Keith MacDonald 9:11 am on Nov 8, 2011 Reply
Tags: ,   

As the ‘new guy’ at Unilytics I have the somewhat unique perspective of transferring from in-house web analytics expert (in my previous role) to web analytics consultant. This perspective, combined with anecdotal stories from fellow practitioners, has lead me to the following conclusion:

For web analytics to be really successful, building trust within the organisation is paramount. Without trust from the organisation, any potential value from web analytics is constantly undermined by fear, uncertainty and doubt.

Building trust in web analytics (as a business service) requires three things: accurate data, thoughtful analysis and end-user adoption.

Accurate data requires ensuring that your technology setup is correct, looking at both on-page code as well as analytics tool configuration. This is as true for free tools like Google Analytics (yes, even “copy & paste” code can go wrong) as it is for paid solutions like Webtrends or SiteCatatalyst, which introduce risk due to the high level of customisation available. Although mistakes will happen, and analysts have an ethical obligation to call them out and correct them when they do, every mistake undermines your efforts to build trust.

Although it may seem like thoughtful analysis is a base requirement for web analytics, good analysis must consider three factors. First, you must account for biases inherent with different data collection and processing methodologies. Second, you must be on top of internally-driven product developments, technology changes and changes to business directives since shifts in any of these three areas will affect user behaviour. Third, user behaviour will be affected by externalities, for example Twitter launching referrals from the domain. Missing the mark on any one of these three compromises the possible accuracy and thoroughness of your analysis which in turn undermines trust in the conclusions you draw and the advice you provide the business.

Mastering your web analytics tool and drawing on the data it provides will go a long way to web analytics delivering value to the business; however, end-user adoption may be the most critical piece to build trust in web analytics as a service. It doesn’t matter what analytics solution you have; if users don’t use it, the effect on the business is the same. While it’s still best practice to have dedicated analytics expertise, training end users on the basics of web analytics means they don’t have to have blind faith – they can look at the data themselves with some level of understanding and confidence, and they’ll have a better idea of how to consume the analysis and insights provided by dedicated resources. Moreover, end users mastering the basics frees in-house resources for longer-term and more complex analysis work.

Combining these three factors together is a cornerstone of establishing good web analytics governance within the organisation, and good governance builds trust.

Even a basic analytics solution implementation with one-size-fits-no-one, regular reporting can provide some value to the business. Without trust however, analytics never graduates beyond this point: it is relegated to an afterthought, simply cut out of the process entirely or is brought in only after-the-fact. Engaging third-party consulting services to help build good governance presents some distinct advantages:

- Consultants are often free from the internal politics and historical problems (e.g. HiPPO) that plague analytics projects and bring a fresh set of eyes to the code and the logical setup
- Bringing in consultants can add a level of importance to the project that brings all of the stakeholders together, often for the first time
- In-house staff are usually so pressed with projects and day-to-day maintenance that user training falls off the radar completely
- Often in-house staff are part of a very small team, if a team at all; bringing in consultants gives them the opportunity to learn best practices, share ideas and discuss strategies

Unilytics presents a number of products and services to help build good governance including RETAIN, Mergence, MARS and the web analytics maturity model. Once these are in place, both the quality of your web analytics service and trust in it will build over time.

Regardless of the solution(s) you use, teaching end users how to involve web analytics (as a discipline) in their daily tasks, while getting them looking through reliable data with expert advice from analyst resources, goes a long way to making web analytics a trusted service within the business. Only then, once analytics is seen as a trusted and essential service, can the business begin to realise serious value from web analytics.

In short: even if you get the implementation right and hire the best people, it’s all for naught if the business doesn’t trust web analytics.