Part IV: Web Analytics: Taming the Information Beast
[The following is part of a six article series on CRM by Ashley Friedlein, CEO of London-based e-consultancy.]
Customer profiling and analytics are clearly core to CRM--you need to be able to understand and measure customer value. For digital channels this is arguably even more important as the 20:80 rule of thumb is even more pronounced. Online, 10 percent of customers currently account for 90 percent of value. And, as the old adage goes, you cannot manage what you cannot measure. Equally, you cannot measure what you cannot define. So how should you understand and approach Web analytics?
The Theory
The Internet is often heralded and sold as the ultimately accountable medium with huge sales and marketing opportunities arising from analyzing the rich customer interaction data. In theory, Web analytics enables the following:
The capture of customers' behavioral and preference data, giving more valuable marketing insight than traditional demographics
Hugely increased speed ("real time") at which marketing messages could be optimized and campaigns turned around
Richer customer profiling for improved targeting and more refined communications to better extract customer value
The creation of additional sales opportunities for better-informed agents talking with customers
An improved capability to predict the profitability of customers and clear insight into how best to acquire them
Equally, the ability better to predict and mitigate customer churn
The platform of knowledge for Web personalization, improving the customer experience and, therefore, increasing competitive advantageThe Practice
As many companies have found, the practice is not quite so easy. The main hurdles include:
Information overload. Online gives you enhanced capabilities for data capture and analysis but this can lead to data paralysis if you are not sure exactly what you want to be analyzing and why.
Data quality. A recent PricewaterhouseCoopers survey of 600 IT executives revealed that fewer than 50 percent had confidence in the quality of their own data. Web data, in particular, has come under fire for its lack of reliability and accuracy. Use of the much vaunted behavioral data for personalization has been shown to frustrate and deter customers in some cases.
Few common standards or metrics. The Internet is still a relatively new channel and although great efforts are being made to standardize around particular protocols, metrics, practices and processes, most companies are finding it hard to build a robust measurement framework.
Privacy and security. Capturing, storing and analyzing customer data, particularly customers from many countries and jurisdictions, presents considerable challenges in meeting legal and customer expectations for privacy and security.
The single customer view still isn't there. This is the "big" problem. Very few companies are really that close to integrating customer data across all channels and touchpoints to be able to analyze and understand customers' interactions with the business in a holistic manner. For the moment analysis still tends to be around silos of data--Internet channel data typically being one of those silos. Companies are progressing from first generation Web site statistics (hits, page impressions etc.) through second generation e-business intelligence (online purchase frequency, conversion ratios, etc.) to third generation e-customer intelligence (propensity by segment, customer value modelling, etc.). However, in isolation of other channels this only goes so far and does not accurately show how customer value is created through channels complementing each other (e.g., research online, then purchase in store).
The Way Forward
A key challenge for Web analytics is not to be tempted into analyzing everything just because you can, but to work out what you need to analyze to run your business more efficiently and analyze that. What are your key business drivers in maximising profitability? What "levers" do you have to pull and what effect will they have? If you have built customer value models, what metrics will you need to report against them? What analysis do you need to understand where customers are in their lifecycle or value chain so that you can effectively move them on and up? By understanding the network of levers and drivers that underpin your business model, you can architect a framework for measurement which has corresponding metrics that allow you to gauge what effect your business activities are having.
Online, once the required metrics and reporting processes are in place, it is indeed possible to have unparalleled levels of real time insight into the performance of your e-business. This insight provides the intelligence to make better and faster decisions in the effort to maximize returns on investment. But it cannot just be about e-business. You must work towards integrating web analytics with enterprise reporting. Only this will give you proper insight into how you can optimize your use of channels to maximize customer value. If nothing else, this will help paint a much truer picture of the real returns on investment that your digital channels are delivering.
The measurement framework that you create, founded on business drivers, is incredibly important. In itself it has enormous intellectual capital value for your business. In terms of Web development, for example, it is hugely important for informing the information architecture of your site (not dictating it--customer needs should do that), informing the data tagging and architecture of your content management system, etc.
Do not be over-ambitious with what the framework attempts to measure. Concentrate on key metrics and building bridges (using quantitative and qualitative techniques) between data silos. Particularly for digital channels, the value of information depreciates very quickly--speed, targeting, flexibility and agility are paramount in order to capitalize on moments of value--so it is better to do a few things really well than attempt too much. You need to allow adequate time for the actual analysis and fine-tuning of reporting processes to ensure the right people get the right information at the right time to deliver actionable insight.
What It Means to You
Do you believe, as I do, that different people want to buy different things through different channels at different times? Do you believe that the customer mindset determines propensity and chosen route to purchase? If so, then the better you understand these people, these mindsets, these purchasing dynamics and how they vary by segment and by channel, the better you can craft your proposition to match customers' needs. If you can do this they will spend more with you.
Web analytics is a vital component of maximizing customer value but it must be seen not as and end in itself but as a way of understanding how the Web and digital channels can best contribute to the overall business goals.