“Must do analytics better” – surely at the top of everyone’s New Year’s resolution list? Digital and non-digital data gets bigger (i.e. more confusing) every year, so how can you get your analytics working for you? We share our top tips.
1. Take your basic metrics with a pinch of salt
Unique visitors are not people. Average time on site is not the average time people spend engaging with your content. Bounces are not always bad. Make a New Year’s Resolution to probe what value the “standard” metrics really give you and don’t be afraid to ignore them if the answer is “not very much”.
2. Small is beautiful
Forget “Big Data”. Unless you have a crack team of stats wizards that map-reduce data for breakfast? Focusing on an optimal (and small) set of KPIs (Key Performance Indicators) will pay dividends time and time again in picking out actionable insights. Invest time in getting those KPIs right, collecting the right data and focusing on outcomes that can drive practical change - not searching for magical trends.
3. Don’t underestimate Google Analytics
The free Google Analytics feature-competes with the paid vendors; but exactly like those paid vendors you have to invest in getting your deployment right to get the most out of it. Don’t assume that just because it doesn’t appear out-of-the-box it can’t be done in Google Analytics. And don’t fall into the trap of thinking issues like “Not Provided Keywords” are Google Analytics issues – they affect all analytics tools. Watch out for the launch of Universal Analytics by Google this year and the benefits that will bring for analysis across devices and online/offline channels.
4. Social is Just a Channel
There is still an overriding sense that social effectiveness is somehow uniquely impossible to measure. It’s not, but requires discipline around getting meaningful data: track every social share a user makes from your site, every viral inbound click you get from those shares, every link you push out through and between networks. Use attribution to understand how it complements other marketing channels.
5. Mobile is Just a Device
Define “mobile”: screen-size, bandwidth, multi-touch, location, apps versus responsive websites? Is a mainframe in a caravan a “mobile” device? The technical capability of your audience should always be a key dimension for comparative analysis – it often exposes where performance gaps exist for subsets of users. The multiple device issue has always existed (anyone ever visited a website on their home and work PC); the proliferation of mobile is just making it more profound.
6. Privacy is not about cookies
The debate was never really supposed to be about cookies… so watch out for the real issues here. Apple, Facebook, Google et al all want your data, and don’t want you to have their data, and don’t want each other to have each other’s data! That’s bad news for consumers and analysts alike, and a plethora of cookie warnings across UK sites is at best a distraction. Think carefully about what data you capture and share, why you do it and what your customers would say.
7. Errors are “good?”
Everyone wants “actionable analytics”. Here’s the easiest (and most commonly missed) one: if it’s worth telling your users that their email address didn’t match, or their postcode was not in the designated format, or it was compulsory to indicate how you found the site… then it’s worth capturing that error occurrence data in your analytics tool. Watch it like a hawk and segment it by device (anybody ever had frustrations completing a form on a smartphone?).
8. Integration starts with you
Cross-channel measurement is tricky because it spans people, targets and sometimes even P&L buckets. Don’t even think of investing in technical solutions until you have an aligned focus, common objectives and consistent cross-channel KPIs. Then think tactical integrations where the sum of specific data is greater than its parts rather than trying to bung everything together into one magic data warehouse.
9. Be bold with testing
Digital opens up a massive opportunity for testing propositions and messaging, and the businesses that lead their sectors are the ones that invest in continuous test and learn. But don’t fall into the trap of only testing things you know the answer to already (“told you so”), or minor but overly contentious points (“which shade of red should the button be?”). If you are doing test and learn properly, you need your tests to fail sometimes, otherwise there’s no learn!
10. Use experts wisely
Kill a pointless report and you might save an analyst. Invest a day in ensuring your data collection is accurate and you could save a year of chasing your tail. Get your objectives, KPIs and requirements successfully aligned from the start and you could save thousands on misdirected technology.
Find a Web Analytics Wednesday near you and share your challenges with other like-minded souls.
For analytics expertise and resource across tools, sectors and channels, tap into our 8 years of experience with clients including HSBC, Experian, Viacom, Canon, Schuh, Virgin Holidays and Cartoon Network. Visit our website to find out more: www.lynchpin.com
Browse the sections to read articles prepared by thought leaders in all the key marketing disciplines. If you would like to deposit a Knowledge Bank report contact firstname.lastname@example.org