The first step is to 'drive' your online business, not just "run" it.....
There was a time when having a website was enough to differentiate a business from its competition. Organisations with a website were able to expand their geographic boundaries and tap into low cost web marketing. New Internet businesses appeared and very quickly grew to challenge the traditional bricks and mortar businesses. How things have changed!
Today, even small “local” businesses have a website and are able to market to whatever audience they choose. This brings them into competition with every other business that provides the same product or message. Differentiation is critical if they are to succeed.
Most businesses think of differentiation as having the best portfolio at the best price and advertising this better than anybody else. Those that succeed however, also think hard about differentiating the way they work. What processes can be improved? Which activities yield the best results? Where are the quick wins? The key to being able to answer any of these questions is the information that a business has available. If every business were to make decision based on the same information, most would make very similar decisions and there would probably be no clear market leader.
Over the last 10 years a whole industry has grown to serve the need for online business optimisation. Those businesses that were early and enthusiastic adopters of this technology are now the leaders in their markets. I have worked with organisations like Thomson Holidays, Oxfam and FindaProperty for many years. Thomson’s online business has grown from 5% of turnover to over 50% during this period. They have used the information available to optimise every part of their online business and outperform their competitors.
Most businesses now have some form of web performance information available. Many have however opted for a free of charge web analytics tool believing that this will give them all of the information that they need. These free tools provide information on traffic and pages viewed but do not generally allow a user to follow a line a thought in order to analyse the cause of something. The ability to analyse why something has occurred enables a user to make the necessary changes to maximise opportunities and avoid problems. Enterprise class analytics systems are designed from the bottom up as analysis rather than just reporting tools. The information is insightful and accessible and probably most importantly can be made available to other systems to enable automated actions to be taken.
Those that really want to drive their online businesses rather than just run them, can achieve business differentiation if they:
1. Use the information available to make decisions rather than using it to justify decisions that have already been made.
2. Analyse segmented website visitor activity rather than just count page views.
3. Profile website visitor behavior and use this to target marketing activities.
4. Match site content to visitor preferences as exhibited by the searches they make and the pages they visit.
5. Understand which marketing activities drive the most valuable conversions.
6. Choose a business intelligence solution that can be moulded to the way the business works both now and in the future.
7. Understand the whole business and continuously compare each element with the others in a consistent way.
Wednesday, January 21, 2009
Sunday, January 18, 2009
How can I cut costs in the right place?
Over the last 6 months, the most common question asked of my client services team is “How can I better understand my conversion rates?” Unfortunately, when you dig beneath the skin, you find that the underlying question is really “How can I cut costs in the right place?”
Marketing budgets are being scrutinised like never before and every element needs to be justified with facts and figures. Each part needs to be analysed and a conclusion drawn as to which yields the best return and which the worst. In order to do this accurately, every organisation needs to have a structured approach to marketing analytics which should include a number of key elements.
Campaign Definition
Some organisations structure their budgets by the technology deployed while others think of a campaign as a multi-channel activity. Whichever mechanism is used, it is important for the definition to be sufficiently granular for analysis of cost and value to be achieved. If a campaign has multiple elements such as email, direct mail and telephone, then a decision must be made as to how value will be attributed. Similarly, if a conversion is achieved after a customer has been touched by more than one campaign, a decision must be made as to which campaign or campaigns are attributed with the sale.
Consistent Measurement
Having decided on a campaign definition, an organisation needs to understand the value of the conversions that each campaign yields. Ad serving, affiliate marketing, PPC and even email systems are all delivered with reporting platforms that provide useful information. Each however uses its own method to measure conversion and each will inevitably claim a conversion if this is possible. The opportunity for double counting and not comparing like with like is inevitable. As a result, what an organisation needs is a central independent system that if flexible enough to analyse every type of campaign and allocate conversions based on the businesses own criteria
Reporting and analysis systems should be able to treat a number of different visitor actions as a “conversion” including the capture of sales information, click-outs, downloads and form fills. It should even be possible to treat a page view as a “conversion”. The system should automatically identify visitors as having come from search engine natural listings or PCC and records the keywords used. This allows an individual keyword to be considered as a visitor's origin and its value compared with other keywords.
It should be possible to compare the value of email campaigns and inbound links with the value of banner adverts and affiliate activity. Users should be able to specify a piece of on-site creative as a campaign and look at how many visitors convert after clicking on the creative and what value the resultant business is. Finally it should be possible to measure the success or otherwise of direct mail that offers promotion codes or directs visitors to landing pages or micro-sites.
Value Identification
Once an organisation has a consistent method of measurement and can attribute a conversion to a campaign or campaign component then it can look at the value that each provides. The value can be immediate or long term and analysis should be performed for both. Immediate value is generally attributed to the campaign that created the visit within which the conversion occurred. It is often the case however that a conversion will occur on a subsequent visit so it should be possible to analyse conversion rates over long periods of time rather than within the visit. In this case a website visitor may have, for example, initially visited as a result of a PPC campaign but then revisited the site a number of times using a saved favourite and purchased some time later.
Organisations should attribute the value of a conversion or sale across all campaigns that either yield direct in-visit conversions or assist in the process . In this way it is possible to identify which campaigns yield high value visitors and which do not. By applying relevant costs, it is generally possible to identify a number of campaigns that fall well behind the norm. It is my experience that "quick wins" are nearly always possible and most organisations can make significant cuts with only a small effect on their overall business.
Marketing budgets are being scrutinised like never before and every element needs to be justified with facts and figures. Each part needs to be analysed and a conclusion drawn as to which yields the best return and which the worst. In order to do this accurately, every organisation needs to have a structured approach to marketing analytics which should include a number of key elements.
Campaign Definition
Some organisations structure their budgets by the technology deployed while others think of a campaign as a multi-channel activity. Whichever mechanism is used, it is important for the definition to be sufficiently granular for analysis of cost and value to be achieved. If a campaign has multiple elements such as email, direct mail and telephone, then a decision must be made as to how value will be attributed. Similarly, if a conversion is achieved after a customer has been touched by more than one campaign, a decision must be made as to which campaign or campaigns are attributed with the sale.
Consistent Measurement
Having decided on a campaign definition, an organisation needs to understand the value of the conversions that each campaign yields. Ad serving, affiliate marketing, PPC and even email systems are all delivered with reporting platforms that provide useful information. Each however uses its own method to measure conversion and each will inevitably claim a conversion if this is possible. The opportunity for double counting and not comparing like with like is inevitable. As a result, what an organisation needs is a central independent system that if flexible enough to analyse every type of campaign and allocate conversions based on the businesses own criteria
Reporting and analysis systems should be able to treat a number of different visitor actions as a “conversion” including the capture of sales information, click-outs, downloads and form fills. It should even be possible to treat a page view as a “conversion”. The system should automatically identify visitors as having come from search engine natural listings or PCC and records the keywords used. This allows an individual keyword to be considered as a visitor's origin and its value compared with other keywords.
It should be possible to compare the value of email campaigns and inbound links with the value of banner adverts and affiliate activity. Users should be able to specify a piece of on-site creative as a campaign and look at how many visitors convert after clicking on the creative and what value the resultant business is. Finally it should be possible to measure the success or otherwise of direct mail that offers promotion codes or directs visitors to landing pages or micro-sites.
Value Identification
Once an organisation has a consistent method of measurement and can attribute a conversion to a campaign or campaign component then it can look at the value that each provides. The value can be immediate or long term and analysis should be performed for both. Immediate value is generally attributed to the campaign that created the visit within which the conversion occurred. It is often the case however that a conversion will occur on a subsequent visit so it should be possible to analyse conversion rates over long periods of time rather than within the visit. In this case a website visitor may have, for example, initially visited as a result of a PPC campaign but then revisited the site a number of times using a saved favourite and purchased some time later.
Organisations should attribute the value of a conversion or sale across all campaigns that either yield direct in-visit conversions or assist in the process . In this way it is possible to identify which campaigns yield high value visitors and which do not. By applying relevant costs, it is generally possible to identify a number of campaigns that fall well behind the norm. It is my experience that "quick wins" are nearly always possible and most organisations can make significant cuts with only a small effect on their overall business.
Tuesday, January 6, 2009
What does the future hold?
What does the future holds for web marketeers and the world of online business optimisation?
User Knowledge
In February 2007 I attended the Technology for Marketing & Advertising show as an exhibitor and spoke to a large number of people. After the show I analysed the enquiries and categorised about 10% of these as “education”; they did not really know what web analytics was and just wanted information. A year later I went to the same show and only listed 2% as “education”. The main reason for the change: Google Analytics. Businesses of all sizes have dipped their toe in web analytics by installing Google. It’s free, it provides the basic stats and, by using it, you soon learn what the issues are and what you really need.
Divergence of free traffic stats and paid for analytics
The recent acquisition of IndexTools by Yahoo has further reinforced the divergence that is taking place within the industry. At one end, the mass market vendors such as Google provide simple, free of charge website traffic statistics. At the other end, the specialist vendors provide visitor behavioral analysis and integrate this information with CMS, CRM and email tools to provide complete closed loop management and optimisation.
Mobile tracking
The use of mobile phones to browse websites is now common and web analytics systems have had to evolve in order to be able to record this traffic. The huge variety of mobile devices creates some interesting challenges. Within the last 12 months most of the serious web analytics vendors have responded with mobile solutions.
Interactive Content – Video, Flash and now Silverlight
As sites become more interactive the concept of the “page view” as the primary unit of measurement for website traffic becomes less relevant. Web analytics systems now have to be able to record the fact that a visitor transitioned from one image to another within a flash object or viewed a particular piece of video. Microsoft’s launch of Silverlight has added yet another format within which web analytics vendors must be able to record visitor activity.
Focus on Marketing ROI
The recent financial turmoil has brought the need for tight control of marketing expenditure into greater focus. The old adage of “you can’t manage without measurement” has never been more relevant. Over the last 6 months, there has seen a sharp increase in the level of interest in marketing campaign analysis. Clients need to know if their budgets are best spent, for example, on email or PPC and within these which email campaign or keyword drove the best results.
Information granularity – the detail and the big picture
In order to succeed, it is no longer enough to simply know how many page views and visits a site receives. It is no good patting yourself on the back when this increases unless you know exactly why it increased and you know you can build on this success. Winning is in the detail. It’s all about completely understanding how visitors use your site. Which campaign drove the best conversion rates for product A? Is this the same for Product B? Does this vary by country, time of day, visitor type or a combination of these? The new breed of web analytics systems provide the ability to easily get at the detail without losing sight of the big picture.
Making complexity easy to understand (visualisation)
Website activity is inherently complex and all of the web analytics vendors have put a huge amount of effort into making their reports easy to understand. Graphical visualisation is now the norm within the specialist vendor marketplace with tables and lists generally only used in support of the visualisation. The mass market stats tools have focussed on page views and visits, avoiding the more complex visitor behaviour analysis and the need for multi-dimensional analysis.
Redefining the Visitor as an Individual
Most web analytics systems think of a browser / cookie combination as a “visitor” and this is how the standards authorities define it. However, real people use more than 1 PC and hence have more than 1 cookie. They browse using their phones and in some cases share an access device with other people. If there is another way to identify a visitor such as through a registration or login process then this can more accurately identify a visitor and potentially allow a system to build up a picture over time that shows the multiple devices used by an “individual”. This allows their preferences to be more accurately analysed. A number of the enterprise class vendors are now differentiating between a “visitor” (used for stats reporting) and an “individual” so as to be able to accurately record a person’s use of a website.
Closed Loop Marketing
The identification of an “individual” enables the integration of behavioral information with Email, CRM and campaign management systems. Recording the fact that a website visitor came from a particular email campaign and feeding this back to the email system with the associated individual’s behavioral profile enables further segmentation to take place. Subsequent follow-up emails can be based upon this information, closing the loop and optimising email open rates and marketing return.
Content Management Integration
The ability to serve personalised website content has existed within the ad serving industry for some time. It is now becoming more common in the website itself. A number of of the leading online publishers now reposition key content based upon the previous hour’s readership by feeding web analytics information straight into their content management systems. Over the next 12-18 months I believe that personalised content will become more common. The current multi-variant testing tools will become a standard part of enterprise strength content management systems and the serious web analytics vendors will provide feeds to these systems. This will enable the content served to be based on the preferences of the individuals as exhibited by their web behavior.
User Knowledge
In February 2007 I attended the Technology for Marketing & Advertising show as an exhibitor and spoke to a large number of people. After the show I analysed the enquiries and categorised about 10% of these as “education”; they did not really know what web analytics was and just wanted information. A year later I went to the same show and only listed 2% as “education”. The main reason for the change: Google Analytics. Businesses of all sizes have dipped their toe in web analytics by installing Google. It’s free, it provides the basic stats and, by using it, you soon learn what the issues are and what you really need.
Divergence of free traffic stats and paid for analytics
The recent acquisition of IndexTools by Yahoo has further reinforced the divergence that is taking place within the industry. At one end, the mass market vendors such as Google provide simple, free of charge website traffic statistics. At the other end, the specialist vendors provide visitor behavioral analysis and integrate this information with CMS, CRM and email tools to provide complete closed loop management and optimisation.
Mobile tracking
The use of mobile phones to browse websites is now common and web analytics systems have had to evolve in order to be able to record this traffic. The huge variety of mobile devices creates some interesting challenges. Within the last 12 months most of the serious web analytics vendors have responded with mobile solutions.
Interactive Content – Video, Flash and now Silverlight
As sites become more interactive the concept of the “page view” as the primary unit of measurement for website traffic becomes less relevant. Web analytics systems now have to be able to record the fact that a visitor transitioned from one image to another within a flash object or viewed a particular piece of video. Microsoft’s launch of Silverlight has added yet another format within which web analytics vendors must be able to record visitor activity.
Focus on Marketing ROI
The recent financial turmoil has brought the need for tight control of marketing expenditure into greater focus. The old adage of “you can’t manage without measurement” has never been more relevant. Over the last 6 months, there has seen a sharp increase in the level of interest in marketing campaign analysis. Clients need to know if their budgets are best spent, for example, on email or PPC and within these which email campaign or keyword drove the best results.
Information granularity – the detail and the big picture
In order to succeed, it is no longer enough to simply know how many page views and visits a site receives. It is no good patting yourself on the back when this increases unless you know exactly why it increased and you know you can build on this success. Winning is in the detail. It’s all about completely understanding how visitors use your site. Which campaign drove the best conversion rates for product A? Is this the same for Product B? Does this vary by country, time of day, visitor type or a combination of these? The new breed of web analytics systems provide the ability to easily get at the detail without losing sight of the big picture.
Making complexity easy to understand (visualisation)
Website activity is inherently complex and all of the web analytics vendors have put a huge amount of effort into making their reports easy to understand. Graphical visualisation is now the norm within the specialist vendor marketplace with tables and lists generally only used in support of the visualisation. The mass market stats tools have focussed on page views and visits, avoiding the more complex visitor behaviour analysis and the need for multi-dimensional analysis.
Redefining the Visitor as an Individual
Most web analytics systems think of a browser / cookie combination as a “visitor” and this is how the standards authorities define it. However, real people use more than 1 PC and hence have more than 1 cookie. They browse using their phones and in some cases share an access device with other people. If there is another way to identify a visitor such as through a registration or login process then this can more accurately identify a visitor and potentially allow a system to build up a picture over time that shows the multiple devices used by an “individual”. This allows their preferences to be more accurately analysed. A number of the enterprise class vendors are now differentiating between a “visitor” (used for stats reporting) and an “individual” so as to be able to accurately record a person’s use of a website.
Closed Loop Marketing
The identification of an “individual” enables the integration of behavioral information with Email, CRM and campaign management systems. Recording the fact that a website visitor came from a particular email campaign and feeding this back to the email system with the associated individual’s behavioral profile enables further segmentation to take place. Subsequent follow-up emails can be based upon this information, closing the loop and optimising email open rates and marketing return.
Content Management Integration
The ability to serve personalised website content has existed within the ad serving industry for some time. It is now becoming more common in the website itself. A number of of the leading online publishers now reposition key content based upon the previous hour’s readership by feeding web analytics information straight into their content management systems. Over the next 12-18 months I believe that personalised content will become more common. The current multi-variant testing tools will become a standard part of enterprise strength content management systems and the serious web analytics vendors will provide feeds to these systems. This will enable the content served to be based on the preferences of the individuals as exhibited by their web behavior.
Subscribe to:
Posts (Atom)