Murdok had something of a Q&A session with a couple of web analytics experts about what businesses can do to improve their implementation. We spoke with Brian Clifton, author of the book Advanced Web Metrics with Google Analytics, and E-Nor President Feras Alhlou, who Google itself has even had as a guest author on the Official Google Analytics Blog.
Do you struggle to get the most out of your web analytics efforts? Discuss here.
Murdok: Why do you think so many organizations struggle to go beyond the basic metrics like visitor counts and pageviews?
Brian Clifton: Simply put, its a lack of resources i.e. staff and expertise. All organisations recognise the importance of measurement, but there is a dearth of expertise of knowing what to do next. Remember web analytics is still a nascent industry around 10 years old – similar to search engine marketing. But unlike SEM, web analytics is maturing at a slower pace – so there is less expertise out there to hire or recruit.
The result is that for many a business owner and marketer, web analytics is simply considered a “hit counter”. That is, keeping track of past activity. However savvy organisations who are investing in this area, are exploiting web analytics as a way of looking forward i.e. optimising the user experience and planning future marketing efforts.
Murdok: How can they become motivated to go further?
Brian Clifton: Free is a good motivator and the use of free products such as Google Analytics and Yahoo Analytics has gone a long way in this regard. Money that was once required to collect data can now be used in analysing and understanding the reports and therefore instigating website change i.e. improvement.
Following the obvious benefit of using a free system, sit down with a consultant and perform an analysis on what benefits are achievable. For example, often the result of website changes can be dramatic for the bottom line. One travel client was able to increase online holiday bookings four fold just by optimising their checkout system which consisted of 7 pages. That’s a dramatic change that generated millions of additional revenue. However, to get started ask the question, if our conversion rate was to increase by 1 percent how much extra revenue (or sales leads) would that bring in? Use that number as your first year target and investment cost. From year two onwards, that additional revenue becomes pure profit and you can move on to the next optimisation project.
Murdok: What is the most common mistake you see organizations making with regards to web analytics implementation?
Brian Clifton: Not configuring goals. Defining goals i.e. pageviews or actions that are more valuable to you than others, is the single most important configuration step you can do. A goal can be many things, both e-commerce and non-transactional such as downloading a catalogue.pdf, completing a subscription/registration form, viewing the special offers page, providing feedback, adding a blog comment, or even clicking on a mailto link. Where possible, these should also be monetized.
Murdok: How can this mistake be corrected?
Feras Alhlou: Fundamentally all businesses establish short and long term goals for business success. The same planning process needs to be applied to the organization’s website. Why does the website exist, what does the organization expect to achieve with the website, what are the website goals? These questions, while core business questions, are often overlooked in planning for the website. The days of only having an online presence are over, website business planning has to be integrated into overall business planning. Education is how to correct this mistake. Education of the value of establishing website goals showing the bottom line investment results. Or even start with some baby steps. For example, and for lead generation type websites, set up a goal for number of whitepaper downloads or number of “contact us” form submissions, and show this data to those in charge of lead generation, get that internal conversation started.
Murdok: Can you give an example or two of how an organization can use analytical data to plan for new campaigns?
Feras Alhlou: This is a very good question Chris! A lot of times this question is an after-thought, “oh, we just launched the XYZ campaign, how did we do?” We want to bring the measurement discussion to the campaign planning phase and proactively identify the campaign success metrics. Then ensure you have your website, analytics tool and reports are properly configured to support measuring these metrics.
Let’s take email marketing campaigns as an example. Your email solution provider typically provides a lot of stats about open rate, click-through rates, number of unsubscribers, etc. In addition to these important metrics, we would want to have post-click visibility. Are email visitors coming to the site and bouncing off right away? Are they engaged? Are they buying? These important questions can be answered and acted upon with Google Analytics (and other analytics tools as well). In order to get these insights, you want to ensure that the clicks from your email campaigns are identified as “email” traffic and show accordingly in the analytics report. There is a process of “tagging” the links in your email campaign with parameters so that your analytics tools can place this traffic in the “email” segment.
Another example is measuring lift in traffic off branding or awareness campaigns. You might want to look for a bump in your direct traffic (people who typed your URL directly in their browsers). In addition, you can look at your organic or natural search traffic. A lot of us these days don’t type the URL in the browser, we get lazy :), and we just type the company name or the product name in the Google search box. Google Analytics captures the keywords that searchers type in the search box. You can now, on the fly, segment this data into two groups, branded and non-branded keywords. If you start seeing a significant lift in traffic from your branded keywords, it’s definitely a strong signal, and your branding folks deserve a pat on the back!
Murdok: How much of a company’s staff should be involved the implementation/interpretation of web analytics?
Feras Alhlou: Depends. Every organization varies. The complexity of the website, the number of online properties, the level of marketing programs, the maturity of measurement culture, and other aspects, all play a role in the resources required for implementation and analysis. Analytics might start off as a task on the webmaster or the marketing manager to-do. Hopefully with some early wins, more resources can be recruited (or lobbied for!). The most important element for success is to have a clear idea of website objectives and take few steps towards measuring these goals. From there if the internal resources are not available, outsourcing to consultants is definitely an option we see companies leveraging. Consultants can bring best practices, instant know-how and can help companies build the required process. But regardless, someone internally needs to own measurement and have some passion for it (and getting an executive champion on your side is almost a pre-requisite :)).
Murdok: Is it better to have more people involved or less so it is more controlled?
Feras Alhlou: For implementation more control is probably better, with a system of checks and balances in place. Once we start getting data and reports, we want all stakeholders to see and act on these reports. Some internal training is recommended, even on things like terminology, so your team knows what “conversion” means, and what “bounce rate” is, etc.. Training and extracting actionable insights from reports won’t happen overnight, so plan for a gradual introduction of these concepts to your team members and focus on what is important to them. The power of tools like Google Analytics is that the marketing & web teams have more flexibility to get things done outside of the IT realm. They can have direct access to the reports, reports and dashboards can be automatically emailed to them (daily, weekly, monthly), the advanced user interface allows for real time segmentation, filtering, custom reporting, and trending, all very useful for the marketer.
Murdok would like to thank Clifton and Alhlou for providing our readers with their insight into web analytics implementation.
How many people do you have involved in the web analyitcs process at your company? Comment here.
