The concept of Voice of the Visitor / Customer data analysis was originally developed by the promoters of Total Quality Management (TQM) back in early 1990s. VOC measurement involves analyzing qualitative data obtained from the customer’s through various methods such as feedback forms, questionnaires, surveys and polls and using this data to optimize product/services and marketing campaigns.
Voice of Visitor approach piggybacks on the VOC concept but goes a step further by promoting analysis of the web visitor qualitative data. Complaints and low ratings are also considered to be the Voice of the Visitor and this data can be used by the companies to changes to their services and/or their product line.
Few years ago, capturing qualitative data required a lot of time and money because of the manual data collection and analysis process. The results were mostly inaccurate because of the smaller sample size and mistakes in manual calculation.
With sophistication and popularity of the web analytics tools, the data collection methods for VOV have also changed. Offline forms and feedback sheets are replaced by sophisticated online surveys, and manual calculation is sidelined by advanced regression analysis of the collected data.
However, it is not easy to quantify the Voice of the Visitor in terms of simple calculations: one cannot say X % of people found this product good and Y % of people found it bad. If Y contains 90% of the company’s most loyal customers, even if X is much larger than Y, the company should consider this as a serious issue and fix it in order to make it suitable for their old customers while laying down several bases for their new customers.
Analyzing and measuring Voice of Visitor data is a fragile business – it has both an emotional as well as a practical side to it. The emotional or the social side deals with loyalty, customer satisfaction or complaints, while the practical or enterprise side deals with completion rates, review scores, speed and ease of use. While customer satisfaction may not always lead to loyalty, it becomes an abstract positive score in comparison to tangible scores in speed tests and reviews.
Here are some best practices for capturing and measuring VOV/VOC –
1. Likert scale model – Always use likert scale model for surveys and feedback forms. Yes/No type surveys popularly used by car dealers are outdated and cannot tell you anything about forecasting customer behavior.
2. Questions – Limit the number of questions per survey to ten or less. There are companies promoting 50 plus survey methodology, and it could only work if you have time, money and patience to wait for months before you can perform actionable analysis.
3. Incentives – Never offer any incentive to your visitors/customers for completing the survey/feedback form. Incentives can influence the user behavior and will result in skewed data.
4. Integration – I highly recommend integrating the VOV data (qualitative) to your web analytics data (quantitative). Integrating both data set can provide you with extremely powerful information about your visitors. You will be able to identify the referral, pathing and buying behavior of your satisfied/dissatisfied customers.
5. Metrics – You can create your own metrics depending on your business objectives and goals. For a retail business, it could be the average acceptance rate of number of steps for an order process, for informational sites it could be the page with the most useful content. Always include the “what is your overall satisfaction with this site/product/service” to get a holistic view of response.
6. Tools – Here is a list of the top five free and low cost tools for the voice of visitor/customer analysis. If you are looking for an enterprise solution then I recommend Foresee.
It is imperative to use this analysis effectively for the expansion and success of your company.
Hope you enjoyed this post.
Please share your thoughts, insights and comments because it really helps me write more effectively.
Here is a list of other helpful blog posts –