How to measure the ROI of web analyticsby
It’s safe to say that web analytics has moved far beyond its development phase, placing itself at the centre of an environment where companies are looking to squeeze the most out of their investments of time and money.
Analytics can be used to prize valuable customer insights from the depths of the web, providing the basis for strategies to tackle high bounce rates, issues with converting and even improvements in customer response times among several other applications.
But increasingly, questions are ringing out regarding exactly how many conversions were driven; how much time was saved, and the spend it took to get there.
Many will point to an air of irony in the fact that web analytics, used by businesses to asses the return of investments pledged to their digital platforms, is now having to prove ROI itself. However, its status as an ‘enabler’ is certainly something to bear in mind when doing so.
The need for return
Standing as a fixed-cost investment, analytics tools enter a business much like the personnel that control them; with a price tag on their head. Competition among service providers means the going rate of analytics platforms is on a very gradual decline, which goes a little way towards counter-balancing the rising cost of talent.
A little way, that is. Online commerce paved the way for small businesses to compete with the goliaths of their industry, and it’s with the same shrewd vision that these enterprises will be looking to forge a clear link between their usage of analytics and the bottom line.
That practice of ROI trawling can even grow into a benefit in itself, as once it’s clear to see where the return comes from, the technology’s value is amplified - a vital consideration if the initial investment is minimal.
But before attempting to measure any return from a solution, those in charge of the tools need to know primarily what it’s being used for. Setting clear goals for results will ultimately make it easier to judge success and there are plenty of ideals to choose from.
“Return will be measured in reference to the key performance indicators (KPIs) which must be defined at the outset of any project or programme,” says Rob McLaughlin, analytics director at global marketing agency DigitasLBi.
“These KPIs will represent the drivers that have been commonly agreed to ‘move the needle’ for a business, coming across a range of areas such as, but not exclusively; reach, sales, service, lifetime value or customer satisfaction.”
McLaughlin’s process dictates that if actions based on analytics insight are being delivered against a KPI, then return is being generated. A similar approach to matching analytics data with the activity it inspired is observed even by the tool providers themselves.
Axel Schaefer, senior manager for product and industry marketing EMEA at Adobe Systems, highlights a rarity in analytics being able to generate additional revenue in a cut-and-dry manner.
“Usually, our customers are looking for the correlation of web analytics with other channels, so for example how did my segmentation of web visitors help us to create additional revenue in the next email campaign?”
There are of course exceptions to this role; instances where analytics will be able to deliver a direct return to the business.
The capability of modern-day platforms means that companies can drill down into every aspect of their site, and if they have goals for performance based on engagement or traffic, they can weed out the under-performers based on the findings. Once adding up the price of a costly image or piece of licensed content, it’s easy to see where money has been saved.
Drawing the line
Even the briefest of conversations with an analytics expert or tool provider about the return generated from their actions will come around to the same point: it’s holistic. Generally, ROI must take into account an ability to influence the performance of something which directly impacts the bottom line.
Perhaps the most salient point to come from the stages following goal creation is that it really is up to the business to design their own graphs from the result (the engagement, the conversion), back to the tool, which stems from a need for analytics to provide a lift to metrics like customer lifetime value (CLV).
One example could see a business using analytics to glean insights across their cost-per-click activity.
If that process unearthed areas whereas spending should be increased, the result of that change could be used to obtain ROI. For customer lifetime value (total revenue from one user’s purchases ÷ the number of months they came in) comparisons between a campaign, landing page or channel’s performance before analytics was used and how it fared afterwards, can highlight the return.
That line to ROI, unfortunately, comes with its fair share of barriers.
Relying on another party to prove value makes communication between tools and in real life all the more important.
For the former, the true challenge lies in collecting and combining data from several different platforms to determine success. On the latter - a consideration for the specialist - it’s all about communication with other departments, like sales and marketing, especially when it comes to measuring the impact of analytics on offline performance.
There are even things that cannot be measured for return, as Mclaughlin states.
“Equating an overall ROI to analytics is further complicated by the variety of ways that return can be understood. Whilst some insight can directly lead to actions which appear to drive sales, other insight may lead to understanding opportunity cost and may justify inaction, with returns impossible to calculate.”
Some businesses won’t require the granular analysis that a high-street merchant would as they go about rewinding a conversion in a brick-and-mortar store back to a landing page. Challenges at entry level come in the form of making the most of the data on hand; using this insight to inform business decisions and prove ROI in the aftermath.
The CMO Survey provides a progress report on this every year. In 2012, 37% of marketing heads claimed to have used analytics before making a decision within a project, but this went on to drop to 29% a year later. In 2014, a slight uptick saw 31% using analytics to inform their business decisions.
While analytics has done so much for improving results across digital properties, there is evidence of the ‘data deluge’ becoming too much for some. It only stresses the value of good talent behind the controls, and working to pre-set goals before moving onto something else.
Appreciation of the role
What certainly comes from learning more about the uses of web analytics is an understanding of how it works, and why it’s needed - with or without a balance sheet highlighting a return of X.
Schaefer insists that prospective clients “usually” won’t come to Adobe with return as one of their first thoughts, although experience of use can eventually make it easier to reach that goal.
“Our customers understand how important it is today to understand your customers and web visitors in order to gain deeper insight, which is most likely the most important element of the digital world.”
On a similar point, McLaughlin says: “Analytics is an enabler to other functions and a fixed cost which is at least one layer removed from return generation.
“Even in the case of optimisation there will always be a layer of user experience (UX) or media activity which will be key in generating returns. In fact, it is the integration of analytics teams and practices within these related disciplines which truly requires scrutiny to ensure that insight is successfully being delivered to the frontline of the business and can be effectively put into action.”
It’s for this reason why any attempt to obtain ROI from an analytics toolset - the “long and winding” process - should be tackled with a simple equation.
As Schaefer states: “Success in web analytics is efficiently using what is available to you and knowing what you want to get out of web analytics in the first place.
“Once you can define those aspects clearly, then you can measure the return more easily.”