Your first question should be extra to what? It’s always great to get five Google Analytics tips, tricks & hacks, its even better to get more of them so why are these extra? Well I am speaking at the ObservePoint Analytics Summit this Thurs 17th Nov and there I am giving the first 25 Google Analytics tips, tricks and hacks. These are extra to that list.
Quick little bit of background. My original plan was to give 10 tips, tricks & hacks in 30 min. But then I decided that wasn’t a big enough challenge. I wanted to take it further, aim for one per minute and get up to 30. It appears I was too ambitious, during recording it became apparent that in order to focus on quality, I needed to cut back. So these are the five that didn’t make the final cut.
Oh yeah, taping, another piece of explanation required. The Analytics Summit is an entirely online conference, all the sessions have been pre-recorded. Not sure about everyone else but I was pretty good… Even better news, the entire conference is free to attend. Register via Analytics Summit and listen to not just myself but other analytics experts from around the world. Did I mention it was free???
The five tips include two related to Enhanced Ecommerce, a use case for two vital pieces of customer information to capture, a general tracking tip to make information more useful and a plea to invest time to understand Google Analytics definitions so you can understand the data you are looking at.
At the end of last year I attended the iLive Conference in Riga where I talked about how to use analytics to improve a business’ bottom line.
Overall, the iLive conference was a great experience. I was really impressed with the quality of presentations and the level of digital expertise shared across a wide range of digital channels, including SEO, analytics, social media, community building and brand development.
iLive Conference 2015 Presentation: Impacting Business Performance with Analytics
The issue for businesses today is not that they don’t have the technology for analytics, it is they don’t know how to harness it. When the potential rewards are as big as they are, this is not acceptable. As shared at the iLive Conference, there are a range of techniques and approaches to make the benefits of analytics accessible and achievable for businesses of all sizes and in any sector.
Over the coming weeks I plan to share the tips discussed at the event in the form of a blog series. The aim here is to help you translate your overwhelming mess of data from within your analytics tool into digestible, actionable insights which will in turn help your business’ digital performance.
Topics covered in this blog series will include:
Analytics Set Up Tips
Analysis & Insights Tips
Company & Process Tips
Interested in Learning More?
Please let me know if you have any questions or would like samples/templates of any of the reports described within the presentation. Of course, if this does inspire you to start using analytics to impact the performance of your business and you need some expert help, please get in touch on 07843 617347 or via email@example.com.
I was kindly invited by Nicolas Malo to speak at the Lille Web Analytics Camp on last Wed 21st Mar, presenting as part of the English stream of talks. It was a great experience (a bit challenging for me as probably the only native English speaker in attendance) but great to see so many people happy to give up an afternoon to talk and learn about web analytics. My biggest disappointment was not being able to stay long for the informal discussions that night, I should have booked Eurostar for the next morning.
When Nicolas first asked me to speak, I threw a couple of ideas for topics at him. His preference was for a talk on Getting Practical Value from Web Analytics throughout the Organisation. This is the sort of topic I find myself returning to more and more when I talk to people, web analytics is not about data but about making use of data within an organisation. Below is a copy of my presentation and, as it is a fairly visual presentation, some brief notes on the ideas behind each slide. It can also be downloaded here – Getting Value from Web Analytics throughout an Organisation.
I wanted to repeat some of the lines probably every web analyst has encountered when introducing web analytics into a company. Some of my favorites are:
You will say (prove) I am not doing a good job
My job is about creativity, not numbers
I don’t have time for this
The last thing I need is more reports
I know our customers & the market better than you (and your data)
Purpose of Web Analytics
It is always worth remembering why companies invest in web analytics and what it is all about. It is not about numbers, statistics and reports. It is about providing the intelligence to allow better decision making leading to organisations improving their performance.
Must have Quality of Data
Why – so internal stakeholders have confidence in the numbers and so you can provide them with answers to their questions
How – ensure all pages & campaigns are tagged and that you are capturing key visitor interactions & key meta-data
You need to have the foundation of web analytics in place before you can build on it.
Build Internal Relationships
Why – so that people tell you when they make changes that will impact performance (the numbers you are looking at) and so they come to you for assistance with their business questions
How – depends on personal approach – walk about the building talking to people, buy them lunch, have a beer with them after work, invite yourself along to every meeting you can find out about
You do need to know (be on first name terms with) people from all departments throughout the organisation – and that doesn’t mean just marketing and development but also product management, user experience, finance, customer service, operations, etc. They all have answers and questions.
Help people to like Web Analytics
Why – web analytics can be very overwhelming with so many reports all full of numbers meaning most people have no idea where to start and so don’t
How – provide them with training and reference guides, run regular workshops where they can ask questions and very importantly, give them dashboards so they can look at a subset of the data all in the one place
Once people get used to web analytics, they discover that it is not so scary after all and that it can provide them with lots of useful information – ease them into this new world of data (preferably without using the word data or showing them much data).
Integrate Web Analytics everywhere
Why – the set-up of web analytics and the use of intelligence from web analytics to improve performance are not one off endeavours, they must be continuously worked on to be effective – this requires web analytics to be integrated within the existing processes of the organisation
How – processes and templates – as simple as possible but documentation is a necessary evil and web analytics must be a consideration when creating any new marketing campaigns, pages, website features, etc.
If it is useful to know the impact of a new campaign/website feature, you need to invest resources into defining business objectives, KPIs, targets, into adding the necessary web analytics tags and into evaluating performance.
Prove Web Analytics works
Why – you need proven results before other people will believe in web analytics (just as the idea of web analytics is to believe in data over opinion)
How – find someone who will give you a chance, use web analytics to improve performance and then create internal case studies
Once it is proven that using web analytics data will lead to improvements in performance, it becomes a question of can you afford to not take advantage of this resource rather than do you have the time to do so.
Reminder of Key Points
Get the foundation accurate and complete ASAP
Make friends with people in all departments
So they tell you what they are doing
So they come to you with their business intelligence needs
Support people in getting used to web analytics
Provide training sessions and Q&A workshops
Give them dashboards with a reduced data set
Develop templates & processes for web analytics
Create case studies to prove Web Analytics can and does deliver against its promises
Through the support of Chinwag, we have being able to gain access to a great venue in central London – 01Zero-One. The doors will be opening at 6pm with presentations starting an hour later. There will be four talks of 10min each then some time for Q&A. The venue is available for us till 9pm for discussions, networking and more drinks, we can also retire to a local pub if the discussions get really interesting.
The presentations will be on four different sources of data from social media. These are:
Your own brand pages on social media networks
Discussions regarding your brand on social media networks
The influence of individuals within social media networks
The impact of social media of your own website
All talks will be tool agnostic, they are intended to give you an overview of data available within each area, what sort of metrics you should be focusing on and examples of ways that you can use the intelligence from this data to improve performance.
** Update – WebTrends has offered to come on board as a sponsor of the event and they will be providing refreshments for the evening. They have a strong interest in the analytics of social media, offering an integration between the leading social media networks and their own web analytics tool. In addition, WebTrends offers a free tool for understanding your Facebook pages called Hoverstats. **
The speakers will be kept under control by Cathy Ma who is the Head of Social Media at IPC Media. She is an experienced user of social media and in her role requires much of the data and intelligence that will be discussed during the presentations.
Joshua March is the co-founder and CEO of Conversocial. His presentation will centre on the metrics available from your brand pages on social media networks. These metrics provide valuable insights into your use of social media and Joshua will provide examples of these insights for companies who use social media for customer service.
As previously reported in the Sun, Christian Howes has the ability to predict Big Brother evictions using the data from Social Media conversations. He will be speaking about these and other insights available through monitoring conversations on social media networks.
Simon Cast is the Head of Products from PeerIndex and a co-founder of ProductCamp London. He will be providing details on metrics that are available to determine influence within Social Media. The key question everyone will want the answer to of course, is how to become more influential themselves.
I will also be presenting some ideas on how to measure the impact of social media on your own website. This will cover tagging requirements but will also include approaches for understanding both the direct and indirect impact of social media campaigns.
Please come along and join in the discussions as we all learn more about the insights possible from Social Media data. Drinks and nibbles will be supplied (sponsor to be confirmed, please get in touch if interested) so there really is no excuse. Except not getting in fast enough to get a ticket, we already have over 50 registrations meaning there are not many left. Please register via the event page on the Social Media Week website.
I was given a speaking slot at Conversion Thursday London last night and, while I have spoken at various events previously, I think this was the first time presenting to my peers in the web analytics community. So thank you Rob & Elisa DBI for the opportunity. And of course thanks to Epiphany Solutions for sponsoring the event.
With vague guidelines of 10 minutes and talking about something relevant and interesting, I decided to practice what I have preached about other presentations and provide practical tips on applying Web Analytics. With that time limit, I restricted myself to five tips and keeping them brief and to the point. While they had to be something that could be applied immediately, all tips were tool agnostic with examples provided using my primary tools, SiteCatalyst and Google Analytics.
Most of the tips have been covered elsewhere in previous blogs but this should prove to be a useful reference guide. Of course it is intended to be incredibly useful if you have not previously read those posts. I believe the talk went fairly well and I managed to not speak too quickly. I do recommend viewing the presentation in full screen mode in order to see the details on the screenshots.
The key outcome for me was I reached my target KPI with over 75% of the audience raising their hands at the end to say they had received at least one useful tip. Unfortunately the SiteCatalyst Campaign URL & SAINT Builder is not yet ready and will be released early next week.
As ever, let me know if you have any questions or would like to discuss any of the tips I provided. And I am looking forward to future speaking opportunities so get in contact if you would like someone to present on topics around the purpose and use of web analytics.
My favourite session at the Adobe Omniture Summit EMEA 2011 was a panel based session for power users from Adam Greco, Jan Exner and Brett Dykes. It was full of tips, tricks and ideas for extending the capability of SiteCatalyst and getting more value from it. They explained potential approaches to resolving ten predefined questions and then took questions from the floor.
I didn’t always record the exact questions and answers, instead jotting down thoughts which were inspired by what was said. The following then is a mixture of the ideas presented and what I took from them (apologies for any mistakes made in the translation). For more ideas from the summit, check out my post on Experiences from the Adobe Omniture Summit EMEA 2011.
Include price in “add to cart” events
When recording an “add to cart” event, Adam recommended including the price of that product within the measurement. This enables calculations around the value of abandoned shopping carts. I wonder if this could be used to give a weighting to the Product Page Success Rate within the Merchandising report used to evaluate your most important products.
Use Counter eVars to understand behaviour pre conversion
A variable I wasn’t aware of are Counter eVars. It allows you to count the number of times an event occurs before a conversion. Examples of this include:
the number of product views before a cart is created
the number of internal searches before a result is selected
the number of articles viewed before subscribing
The process for a successful implementation of SiteCatalyst
While it might seem obvious, the next piece of advice was to take an implementation or re-implementation seriously or web analytics will fail within your organisation. There are three elements to doing this properly:
Plan – the process you will take during the implementation
Talk – to all key stakeholders about their business/reporting requirements (what they need to know)
Document – the business/reporting requirements, what is being tagged & configured
Matching up online and offline data
Use the Transaction ID variable to match up online data with offline data that you load into SiteCatalyst at a later date. I knew that you needed a unique identifier/key to match online & offline data but didn’t realise there was a specific variable available for this.
Streamline the SiteCatalyst UI for non-analyst users
Streamline SiteCatalyst through:
renaming the menu items to make more relevant to your business
hiding reports that are not relevant
reordering the reports so that users can easily find the reports they need
creating custom reports
writing instructional notes for each reports
I recently worked with a client where the campaign section of the menu was completely reworked with custom reports and labelled for different stakeholder groups. Putting this work in can be the difference between SiteCatalyst being used by non analysts or being just another expensive reporting tool.
Create custom dashboards with the API
Use the API to build custom dashboards – this does only apply to online dashboards or those ones you see on big screens in the office. For Excel dashboards, use ReportBuilder (will work nicely with my free Excel Dashboard Templates).
Success Event Pathing
An idea Adam had previously written about is to capture all events as a sProp. He suggests you then turn on pathing for that sProp and you can do Success Event Pathing.
I was quite annoyed when I first read this this post as I had been planning to make the same suggestion although for a different reason. I love being able to look at the % of visits in which an event occurred and this allows you to do so. It can even give you an approximate funnel if you are tracking the key stages as events.
Evaluating the success of content
There were some ideas around how to evaluate the success of content including:
Applying an eVar to capture the last page viewed prior to a success event
Applying an eVar to capture the second last page viewed prior to a success event
Tracking and reporting on Internal Search
Internal search is one of the fundamental things that should be captured in any web analytics implementation and some ideas around doing so with SiteCatalyst were mentioned here:
Use an event to capture use of internal search
Use an eVar to capture the search term used
Use an event for the visitor clicking on a search term result
Create calculated metrics to evaluate the performance of search terms
It was not mentioned but you should also capture internal search categories (as per Google Analytics)
Also not mentioned is capturing the number of results, especially so you can identify search terms with zero results (as per Sitestat/Digital Analytix)
I do continue to be surprised that some specific Internal Search code and relevant reports have not been developed for SiteCatalyst instead of having to use traffic and conversion variables.
Measuring Visitor Engagement
It got really advanced at the end but counter eVars can be used to give points to visitor behaviour to calculate a measure of visitor engagement. I wrote down something about an incrementor event which I think is an alternative way of capturing this information, allowing you to compare the performance of other dimensions such as traffic source based on their level of visitor engagement.
Pathing for products
A bonus idea thrown out there was to enable pathing on products, allowing you to view how visitors navigate through the different products on a website (although I am not sure how this information could be used).
Like many sectors, any serious business working in real estate has a website these days. And like any business operating in the online world, the intelligence from web analytics can be used to improve business performance. There are two main types of real estate websites, the web presence of the real estate agents and websites that aggregate the content across multiple real estate agents. For the purposes of this, I am only covering real estate agency websites.
Outcomes and online behaviour for optimisation
Real estate agency websites exist as advertising mediums for these businesses. They are the modern equivalent of shop windows containing properties for sale or rental. However they go far beyond the real estate window or newspaper in terms of available content. They contain photos, floorplans, virtual tours, details of the local area, reviews of local school and other valuable information that influences the selection of a property. The websites also advertise the real estate agents, proving how well they publicise their properties and detailing their other services.
The primary desired outcome from any website visit is contact made with an office via a contact form, email, phone or in person. The reason for the contact can vary depending on whether the client is looking to rent/buy a property, to sell/rent out their own property or for one of the other services offered by the real estate agency.
Suggestions for applying Web Analytics
While the desired outcome is contact made with the office, there are certain website actions that contribute to this desired outcome. Improving the performance of these micro conversion actions will lead to an increase in the number of contacts made. They should be tracked through the web analytics tool and insights gained from the data used to make improvements. Below are some examples of key visitor behaviour and suggestions for how their performance can be measured with a view to improving this performance.
The property search is the heart of any real estate website. When working optimally, it enables a visitor to find all properties relevant to them with a minimal number of clicks. Real estate agency websites may offer different methods of searching for properties with some ideas here for evaluating their performance:
Which type of search is most popular?
Searches by search type
How did the visitor access the search?
Previous page viewed by search type
Clicks on navigation options to select search type
Does the search return relevant results?
Search refinements (less is better)
Search results success rate (clicks on search results / views of search result pages)
Is the search easy to use?
visits in which search was performed / visits in which search page was viewed
Which type of search is most effective?
Search conversion rate based on contacts made with office regarding properties found through using that search type
Real estate agencies can include different types of information about a property on its website but there are costs involved. Given this investment it would be nice to know:
Do visitors consider this information relevant?
Views of each type of content
Does the content influence a prospect to contact the office?
Contact made on properties where content type was viewed / properties where content type was viewed
What do visitors consider to be the most useful content?
Order in which content is viewed
Do visitors understand the content or is there too much jargon?
Clicks on help or jargon busters by content type
Methods of contacting agency
This is the key outcome for the real estate agency. It is definitely not the point where you can risk losing a prospect due to website issues. So some key questions are:
Are the calls to action optimised?
Are they above the fold?
What is the impact of changing the wording or colour?
Are people completing enquiry forms?
Do they require too much information?
Does validation of the forms prevent some people from submitting them?
Do they work equally well in all browsers?
Is the form generic or partially pre-populated based on visitor behaviour?
Are contacts made via the phone counted as conversions?
Are you using unique phone numbers?
Do the calls appear as an action in your web analytics tool?
Optimise Marketing Spend
Real estate agencies can invest in various online marketing channels including SEO, Paid Search, Aggregators and Email let alone all the offline marketing that they do. Optimising marketing spend has twin targets of reducing spend and increasing leads generated. Areas to look at include:
Do all the online marketing links contain web analytics campaign tags?
Does offline marketing use vanity URLs so it too can be identified?
Do all marketing channels deliver a positive ROI?
The calculation might by “applying average revenue to number of leads generated / cost of marketing” but which attribution model?
Can you differentiate between marketing that generates research visits and marketing that generates leads?
Is the agency paying for existing customers (i.e. visitors are simply using the marketing to access the website)?
Is initial website experience matching the expectation created by the marketing?
Usually measured by the Bounce Rate
Do you need to create customised landing pages?
People registered with agency
Once someone has signed up with the real estate agency, they can be sent details of new relevant properties via channels like text or email. With the hard work of acquiring a prospect done, is this tool helpful in converting them into customers:
Can the prospects easily find the properties they are sent details of?
Click through to property directly or search by property ID
Are they tracked in the web analytics tool as Alert visitors?
Do they have a different website experience compared to new visitors or is it exactly the same content?
You know all they want to do is find relevant properties for viewings so why display any content not relevant to this?
Is the call to action more relevant to the stage of the process they have reached?
If they choose a property for a viewing, does this become a one click experience as the agency already has all their relevant details?
While not a website action, it is worth noting that real estate agencies are sitting on a wealth of valuable information. Visitors to their websites are telling the agencies exactly what they are interested in through the searches they are making and the filters they apply. Real estate agencies can use this data to identify key trends in the market, what matters to people looking for a new property, what areas they are looking in, what price range is appropriate, etc
This information can be captured using custom variables (whatever the web analytics tool) on the type of filter and the value selected. It is likely the data will need to be extracted via API and then manipulated to get at the insights.
Those are a few ideas but it should be sufficient to demonstrate how real estate agencies, like any online business, could derive immense benefits and improvement in performance through the use of web analytics. Not in reporting on the number of people using the website and what pages they are viewing but in making smarter business decisions. Their websites and marketing can be optimised with a new source of intelligence available to add to their knowledge of the market.
Web analytics may not be receiving much attention right now from real estate agencies but what would 20% more leads generated through their website be worth??
The top five UK real estate agencies based on the number of UK visits in Feb ’11 are (data from DoubleClick AdPlanner and hopefully I didn’t miss any of the top 5):
Peter kindly asked me to share a tip or trick that I’ve found particularly helpful or useful when working with analytics. Unfortunately I’ve failed – there are just too many of the little suckers to put one above the rest. Instead I’ve put together some rough and ready advice that I‘ve mostly learnt from doing it the wrong way (please take it with a hefty pinch of salt):
If you’re analysing data but can’t change anything as a result, you’re wasting your time. Stop. Do something else or find another job.
If you’re just producing reports but don’t know what they get used for, you’re probably wasting your time. Stop. Get out of your chair and go to where the reports go. Talk to the people who use them. Make them better or bin them (the reports, not the people).
Agree success metrics before you undertake a project. Then measure them. Seriously. It’s surprising how often agreed metrics can’t actually be measured.
When you finish a project go back to the success metrics. Did you get there? Was it the right thing to go for? Have you done what you set out to achieve?
Capture usable data. There’s no point in collecting it if you can’t do anything with it.
Forget about absolute values. Think in terms of trends, correlations, differences and changes.
Assess the validity of your stats in context. If they feel wrong, they probably are. Triple check everything.
Don’t get too busy implementing analytics solutions to look at data you’ve already got. Stop. Ask yourself what you can look at now to improve things. Performance improvement is the goal, not perfect data.
Don’t allow data to suffocate design decisions with a perpetual quest for justification. <cough>Google</cough>.
Develop hypotheses, design solutions to test the hypotheses, put them into experiments and analyse the results. Repeat. It’s called iteration. It works. Do it continually.
If you only do phase one and never get to phase two, reread point 10.
If the above two points don’t make sense, Google “agile development”.
When developing hypothesis and designing solutions, don’t be precious about job roles. Give everyone a Sharpie and start sketching and talking.
The word hypothesis is just another word for idea, but it sounds more smarterer.
Understand statistical significance and sample size.
Don’t always trust statistical significance and sample size. In split tests, if you’ve got the traffic volumes, use two duplicate control groups to benchmark against each other.
Use A|B tests to test radically different solutions until you get an optimal base to build from. Then continue to optimizing with MVT.
If you haven’t done A|B tests on something, don’t do MVT. It’s intensive, expensive and will distract you from the radical changes.
Understand natural variation, standard deviation and control limits.
Spend a disproportionately large amount of time thinking about how to present your discoveries. Rarely can one person make significant changes to a site single-handedly. You need other people to understand the problems so they can help with the solutions.
Continually work on your influencing skills.
Read Tufte. Understand his ‘fundamental principles of analytical design’.
Surround yourself with meaningful and clever infographics. Push yourself to make your analysis, discoveries and insights clearer. It’s hard, but one picture that clearly illustrates a problem can change a company’s agenda.
Understand the difference between data, information, knowledge, insight, actionable insight and wisdom:
Data: number(s): e.g. visitors = X
Information: a set of numbers: e.g. visitors per day during June
Knowledge: sets of numbers in context: e.g. visitors per day versus visitors per day from search engines during June
Insight: knowledge + awareness: visits per day decreased in June but more traffic came from non brand keywords due to SEO activities on promotional pages
Actionable insight: insight + plan: Rolling out SEO updates to a particular template will increase visitors by X% which should lead to Y extra sales.
Wisdom: actionable insight + experience: Rolling out SEO updates to a particular template will increase visitors by X% which should lead to Y extra sales although it could cause ABC to happen which would mean Z extra sales instead.
Never circulate just data or information. Everything must be in context or people will misinterpret it.
Scrutinize your own insight. Play the devil’s advocate against your own discoveries. Everyone else will.
You can’t know everything. Do all that you can to help decision makers really understand what’s happening so you can both turn it into actionable insight and wisdom. That usually involves more listening than you’d think. I’m still really rubbish at this one.
Geek out with other analytics folk. You’ll learn a lot and you may have quite a bit of fun along the way.
Everyone likes to know if the people visiting their website are seeing it for the first time or are regulars. This is even more important when they are paying for the traffic, if the money is going on acquiring new visitors (potential new customers) or is it just providing a convenient entry point for people who would be coming to the site anyway.
Due to cookie deletion and multiple computer usage, it is difficult to get a true picture of the split between people who have never seen a website before and those who have. However, recording whether the visitor had a cookie from this website previously does at least give an indication of this new/returning split.
What I like to be able to do is to segment out new visitors for a time period (week or month) and examine their behaviour on the website compared to visitors who had visited previously. The new visitor segment should include all visits during that time period by these visitors, not just their initial visit.
Frustratingly, this information is usually not available as default in a web analytics tool unless you can segment at visitor level. However, as long as you have one of the four metrics from New and Returning Visits or Visitors, you can calculate the other three. And most tools will give at least one number. As examples:
Google Analytics gives New Visits and Return Visits
SiteCatalyst provides Return Visits
HBX contains Returning Visitors
The key to this is knowing that the first time a site is visited, that is both a new visit and a new visitor. And as any subsequent visits by these people will be reported as a return visit, the number of new visits equals the number of new visitors.
With that logic in mind, it is simple to calculate all four metrics once you have a single one. For example, assume that the tool available is SiteCatalyst (without access to visitor level segmentation via Data Warehouse or Discover):
The number of Return Visits is available but none of the other three metrics
Total Visits minus Return Visits gives New Visits
New Visits equals New Visitors
Total Unique Visitors minus New Visitors gives Return Visitors
And now it is easy to calculate the proportion of Visits that were New or Returning or to calculate the proportion of Visitors that were New or Returning.
The same principle can be applied to Google Analytics:
New and Returning Visits is available (note that this metric is visits, not visitors as it is titled in the report)
New Visits equals New Visitors
Total Unique Visitors minus New Visitors gives Return Visitors
Of course, these numbers don’t mean that much on their own but do become more useful when trended over time or across different segments.
An interesting thing to look at can be the split in New and Returning Visitors for different time periods – day, week and month. This can indicate the scale of the issue with cookie deletion, but more on that another time.