guides to help setting up your account and adding the tracking to your site. To read more go to http://www.google.com/support/analytics
New Vs return visits Browser/ operating system used Country visiting you from How many pages they looked at Top landing and exit pages Products sold Revenue ROIAdding the tracking to your website is very easy and Google have made it as simple as copying and pasting. Just follow their step by step guide and you will be up and running in minutes
Heat maps
Analytics software can report which links your visitors are clicking on but a heat map will show you which parts of your pages your visitors are clicking on even if it isn’t a link. They also help to visually show which parts of the page are grabbing people’s attention.Using this information you can see any elements of the page’s design that are confusing visitors as well as the popular hot spots. One of the leading heat map providers is http://www.crazyegg.com
User / Customer Feedback
Feedback forms and surveys are great ways to collect data about how people use your website, what they think about your services and how they rate your products. However, writing a good survey with results you can analyse and data you can actually use is not as simple as many people think.Keep it short: No one wants to be faced with dozens of questions especially online where attentions spans continue to fall. To ensure as many questions are answered as possible and reduce the dropout rate aim for 10 questions or less. If you decide to split the questions in to multiple pages
don’t create so many stages that it looks like they are climbing a mountain and always let them know where they are in the process e.g. step 2 of 5
Give them a reason to take part: Provide a strong incentive for the participant to complete the survey and give you their contact details. For example, if you are running a survey to benchmark your industry you could give them access to the results or if it is a survey about your service a competition is always popular e.g. Win a laptop.
Structuring the questions: How you structure your questions will make the difference between having data you can apply and having a large spreadsheet sitting untouched on your computer. Some tips include...
Don’t weigh the answers in favour of a certain answer e.g. Excellent, Very good, Good, Poor (3 positive and only 1 negative) Don’t put “don’t know” or “NA” as an option in middle, it will make statistical analysis a nightmare Avoid leading questions e.g. “Do you agree that...?” Keep the question specific e.g. don’t use words such as “Occasionally”, “Often”, “Regularly” etc. What they mean from person to person differs What does average mean? Don’t use it for the sake of it Don’t fall in to trap of asking two questions in one sentence. “Did you buy our red balloon and what did you think of it?”Open and closed questions: Open questions allow the respondent to answer a question with no limitations on what they can say e.g. “What do you like about your current supplier?” A closed question has a limited number of responses to choose from e.g. “How would you rate our customer service?” Very Good, Good etc.
In theory open questions provide you with more qualitative data but they can be difficult to analyse. If you have thousands of responses you will have to go through each one grouping the answers to be able to spot any common themes. Closed questions make it very easy to analyse the data but you could miss out on some valuable feedback.
Piloting the survey: Good practise dictates you should run a pilot survey to test the questions amongst a limited sample before launching it to the general population. Doing this will highlight any problems with the questions and any in data collection and analysis.
Choosing the software
There are literally hundreds of software suppliers to choose from covering all the stages of planning, implementing, distributing and analysing your survey. A couple of popular suppliers include: http://www.zoomerang.com http://survetmonkey.com http://www.keysurvey.co.ukUsing the data
Being able to look at all the wonderful information and pretty graphs is all well and good, but how do you actually take the data available and apply it to improve your website?Traffic Beyond simply knowing how many people have visited your site, checking your traffic stats every day also provides an insight into how the traffic changes throughout the day, week and month along with seasonal fluctuations, thus allowing you to make plans accordingly, for example…
Maintenance down time in the quietest periods of the day/ night Stock management during busy months Planning advertising schedules offline and online Periods of the day with a higher likelihood of sales/ conversionsYou can also use the data to highlight if your site had any technical issues or, even worse, if it has any ongoing technical issues.
The specific pages people are going to also provide an understanding of how they are interacting with your site. For example, if the ‘Help’ page is one of the most visited pages then clearly people are having trouble finding what they are looking for and you need to look at making changes to either the content, navigation or structure of the site.
Time spent on a page
This is one metric particularly useful when analysing the popularity (or not) of a blog post. If people are spending very little time on a post then that tells you the subject matter is not an area the readership are interested in and you shouldn’t spend time doing any more like that. On the flip side, if they are spending a long time on the page, then you should look to write more posts about that subject or in that format e.g. an interview.Funnels
These are especially useful if you have an order process, you want people to sign up to something or you want people to contact you. Once you have set up your funnel’s goals (the pages along the road to the end result) you will be able to see where people drop out of the funnel, along with how many that started actually went on to finish. From this you will be able to clearly identify which pages are preventing visitors from completing the process and focus on making changes there. Over time, you should be able to see how your changes improve the conversion rate from one page to another.Keywords
The organic keywords report, which shows the keywords people used to find your site, is a fantastic insight into which keywords you should be optimising your site for, along with those that you are already successful with. You can also use the data to find ‘Long Tail’ keywords (i.e. the large list of rarely used keywords you wouldn’t have thought of yourself) to expand your Search Engine Pay Per Click (SEPPC) campaign list further.Bounce rate
Bounce rate is the percentage of visits that land on a page and leave the site without going anywhere else on the site i.e. they literally ‘bounce’ away. Use this data to find the pages that are putting people off going further into your site or to monitor how well targeted any advertising you are running has been e.g. a paid search keyword with a bounce rate of 80% needs either its creative reviewing or pulling altogether.Testing the changes
Once you have a better understanding of how visitors navigate your site, gathered their feedback and put together your ideas to make changes you can now start testing their impact. Website optimisation centres on the ability to scientifically test how your visitors interact with various components and/ or pages on your website and the impact of changes upon that.
By creating more than one version of a page (changing one component at a time) you can test multiple variations to see which delivers the best results. Rather than guessing which colour, image, content order, banner or call to action is most effective you can and let your visitors do the work for you. Variables to consider testing include...
Page headline (size, position, colour, copy...)
Amount and depth of content
Navigation bar
Pricing
Special offers
Testimonials
Logo design
Anchor text in the body
Images
Call to action (where it is on the page, what it looks like and what it says)
A/B split test & multi-variate testing
The main difference between these two methods is the number of variables you are testing at the same time. As the name suggests, A/B split tests compare the performance of two versions whereas multi-variate testing allows you to compare multiple variables at once.
A/B split test
Version A Version B
Multi-variate test
Version A Version B
Version C Version D
Using the data you gathered from your web analytics, user feedback and heat map reports identify which page you feel needs immediate attention. For example, if you find a page in the order process has a high dropout rate or your blog has a high bounce rate. Everything on the page is up for change, but you will see the biggest impact testing the most prominent elements such as the page’s headline or call to action. For example we could test promoting different benefits in the headline:
“UKs premier Swiss watch supplier” (Quality)
“Great value designer watches” (Price)
“Designer watches with free delivery” (Added value)