The explosion of connected smart phones, tablets and desktop computers has made the individual web user more connected than ever, enabling marketers to monitor and collect engagement information at every touchpoint of the purchasing funnel.
At the same time, the evolution of cloud technology and near-infinite computing power connected through real-time networks, has made data collection more attainable for more businesses. The amount of data available to marketers now is unprecedented, but what is it for? How is it useful?
The author of Web Analytics 2.0, Avinash Kaushik, once gave a brilliant definition of digital analytics:
“Digital analytics is the analysis of qualitative and quantitative data … to drive a continual improvement of the online experience that your customers and potential customers have which translates to your desired outcomes (both online and offline).”
He goes on to say that only with the right skills, processes and technologies, can you uncover the vital information about the effectiveness of your engagements and turn them into business results. In this post we’re going to deconstruct this definition and explore some challenges it highlights:
- What do we mean by outcomes?
- What are qualitative and quantitative data?
- What Are Attribution challenges?
- The analytics cycle
What Do We Mean By Outcomes?
One of the most important steps, and usually the first step in any marketing activity, is to determine your ultimate business objectives: what do you want achieve and how are you going to measure it?
A clearly defined measurement strategy will guide the implementation and analysis.
Outcomes vs outputs: an output is a piece of data with little or no meaning, for instance, traffic. A 20% increase in organic traffic is a great output, but what is the implication here on the ultimate goal, say, profit?
The outcome is the final result of all outputs that success should be judged against. In the connected world, the majority of business objectives can be categorized into five segments and each usually has a measurable primary outcome. These are:
eCommerce: measured by online sales
Lead generation: measured by capturing contact details of leads
Content publishing: measured by engagement – which varies depending on your monetisation strategy. Visits, repeat visits, duration on site, etc. Or a combination of the aforementioned metrics.
Online information and/or support portals: measured by the ease of finding information – quantified by user journey length, onsite time, combined with bounce rates and exit rates. Qualitative feedback is key to gauging real success with this one.
Branding: measured by brand penetration, engagement and loyalty. Sign-ins, social follows, shares and comments on content.
What Is Qualitative And Quantitative Data?
We just mentioned qualitative and quantitative data. Let’s talk about that.
Using pageview and event tracking, we can track engagements taking place on websites – a page loading, a link clicked, a bounce, etc. These are all things that you can count, objectively.
Qualitative data helps analysts identify the reason(s) behind changes, trends, etc. and aims to ask the audience directly about things that quantitative data can only suggest.
Examples of qualitative data are: heatmaps, consumer surveys, user reviews and ratings, and live chat feedback. All of these methods help us to understand the user experience. It’s notoriously more labour intensive to analyse this kind of data, but it is increasingly valuable.
For example, live chat logs could indicate that soft enquiries are more valuable than online sales because of the opportunity for a salesperson to cross-sell a more profitable product. This could inform how you target and measure the success of your campaigns.
What Are Attribution Challenges?
Collecting all this data is great, but solving attribution is where data becomes useful to marketers.
Tracking users continuously across multiple touchpoints both online and offline can be difficult but, depending on your situation, there may be ways to improve the accuracy of your attribution.
There are a number of attribution challenges, i.e.:
Users switching devices
Challenge: can’t identify a conversion that started on mobile and converted on desktop
Solution: user accounts, analysing attribution models to understand user behaviour on different devices and in different scenarios
Offline to online
Challenge: how do you know when offline advertising is driving online sales?
Solution: using specific campaign URLs in your print, TV or radio advertising allows you to track the success of that landing page.
Online to offline
Challenge: how do you know when your online efforts are driving offline conversions, ie. Phone calls?
Solution: call tracking software overlays tracked numbers onto your website and online advertising, allowing you to combine call and website data.
Challenge: when users interact with more than one channel before converting, where should the marketing budget go?
Solution: make sure you’re using an appropriate attribution model(s) for each campaign. Some campaigns are designed to generate traffic early on in the user journey, like a brand campaign, and so conversions should be attributed on a first-click basis. Other campaigns, like remarketing, are about getting people across the line, and are more appropriate for last-click attribution.
Assisted conversion reports can also shine a light on how much revenue each channel has played a part in generating:
The Analytics Cycle
Digital analytics should run in parallel with other marketing activities providing the business with real-time supportive information. Our analytics cycle consists of four steps: measure, analyse, report and test.
Measure: collect all the data needed for you to understand your audience, both quantitative and qualitative. An Analytics suite should be combined with live chat, user surveys, heatmaps and videos.
Analyse: extract actionable information and identify reasons for any patterns and anomalies in your data. This includes segmentation, attribution and competitive analysis. You’ll want to use industry benchmarks to put the data into context.
Report: condense the analysis into easy to digest information, with clear explanations of how the data should inform business decisions.
Test: find the best solutions to the problems being identified during the analysis. Testing eliminates any preconceptions and biases from the decision-making process.
The analytical cycle should run in parallel with other marketing activities; it should monitor ROI, provide unbiased actionable recommendations, give additional insight into the results of marketing and drive continual improvement.
A Simple Example: Car Leasing
Business: Car Leasing
Outcomes: ecommerce transactions and telephone sales
- Ecommerce tracking in Google Analytics
- Call tracking, integrated with Google Analytics
- Live chat feedback to collect qualitative data
2. Analyse and Report
- CRM integrated with Google Analytics show that a particular AdWords campaign is responsible for more profit than is indicated by ecommerce figures
- Live chat logs indicate this is because of the opportunity for a salesperson to cross-sell a similar but more profitable product
- Increase spend on successful AdWords campaign and closely monitor ROI
- Split test ad copy to encourage calling or enquiring for the best deals
- Split test messaging on the landing page to encourage calling or enquiring for the best deals
4. Continued Measurement
- Measure test results against expectations
5. Redefine Desired Outcomes
- PPC – generate telephone enquiries
- Ecommerce and telephone sales
Have you spoken to our digital analytics experts about your ultimate business objectives? Do you have a comprehensive digital analytics strategy, implementation and measurement plan that will safeguard your marketing investments?
Do you want to drive more ROI by getting a picture of your business? What are the common obstacles and challenges you face in your business with regard to digital analytics? Leave a comment or drop us a line via the Contact Us form, we’d be more than happy to discuss your goals with you.