Making analytics make a difference - our interview with David Sealey
How do you try to make use of your analytics? We all know that in theory web analytics tools give us fantastic insight to help us improve results for our digital marketing. In practice you need to know the right questions to ask and use the right reports and measures to find the answer. You also need to know the intuition to follow-up on a hunch.
Every marketer and every analyst will have their own unique style. As a consultant I'm always keen to learn how others tackle a problem so I can learn from their ideas. In this interview, David Sealey of Design and Analytics Agency Quba shares his approach on how to avoid "analysis paralysis".
Current marketing challenges for 2011
Question 1: What do you see as the current major challenges faced by marketers and how can analytics help?
In this tough economy, marketers are under pressure to deliver more leads with less budget and time. Lead quality is a further issue for marketers who are working alongside sales teams in their business. These challenges are amplified on the digital side due to the visibility of effort expended to results.
We also live in a day and age with multiple marketing channels which need to be assessed and managed. Digital channels alone include web, video, email, SEO, PPC, affiliates, mobile, social media and online PR. If you"€™re not following the latest trend then you"€™d better have the evidence ready to defend that choice.
Today"€™s data transparency is a mixed blessing for marketers. On the one hand, it provides real insight into clients"€™ mindset and behaviour. On the other, it leads to confusion, half-baked conclusions and "€œanalysis paralysis"€.
When I started my career in digital marketing, traditional marketers were playing catch up. Our knowledge of HTML, SEO and new media made us kings of schoolyard. There is now a reverse of this as digital natives find that they have to understand the basics of marketing as a discipline in its own right. So whilst the 4Ps (product, place, promotion, price) or AIDA (attention, interest, desire, action) won"€™t replace new technology, they do need to work hand in hand. A perfect example of this is the merging of old school PR with the new kid on the block that is Social Media.
Where to focus investment
Question 2: What focus of time and budgets do you recommend to get better approach?
The projects I"€™ve seen over this last year have had more to do with enhancing the efficiency of a website over completely rebuilding it. When budgets were available marketers could just push the entire website out to an agency as a project but with budget restrictions and a focus on the bottom line marketers are pursuing sales and leads rather than design.
The focus is definitely on wringing the maximum value out of the web channel before completely changing it.
Obviously this is a much more cost-effective exercise and the results should be clearly visible through basic site metrics. With analytics tools that are ever easier to use, identifying the sticking points is a more straightforward exercise.
I personally appreciate this approach as the focus is on the results (an objective measure) rather than the creative (a more subjective matter). With the focus fixed on definable goals (e.g. double the site conversion rate) it is easier for everyone to pull together their ideas.
There is a downside to enhancing a website. We all know and accept that any technology will have bugs that occur. In enhancing a website, these bugs can be amplified or create additional errors. This adds to the overall technical debt which the system carries and your repayments (i.e. the amount of technical fixing time required each month) will only increase.
Whilst enhancements to an existing site may be a cost effective strategy, innovation may suffer due to technical platform limitations. Such innovations are needed to engage with users and be competitive.
Metrics that matter
Question 3: Which metrics do you think are important for marketers to focus on?
I"€™ve talked with marketers who are stressing over metrics that really don"€™t matter. "€œMy average time on site is down by 30 seconds! What do I do?"€ they ask in a blind panic, fearing the approach of the MD. The metrics I"€™m primarily interest in are:
- Unique Visitors
- Bounce rate
- Conversion rate to goals
- Average order/enquiry value
These stats really tell the story about the state of a business website. By addressing and segmenting these numbers first, I identify what can be improved and the goals that can be set.
User testing can then be used to provide much more insight on the causes. For instance if your landing page has a high bounce rate, submit it to a brief round of user feedback sessions. Show visitors the ad that leads to the page as well to give it context and then ask what they think of the page? Was it what they were expecting? What is the main offer? What do they think the business is offering? I saw one promotion where the subject line of the email marketing offered an Amazon voucher but the site offered a M&S voucher. This confused visitors, making them wary of the offer. This friction resulted in the user exiting the site. Submitting this to user testing identified the issue and provided a quick fix.
The value of the feedback you receive will give you the insights necessary to make (or at least a/b test) improvements.
Outside the stats you need to be considering the entire end-to-end user experience. From the ad that they see before the site to any interactions afterwards, everything should be reviewed for consistency and ease of use. For instance there was one project I worked on where the email sent after registration didn"€™t contain the name of the company or a link to the site. Make it easy for people to do business with you.
I also find that digital marketers are either thinking only in short-term or long-term timelines. The web should make use of both short-term tactical campaigns though PPC or advertising and long-term activities such as social media and SEO.
How to apply web analytics
Question 4: How do you think marketers should best use analytics to get better results?
Analytics tools have greatly improved and Google Analytics has played a large part in moving both the industry and user intelligence forward. Its reports and interface have made web analytics accessible to everyone in large and small organisations.
However for analytics to really move businesses forward and provide greater value, the tools need to become smarter and simpler. With web analytics there is still a large amount of work to be done in interpreting the results into an insight that can affect change. Knowing that your bounce rate was 35% for the last twelve months isn"€™t actionable insights yet many marketers are sweating over figures like this.
More visual representations are required too. I use CrazyEgg"€™s excellent click tracking tool and when I sat down with some clients recently to review the results they were blown away by its simplicity. One immediate insight that they"€™d never gathered from Google Analytics was that the two videos were never clicked on. In 10,000 visits only one person had clicked the play button. The change was clear, remove the videos and free up screen real estate.
Rapid and simple measurement needs to become part of the analytics toolkit. I"€™m always frustrated that I can"€™t reverse test goal funnels or instantly process event tracking data and that I need to check back in up to 24 hours that changes have been processed. Furthermore complex Javascript is required to complete the recording of metrics that should be straightforward (such as first and last click analysis).
In the future I think Analytics needs to move forward to a point where it can be easily configured by non-technical teams to record what is important to marketers rather than analysts. Deeper integration is also required with CRM systems to improve attribution of customer registrations and orders.