5 common Google Analytics setup problems
Can you rely on the data you are seeing in Google Analytics? If you use it daily in your business you should really give some time to auditing how the data is captured, and what glitches could be lurking unseen. The notifications feature in Google Analytics now alerts you to some common setup problems, but there are more simple ones you could check today. Here are 5 aspects of your Google Analytics account to check now. Are you running the latest Universal Analytics tracking code? Is your overall bounce rate below 10%? Are you getting referrals from your own website? Are you getting ‘referrals’ from your payment gateway? Have you got the correct website default URL set in GA? Are you getting full referring URL in reports? 1. Are you running the latest Universal Analytics tracking code? You may have clicked upgrade in the Google Analytics admin console, but have your developers successfully transferred over to the new tracker code? Use our handy tool to test for universal analytics (make sure you copy your URL as it appears in the browser bar). 2. Is your overall bounce rate below 10%? The 'bounce rate' is defined as sessions of only one page. It’s highly unlikely to be in single digits unless you have a very unique source of engaged traffic. However, it is possible that the tracking code is firing twice on a single page. This double counting would mean Google Analytics sees every single page view as two pages – i.e. not a bounce This is more common on template-driven sites like Wordpress or Joomla, where you may have one tracking script loaded by a plugin – and another pasted onto the main template page. You can check if you have multiple pageviews firing by using the Google Tag Assistant plugin for Chrome. 3. Are you getting referrals from your own website? A self-referral is traffic coming from your own domain – so if you are www.acme.com, then a self-referrals would be appearing as ‘acme.com’. Have a look at the (recently moved) referrals list and see if that is happening for you. This is usually caused by having pages on your website which are missing the GA tracking code, or have it misconfigured. You can see exactly which pages are causing the problem by clicking on your domain name in the list and seeing the referring path. If you are on universal analytics (please use our tool to check) you can exclude these referrals in one step with the Referral Exclusion list. For a fuller explanation, see the self-referral guide provided by Google. 4. Are you getting ‘referrals’ from your payment gateway? Similar to point 3: if you have a 3rd party payment service where customers enter their payment details, after they redirect to your site – if you are on Universal analytics – they will show up as a new visit… but originating from ‘paypal.com’ or ‘worldpay.com’. You need to add any payment gateway or similar 3rd party services to that referral exclusion list. Just add the domain name - so PayPal would be 'paypal.com' 5. Have you got the correct website default URL set in GA? When Google Analytics was first set up for your website you may have set a different domain name than what you now use. Or maybe you have switched to run your site on https:// rather than http://. So you need to change the default URL as set up in the admin page. For this go to Admin > Property > Property Settings. Once that is setup correctly, the ‘All Pages’ report becomes a lot more useful – because you can click through to view the actual page using the open link icon. Advanced: Are you getting full referring URL in reports? If you run your website across different subdomains (e.g. blog.littledata.co.uk and www.littledata.co.uk) then it can be difficult to tell which subdomain the page was on. The solution to this is to add the hostname to the URL using a custom filter. See the guide on how to view full page URLs in reports. What other setup issues are you experiencing? Let us know in the comments or by tweeting @LittledataUK.
Best enhanced ecommerce plugins for Magento
Agriculture in Uganda: Measure and Improve
I had a truly inspiring day visiting Send a Cow project near Masaka in Uganda. A group of 30 farmers underwent 4 years of training, supported by weekly visits from a social worker and agricultural trainer. From a group living in absolute under-a-dollar-a-day poverty, there are now farmers owning thousands of dollars worth of livestock and selling export crops like coffee. This education and support, plus the capital grant of one animal per household, has transformed their community. Although the success relied on a solid base of family and group cohesion, organised labour and animal husbandry, I want to focus on three aspects which have ongoing potential for the community. 1. Record keeping Yep, data to you and I. Writing daily details of milk yields, crop inputs, market sale prices and even visitor numbers enabled the farmers to measure and improve. Data also allows farmers to forecast and be inspired. Selling a regular surplus of milk from two cows (after family consumption – yes, they have great teeth!) gave the farmer a regular income of US$3.50 per day at the farm gate. That is more than a teacher’s salary in Uganda. With tender care and back-breaking forage harvesting, they now have a calf being reared – and can count just how much that will mean in further milk and profits. Maybe in 10 years they will be entering yields into a smartphone app, and have market prices forecast automatically. 2. Organic agriculture Oil derivatives (like diesel and fertiliser) are nearly as expensive in Uganda as the UK – in ridiculous contrast to the local market prices for vegetables. Efficient farming therefore has to rely on minimal imported inputs, and maximise the local bounty of sun, rain … and manure. Every precious drop of animal urine is captured – to mix with ash and chilli as an insect repellant for plants – or used neat as a fertiliser. In dry season, every rainfall is maximised, with lots of mulching of vegetables to prevent evaporation; and with a permaculture approach of shading coffee bushes with banana plants, and vegetables under the coffee. I am a fan of organic farming for health and environmental reasons, but out here I just do not see an alternative, cost-effective way to increase crop yields. 3. Peer-to-peer lending Developed-to-developing country lending networks, like Kiva.org, have grown rapidly – but with inevitable problems in vetting funding applications at distance. What farmers need are equivalents of 19th century Europe’s co-operative societies – where savers and lenders from the same area are brought together. These farmer groups operate a very effective local system. All members pledge to save every month: from just 1 cent a week. Then any member can ask for a short term (maximum 3 month) loan from the fund – which is now $2000. The default rate is low – around 2% - as members know the debtors ability to repay, and can monitor progress in person. Plus every debtor has savings in the scheme – so wants to preserve their share of the capital. Three month loans (and flat 10% interest) make repayments easy to predict – and work in a country where planting to harvest is only 3 months. Uganda’s government abolished co-operatives in the 1990s when they started sponsoring political campaigns. But if these lending clubs can grow they could go some way to unlocking the capital that Africa needs to grow. This post was written by Edward Upton, Founder of Littledata, @eUpton
Under the hood of Littledata
Littledata tool gives you insight into your customers' behaviour online. We look through hundreds of Google Analytics metrics and trends to give you summarised reports, alerts on significant changes, customised tips and benchmarks against competitor sites. This guide explains how we generate your reports and provide actionable analytics. 1. You authorise our app to access your Google Analytics data As a Google Analytics user you will already be sending data to Google every time someone interacts with your website or app. Google Analytics provides an API where our app can query this underlying data and provide summary reports in our own style. But you are only granting us READ access, so there is no possibility that any data or settings in your Google Analytics will change. 2. You pick which view to report on Once you've authorised the access, you pick which Google Analytics view you want to get the reports on. Some people will have multiple views (previously called ‘profiles’) set up for a particular website. They might have subtly different data – for example, one excludes traffic from company offices – so pick the most appropriate one for management reports. We will then ask for your email so we know where to send future alerts to. 3. Every day we look for significant changes and trending pages There are over 100 Google Analytics reports and our clever algorithms scan through all of them to find the most interesting changes to highlight. For all but the largest businesses, day-by-day comparisons are the most appropriate way of spotting changing behaviour on your website. Every morning (around 4am local time) our app fetches your traffic data from the previous day – broken down into relevant segments, like mobile traffic from organic search – and compares it against a pattern from the previous week. This isn’t just signalling whether a metric has changed – web traffic is unpredictable and changes every day (scientists call this ‘noise’). We are looking for how likely that yesterday’s value was out of line with the recent pattern. We express this as signal bars in the app: one bar means there is a 90% chance this result is significant (not chance), two bars means a 99% chance and three bars means 99.9% certain (less than a 1 in 1000 chance it is a fluke). Separately, we look for which individual pages are trending – based on the same probabilistic approach. Mostly this is change in overall views of the page, but sometimes in entrances or bounce rate. If you are not seeing screenshots for particular pages there are a few reasons why: The website URL you entered in Google Analytics may be out of date Your tracking code may run across a number of URLs – e.g. company.com and blog.company.com – and you don’t specify which in Google Analytics The page may be inaccessible to our app – typically because a person needs to login to see it 4. We look for common setup issues The tracking code that you (or your developers) copy and pasted from Google Analytics into your website is only the very basic setup. Tracking custom events and fixing issues like cross-domain tracking and spam referrals can give you more accurate data – and more useful reports from us. Littledata offers setup and consultancy to improve your data collection, or to do further manual audit. This is especially relevant if you are upgrading to Universal Analytics or planning a major site redesign. 5. We email the most significant changes to you Every day - but only if you have significant changes - we generate a summary email, with the highest priority reports you should look at. You can click through on any of these to see a mobile-friendly summary. An example change might be that 'Bounce rate from natural search traffic is down by 8% yesterday'. If you usually get a consistent bounce rate for natural / organic search traffic, and one day that changes, then it should be interesting to investigate why. If you want your colleagues to stay on top of these changes you can add them to the distribution list, or change the frequency of the emails in My Subscriptions. 6. Every Sunday we look for changes over the previous week Every week we look for longer-term trends – which are only visible when comparing the last week with the previous week. You should get more alerts on a Sunday. If you have a site with under 10,000 visits a month, you are likely to see more changes week-by-week than day-by-day. To check the setup of your reports, login to Littledata tool. For any further questions, please feel free to leave a comment below, contact us via phone or email, or send us a tweet @LittledataUK.
What's new in Google Analytics 2014
Google has really upped the pace of feature releases on Analytics and Tag Manager in 2014, and we’re betting you may have missed some of the extra functionality that’s been added. In the last 3 months alone we’ve counted 11 major new features. How many have you tried out? Official iPhone app. Monitor your Google Analytics on the go. Set up brand keywords. Separate out branded from non-brand search in reports. Enhanced Ecommerce reporting. Show ecommerce conversion funnels when you tag product and checkout pages. Page Analytics Chrome plugin. Get analytics for a particular page, to replace old in-page analytics. However, it doesn’t work if you are signed into multiple Google Accounts. Notifications about property setup. Troubleshoot common problems like domain mis-matches. Embeded Reports API. So you can build custom dashboards outside of GA quickly. Share tools across GA accounts. Now you can share filters, channel groupings, annotations etc easily between views and properties Tag Assistant Chrome plugin. Easily spot common setup problems on your pages using the Tag Assistant. Built-in user tracking. See our customer tracking guide for the pros and cons. Import historic campaign cost and CRM data (premium only). Previously, imported data would only show up for events added after the data import. Now you can enter a ‘Query Time’ to apply to past events, but only for Premium users. Get unsampled API data (Premium only - developers). Export all your historic data without restrictions Better Management API (for developers). Set up filters, Adwords links and user access programmatically across many accounts. Useful for large companies or agencies with hundreds of web properties.
Pulling Google Analytics into Google Docs - automated template driven reporting
The Google Docs library for the analytics API provides a great tool for managing complex or repetitive reporting requirements, but it can be tricky to use. It would be great if it was a simple as dropping a spreadsheet formula on a page, but Google’s library stops a few steps short of that - it needs some script around it. This sheet closes that gap, providing a framework for template driven analytics reports in Google Docs. With it you can set up a report template, and click a menu to populate it with your analytics results and run your calculations - without needing to write a line of script - the code is there if you want to build on it, but you can get useful reports without writing a line of script. Prerequisites While you don't have to write code to use this, there are some technical requirements. To get the most out of it you'll need to have: your Google analytics tagging and views set up familiarity with Google’s reporting API familiarity with Google Docs spreadsheets - some knowledge of Google apps scripting is an advantage If you are looking for something more user-friendly or tailored to your needs, contact us and book a consultation to discuss - we can help with your analytics setup and bespoke reporting solutions. Getting started Setting this up takes a few steps, but you only need to do this once: Open the shared Google spreadsheet Make a copy Enter a view ID in the settings sheet - get this from the Google Analytics admin page. Authorise the script Authorise the API - in the API console - this is the only time you need to go into the script view using Tools|Script Editor Once in script editor select Resources|Advanced Google services On the bottom of the Advanced Google services dialogue is a link to the Google Developers Console, follow this and ensure that Google analytcs API is set to On You're done. You can go back to the spreadsheet and run the report (on the Analytics menu). From now on all you need to do is tweak any settings on the template and run the report. Setting up your own report template You can explore how the template works using the example. Anywhere you want to retrieve value(s) from Google Analytics, place this spreadsheet function on the template: = templateShowMetric(profile, metric, startdate, enddate, dimensions, segment, filters, sort, maxresults) This works as a custom spreadsheet function, for example =templateShowMetric(Settings!$B$2,$B7,Settings!$B$3,Settings!$B$4,$C7,$D7,$E7,$F7,$G7) Note that in the example, several of the references are to the settings sheet, but they don't have to be, you can use any cell or literal value in the formula - it's just a spreadsheet function. To get the values for the API query, I'd suggest using Google’s query explorer. To set this up for a weekly report, say, you would have all the queries reference a single pair of cells with start and end dates. Each week you would change the date cells run the report again - all queries will be run exactly as before, but for the new dates. Using spreadsheet references for query parameters is key. This opens up use of relative and absolute references - for example if you need to run the same query against 50 segments, you list your segments down a column, set up segment as a relative reference, and copy the formula down spreadsheet style. You can use this to do calculations on the sheet and use results in the analytics API, for example you might calculate start and end dates relative to current date. Future posts will cover setting up templates in more detail. Under the hood The templateShowMetric function generates a JSON string. When you trigger the script, the report generator copies everything on the template to the report sheet and: runs any analytics queries specified by a templateShowMetric function removes any formulas that reference the settings sheet (so you can use the settings sheet to pass values to the template, but your reports are not dependant on the settings staying the same)
Analytics showing wrong numbers for yesterday's visits
We've noticed a few issues with clients using Universal Analytics this last month, when visits for the last day have been double the normal trend. It then corrects itself a few hours later - so seems to be just a blip with the data processing at Google. Others have noticed the same problem. The temporary fix is to only generate reports with time series ending the day before yesterday. i.e. ignore yesterday's data. Now Google have officially acknowledged the problem Looking forward to seeing that one fixed!
Measuring screen resolution versus viewport size
There’s a difference between the ‘screen size’ measured as standard in Google Analytics and the ‘browser size’ or ‘browser viewport’. Especially on mobile devices, there are pitfalls comparing the two. Browser viewport is the actual visible area of the HTML, after the width of scroll bars and height of button, address, plugin and status bars has been allowed for. Desktop computer screens have got much bigger over the last decade, but browser viewports (the visible area within the browser window) are not. The CSS tricks site found only 1% of users have their browser viewing in the full screen. While only 9% of visitors to his site had a monitor less than 1200px wide in 2011, around 21% of users have a browser viewport of less than that width. Simply put, on a huge monitor you don’t browse the web using your full screen. Therefore, 'screen resolution' may be much larger than 'viewport size'. The best solution is to post browser viewport size to GA as a custom dimension. P.S. Google Analytics does have a feature within In Page Analytics (under Behaviour section) to overlay Browser Size, but it doesn’t work for any of the sites I look at.
Subscribe to our blog
Get the latest posts in your email
Get the Littledata analytics app
Complete picture of your ecommerce business. Free Google Analytics connection, audit and benchmarks.Sign up