{{ searchResult.published_at | date:'d MMMM yyyy' }}

Loading ...
Loading ...

Enter a search term such as “mobile analytics” or browse our content using the filters above.

No_results

That’s not only a poor Scrabble score but we also couldn’t find any results matching “”.
Check your spelling or try broadening your search.

Logo_distressed

Sorry about this, there is a problem with our search at the moment.
Please try again later.

According to a recent Econsultancy report, the proportion of companies exclusively using Google Analytics for their analytics needs has risen to 44%, compared to 38% last year and 23% in 2009. 

Our Online Measurement and Strategy Report 2011, produced in association with Lynchpin, looks at the extent to which companies are using Google Analytics, paid for analytics tools, and which tools they are using for which reporting requirements.

Below, we include some charts from the report and also some opinions on whether Google Analytics can meet all your needs. 

Who's using Google Analytics (GA)?

Does your organisation use Google Analytics? 

What are companies using GA for? 

87% of companies exclusively using Google Analytics use it for understanding traffic and conversion KPIs, 68% for campaign tracking, and 60% management reporting.    

Just 17% of client-side respondents use GA for product and cross sell analysis, and 9% for rich media, Flash and video tracking. 

Do you use Google Analytics for any of the following types of reporting or insight? 

The report has further charts about which tools (i.e. Google or a paid-for tool) are being used for different types of reporting or insight.

For example, those companies using both GA and a paid analytics tool are more likely to use Google for PPC optimisation, site search usage and campaign tracking.

Why do companies use paid analytics tools along with GA? 

The main reason given here (by 28% of respondents) is that Google Analytics is not sophisticated enough for their requirements. 

Brian Clifton, former Head of Web Analytics at Google EMEA and author of Advanced Web Metrics with Google Analytics, is sceptical about the perceived limitations of GA:

'You don't know what you don't know' is a phrase I use when describing this situation. The issue with Google Analytics, and it's a nice issue to have, is that it is so easy to set up. It really is a matter of minutes before you can be collecting visitor data.

But that's also the problem - it creates the expectation that's all there is to it. The sophisticated stuff of tracking social media engagement, file downloads, rich media/Flash usage, visitor labelling, transactions, KPIs etc., requires additional thought and configuration. Often this goes unnoticed and hence a perception of "GA cannot do x, y or z" grows.

I have yet to come across a customer who has exhausted the possibilities of GA. Often they say "there is too much to do..."

According to Lovehoney e-commerce manager Matthew Curry, GA has enough features for many companies:

The hard part is justifying why you would spend hundreds of thousands of pounds on a paid tool. The argument needs to be completely compelling and the paid vendors aren't stepping up to the plate - they're pushing the "add on" services like MVT, mobile and social, rather than differentiating the core.

What is the main reason you continue to use a paid-­for tool as well as Google Analytics? (Companies not exclusively using Google Analytics)

The 'other' reasons given included the fact that Google only promises to keep data for up to a year, as well as security and reliability issues. According to Matt Curry: 

You've of course got the issue of handing everything over to Google, how reliable that it is, and how much data they keep. The issues in April with lost data bring a reminder of how vulnerable you are.

There is of course, the option of going down the Urchin route [another Google product], to get a reasonably comparable feature set. I don't know why this isn't more popular? Maybe I don't understand it properly yet.

Lynchpin MD Andrew Hood also acknowledges the issue with data ownership:

Data ownership, retention and integration are common tipping points for firms looking beyond GA to paid-for vendors.

Although Google has been generous to date in keeping historical data available and proven reliable in delivering the service, IT departments wince at the lack of formal commitments to uptime and access to data.

While marketing departments love the intuitiveness of GA reporting, traditional analysts quickly become frustrated with the inability to get to the underlying data. The scope to integrate data with other sources (e.g. offline) is extremely limited by the lack of visitor level data.

Depesh Mandalia, Senior Marketing Manager at Tesco, suggests that verifying a paid-for package's numbers is one reason, along with paid search reporting:

For me it seems GA's deep integration with AdWords is one of the key drivers to dual running with a paid-for package, which can take a little more effort to fully integrate. Another interesting take on this is in verifying a paid-for package's numbers.

Whilst the majority of analytics packages differ in the way they track visits and visitors, I've used both GA and other free analytics packages to benchmark paid-for providers and verify, within a certain tolerance, that the numbers correlate. The worse thing to happen is to completely trust the numbers and find a glaring mistake which could add +/- 20% to your reporting (as we did!)

Reasons for not using Google Analytics

42% of companies say the main reason they do not use GA is because they're happy with their existing web analytics supplier, while 19% aren't happy about Google having access to their site data, and 15% say it isn't sophisticated enough for their requirements. 

Adobe Product Marketing Manager Mike Quinn argues that paid analytics have a number of advantages: 

Are they happy with the often restricted and simple high level views of their data or do they want to dive deeper and analyse data further? Analysis is so much more than simple page views and click through, to get real value from your analysis tools you need to be able to gain insight across all your digital channels including mobile, social, video. 

Can you get a full view of your customers? Customers interact with you in so many ways and with so many touch points it’s critical to bring in your CRM, e-mail, content management and ad-serving applications. If these systems cannot easily integrate with your analysis tools then you will be missing out on some important customer insights.

What is the principal reason you don't use Google Analytics? (Company respondents not using GA)

According to Matthew Curry, there some reasons to move to a paid vendor:

It's become cripplingly slow. You're no longer waiting seconds for a report to load, but minutes, and the Fast Access Mode has a degree of inaccuracy that isn't very helpful, especially on statistics where a 5% change means something, like conversion rate or page load time.

If we were to move to a paid tool, speed would be the reason why. I'm much less productive if I'm waiting two minutes per click. 

Also, because of the GA terms of service, you can't pass anything that could singularly identify a customer, so for example a customerID if you're looking at lifetime value. You have to be creative here, and the actual terms of what you can and can't do are still very grey.

According to Andrew Hood, there are other reasons why people may opt for paid analytics tools: 

GA has a natural strength for things like Adwords analysis - in reality it's hard for any paid-for vendor to streamline this process quite like Google can. Set against that, GA has limited page flow and pathing options for content-heavy sites compared to the paid-for vendors.

There is also a flexibility/accuracy trade-off. GA's advanced segmentation is on the face of it an extremely powerful tool (giving the enterprise vendors a run for their money) - but it often samples (approximates) data to the extent that the figures cannot be relied upon for decisions.

Graham Charlton

Published 13 July, 2011 by Graham Charlton

Graham Charlton is the former Editor-in-Chief at Econsultancy. Follow him on Twitter or connect via Linkedin or Google+

2565 more posts from this author

Comments (31)

Comment
No-profile-pic
Save or Cancel
dan barker

dan barker, E-Business Consultant at Dan Barker

I think 'the' gap between google analytics & paid-for platforms is that you are not allowed to use it to the level where you can 'personally identify' a user.

That becomes an insurmountable problem for some companies, for example when trying to join up customer history with what people are doing on the website.

about 5 years ago

Matthew Curry

Matthew Curry, Head of Ecommerce at Lovehoney

Hey Dan - I mentioned this in my email to Graham, it's a bind if you're doing any kind of lifetime analysis - there "is" ways of getting around it, generating UUIDs as custom variables at Visitor level, as long as you can't tie back to any SOP or CRM system. That it wasn't covered in this post, and generally isn't talked about makes me think that the GA ToS a) aren't read and b) are vague in places - so I fear that a lot of people tie up GA to other systems without realising that they're actually not allowed to.

about 5 years ago

Graham Charlton

Graham Charlton, Editor in Chief at ClickZ Global

Hi Matt / Dan

I've added the quote from Matt to the article now. I had 1,000+ words to choose from, and obviously omitted a key point...

about 5 years ago

Matthew Curry

Matthew Curry, Head of Ecommerce at Lovehoney

I'm amazed you managed to make any sense out of it at all :-)

about 5 years ago

Avatar-blank-50x50

Daniel Peden - Epiphany

+1 on what Brian has said, its a huge misunderstanding within the online community that Google Analytics is a limited product. It's very simple to set-up but its also very simple to add advanced features on. The majority of Google Analytics partners (http://www.google.com/analytics/partners.html) have a lot of custom JavaScript than can be used to quickly begin tracking the most complex of website elements.

On this flip side I agree in part with the speed issue and 100% with the sampling issue. Google Analytics is much quicker than a large number of paid providers. Trying to do deep analysis on websites with high traffic volumes is sometimes very frustrating but I don't see any solid reason that Google Analytics shouldn't be on almost every site interested in web analytics - it's free, powerful and as a minimum you can use it as a sounding board for your other analytics packages.

As Depesh said: "The worse thing to happen is to completely trust the numbers"

about 5 years ago

Matt Clarke

Matt Clarke, Ecommerce Director at B2B

Completely agree with Dan and Matt. It's an amazing tool to have for free and event tracking and custom variables do make it quite extensible, but there's also so much it doesn't do yet, which is why we need to use other tools or build our own to integrate with it.

If we could identify individuals, do CLV analysis within GA, see actual net profit from Adwords advertising, manually enter data on costs of marketing activity or perform custom calculations it would become considerably more powerful. For now, the API will suffice but it would be nice to have integration - and a faster UI.

It will be interesting to see what happens over the next 12 months, given the impending doom of the EU cookie legislation. If only 10% of customers are opting in to being tracked it's going to make even the horrible fast access mode data look reliable!

about 5 years ago

Avatar-blank-50x50

Amor Perez

I have more than sufficient with the information that facilitates Analytics

about 5 years ago

Avatar-blank-50x50

Richard Price

Matt makes a very good point. The elephant in the room here is the EU Cookie Directive. The Information Commissioner's Office website has apparently seen a 90% fall in Google Analytics recorded traffic since the introduction of explicit cookie opt ins in May. http://www.flickr.com/photos/vickyb/5859873960/

about 5 years ago

Avatar-blank-50x50

Marcel Soleda

What? An article about web analytics and there's no mention of Avinash???

I think that most of the companies that are spending money in non-free analytics tools are not aware of the 90-10 rule: 10 in tools and 90 in people.

about 5 years ago

James Lang

James Lang, Senior Research Consultant at Join the Dots

As researchers, our major issue with GA is the inability to link with other datasets (e.g. visitor survey data) at an individual level in order to crossbreak the data in new ways. Combined behavioural and attitudinal data is more powerful than either alone. With GA you can't get this integration, whereas with paid packages you generally can.

That said, we love GA as a free tool in our own marketing, particularly in terms of the AdWords integration.

about 5 years ago

Alan Williamson

Alan Williamson, Director eCommerce at Goldmedal Travel part of the DNata and Emirates Group

I agree with Marcel re the people element. Although it’s hard to find good analytics people so in many instances I think business have the intent but struggle to find the right people.

about 5 years ago

Robert Mobberley

Robert Mobberley, CEO at Peformance Motorcare Products Ltd

Agree that with the right people asking the right questions you can get a lot more out of GA than many realise and Avinashs' book is certainly worth a read.

However I would also wholeheartedly agree with Dan & Matt - the missing link is exactly that - the link between the data and the individual customer.

Unfortunately for many smaller businesses there is the big issue of cost to be able to make this leap.

about 5 years ago

Linus Gregoriadis

Linus Gregoriadis, Research Director at Econsultancy, Centaur Marketing

@Marcel @Alan Fair points, this post is focused very much on web analytics technology, though the people are of course the most important thing to make sense of the data.

According to our report referenced above, an average of 52% of analytics expenditure goes on internal staff, 29% goes on technology and 19% on consulting and services. We are quite a long way from the Avinash 10:90 rule but, anecdotally, the penny has dropped for many companies that they have to invest in the people side of things.

about 5 years ago

Dave Wieneke

Dave Wieneke, Director of Digital Strategy - ISITE Design at www.UsefulArts.us

Can Google meet "all my analytics needs?" Nope, but they have a big bite of it.

Let's say these needs can be grouped this way:

A: Web metrics
B: Lead measurement
C: Quantitative modeling
D: Business intelligence

Google has run the board on the first category. There really isn't an at-scale competitor for web metrics, and once GA code goes on, it persists.

There are some scripts which allows you to tag individuals and measure number of visits, order value and such. This isn't out of the box, but its fast work -- and more full featured lead mgt systems (Marketo, Eloqua, Pardot, Neolane) are more robust paid options.

Omniture pulls ahead in analysis - and IBM / Cognos and big data are taking the top of the BI market -- can anyone add a down market solution? Google is the common starting point for web metrics - and to displace Google vendors have to work on a layer above this.

about 5 years ago

Avatar-blank-50x50

Keith Lavender

Google Analytics is a great tool and because it is so simple to implement there is less room for error.
However the limitation is that GA aggregates all of its data and will not deliver data down to visitor level.
I think with the new campaign attribution method that they are discussing will improve the amount of actionable data and it is looking very impressive - just wish they would provide visitor level path information.

about 5 years ago

Anna Lewis

Anna Lewis, Google Analytics Analyst at Koozai

I've been trying to work out whether Google Analytics is sufficient for our website and our clients recently. For the most part I believe it is more than sufficient.

The one problem I have, which is something you highlighted, is the fact that all your data is in one pot and Google has control over this. There was a major panic for many people when April's GA data went wrong for everyone. This really proves that a backup outside Google is definitely required, in my opinion.

about 5 years ago

Avatar-blank-50x50

James Standen

Regarding the comments on a business intelligence solution, what I hear from our customers is that they are amazed at the cost of paid reporting/business intelligence solutions for web analytics platforms, and the lack of flexibility.

While Google Analytics it self doesn't have a lot of functionality, the API makes lots possible, and there are a number of tools available (Google has many of them listed in their Application Gallery).

I think that this is an area where third party tools are going to quickly provide equal or better capabilities to paid solutions, and the overall cost is going to be very compelling. (Of course I'm biased, since that's the area I'm working in :-) )

about 5 years ago

Brian Clifton

Brian Clifton, Author, CEO & Web Metrics Strategist at Advanced Web Metrics

It surprises to me to hear that "identifying individuals" is cited as a key limitation of using Google Analytics. Surely you guys are aware of the new EU privacy directive that came into law on 25th May 2011...?

Overview and discussion here:
http://www.advanced-web-metrics.com/blog/2011/05/20/google-analytics-and-the-new-eu-privacy-law/

Essentially, although poorly worded as a directive, the EU law is about protecting the privacy of individuals visiting a website. Unless the visitor is a customer or subscriber, identifying them as an individual is a big no-no (rightly), that at last has been written into law.

On the other hand, IF a visitor is a customer, subscriber, or in some other way identified themselves to you, then tracking their web activity and passing this to your CRM system is not a problem - either in law (though your privacy statement needs to be clear and transparent on this), or using Google Analytics. In fact, I describe this method in Chapter 11 of my book.

I agree the Terms of Service for GA are grey in this area for the simple fact that the legal team at Google struggles to keep up with the fast moving pace of the Internet. Remember G have many, many products and all are internationalised. For example, GA is available in 30 languages. So its not a simple task for them to be up-to-date in the time frames we, web managers/owners, work in.

To respond to Mathew Curry's point - passing a customerID from/to your CRM system is NOT a problem in GA and I have many clients configured this way. To keep within the GA terms:

1. The customerID value used in GA must not contain personal identifiable information i.e. no username, email address etc.

2. The customerID value should only be applied when the visitor transacts or logs in. That is, explicitly consents to being tracked as an individual, according to your privacy policy. Do not use this method if a returning customer is simply browsing your site.

As long as you follow these, best practice approaches, there is no problem from Google, and I expect them to make that much clearer when their terms are updated (written in 2005).

Correcting some inaccuracies in the article:
--------------------------------------------
*Google has committed to retaining GA data for 25 months, not 12 months as quoted in the article.

**There was no GA data lost in April of this year. Yes, some reports did contain errors. However, once the problem was solved, all data was reprocessed and reports corrected. All web analytics tools (paid or free) have reporting issues from time to time. For example, search for "omniture outage" (http://www.google.com/search?q=omniture+%2Boutage). I really do not see system reliability as valid reason when we are discussing the networks of Google, Microsoft et al.

Best regards, Brian
Author, Advanced Web Metrics with Google Analytics
Former Head of Web Analytics, Google EMEA

about 5 years ago

Matthew Curry

Matthew Curry, Head of Ecommerce at Lovehoney

Cheers for clearing that up Brian - the "identifying individuals" issue has been around for years, specifically because of the grey areas in the terms. I remember when I originally asked, I was told (by a Google rep at Internet World), that it was "outside of the spirit of free web analytics". It's this line in the TOS..

"nor will You (or will You allow any third party to) associate any data gathered from Your website(s) (or such third parties' website(s)) with any personally identifying information from any source as part of Your use (or such third parties' use) of the Service."

which is the sticking point. If you're passing a CustID, which you can then track back to your own CRM, then essentially you're breaching this? This is why I was suggesting an anonymised UUID.

As always, this is up to the interpretation.

I would guess what most people are trying to do is a Customer Lifetime Value based on source/medium/keyword of first visit? That would be a VERY handy figure to know.

Regarding the lost data, indeed, lost is the wrong word. Data was however, missing for several days in April. Trying to do any segment analysis based on April's data was impossible for a week or so. Of course I'm sure every system has issues at some point.

Matt

about 5 years ago

Matt Clarke

Matt Clarke, Ecommerce Director at B2B

@brian Sorry, I didn't phrase that well. Obviously I'm aware of the problems associated with PII and wouldn't use GA for storing anything which identified an individual. What I meant was that it would be useful to be able to link a customer's past purchases together in GA to enable CLV or other calculations to be performed in GA itself.

We can already do this using custom variables, and can calculate CLV externally, but having the ability to have it down in GA itself would be lovely.

about 5 years ago

Avatar-blank-50x50

Gareth Beer, Ecommerce Manager at Speedo International LimitedEnterprise

GA is fine for the basic level of analytics, but when your looking to go so much deeper and justify marketing spend for example GA falls short.

about 5 years ago

Avatar-blank-50x50

Sharon Dawson

Has anyone experienced a huge discrepancy between data reporting through Google Analytics and reporting from other sources - predominantly media agencies who are managing display campaigns? I've recently encountered such an issue with a very large scale of variation between the numbers, and have yet to understand why. Any help or advise gratefully received!

about 5 years ago

Avatar-blank-50x50

Peter Clarke, Head of Marketing at Pumphouse

I fall into the category of using many different tools such as link grader and keyword grader from Hubspot. Does GA do this as well? Have I not looked hard enough at this product. I use it for basic stats and Adwords analysis. Thanks for any responses to this in advance. Peter

about 5 years ago

dan barker

dan barker, E-Business Consultant at Dan Barker

Gareth: I think that depends how you set things up & what exactly you're trying to do. I think GA goes way beyond a 'basic' level of analytics, but sadly it doesn't do the best job of showing a casual user how to move from 'basics' to the really useful stuff that's buried in there.

Sharon: Yes, that's really common. It's common for numbers to differ even from other web analytics tools (eg. Omniture SiteCatalyst). I worked with a client recently with exactly that problem - their display ad & PPC partner was reporting vastly different numbers to GA. It can be any number of reasons depending on the setup sadly, but some include: 1) The display numbers report 'clicks' whereas GA only reports if those clicks result in a successful visit (ie. they don't hit the 'back' button in between), 2) The display ad tag fires in the head vs GA in the foot of the page (or vice versa), 3) The display provider reports on conversions within a window of time (eg. 30 days) vs GA that reports (usually) on 'last click' sales. Perhaps post more info, post the question in the forum, or feel free to drop me a note.

Peter: Those do exist (sort of) in different forms. I prefer doing them in GA (in combination with other tools), but they're nowhere near as simple as Hubspot's tools. There's a tutorial I once wrote here that does something very vaguely similar to Keyword Grader: http://scr.bi/o0BK0u .

Hope something there is of use!

dan

about 5 years ago

Avatar-blank-50x50

David Brown

It's also worth pointing out that (as far as I'm aware) Google Analytics is not certified by any of the international auditing bodies like ABCe in the UK, TUV in Germany, and OJD in France.

This is a major problem for sites wishing to sell advertising space on their site.

All of the major paid web analytics solutions (AT Internet, Webtrends, comScore, Adobe and IBM) can provide this functionality.

about 5 years ago

Brian Clifton

Brian Clifton, Author, CEO & Web Metrics Strategist at Advanced Web Metrics

@Sharon - take a look at my vendor agnostic accuracy whitepaper:
http://www.advanced-web-metrics.com/blog/recommended-reading/

My blog category also has further examples:
http://www.advanced-web-metrics.com/blog/category/privacy-accuracy/

@David - I looked into auditing extensively while I was heading the team at Google. The issue for me was this...

- Every country has its own auditing body, some have multiple ones.
- Each auditing body e.g. ABCe, charges a fee to the vendor in order to "certify" their product. Essentially that means they spend some time understanding how a vendor collects data - it never was clear to me why the vendor should pay for this.
- The certification fees are considerable and annual. From memory, ABCe was around £3000 per year, and that was in 2007.
- GA is a global product. In keeping with that spirit, we would have to be certified in all countries, certainly the ones it is localised in, which is 30 languages.
- That all adds up to a lot of money each year to be certified (we are talking about hundreds of thousands of pounds each year).
- I simply could not justify this, so it was dropped (I did have Urchin 6 certified as a test case).

To be honest, given the extensive free documentation and help material that G provides, I simply could not see why we (Google) should pay for the privilege of an auditing body to understand the data collection technique(s). After all, clients of ABCe pay significant money to be audited.

@Matthew @Matt - noted and thanks for coming back on it.

about 5 years ago

Brian Clifton

Brian Clifton, Author, CEO & Web Metrics Strategist at Advanced Web Metrics

@Gareth @Dan

You need to get hold of a book to understand the full capabilities of GA and move beyond the basics...

Happy to recommend one if you wish, though I am bias ;)

about 5 years ago

Ashley Burgess

Ashley Burgess, Head of Web Analytics at Periscopix

@Gareth @Dan I agree with Brian on the book front (I think we all know which one we're talking about). Beyond that there is a great network of Google Analytics Certified Partners out there, many of whom provide training. To my mind, its unrealistic to expect to use advanced web analytics techniques without some kind of time or money investment. The beauty of GA is that you can invest your money in growing knowledge not just the WA solution licence fee.

@David @Brian - although GA certainly used to be a no-no for ABCe audits a while back they have been making moves towards allowing GA users to be certified through ABCe. We were asked a while back to certify GA implementations inline with some new tests ABC were doing (becoming certified ourselves was ruinously expensive). I'm not sure where they are up to with this but their approach at that time is outlined in the below link. Effectively mirroring GA data in a local logfile.

I presume they have had to go down this route as they are under-fire themselves and GA has such a huge chunk of the publishing market.

http://coi.gov.uk/documents/guidance/abce-google-mirror-meth.pdf

about 5 years ago

Avatar-blank-50x50

Henrik Haller

The correct issue here is that Google OWNS your data.
So Google uses the data to optimize their price fpr paid search words.

And all users of GA help them getting more money.

You open your own internal doors for a wendor there then can decide how much he can charge you because he knows your succes level, and he compare it with the data he is getting from your worst competitor.

And off course also the other way round.

Google create a result (price) using your data to sell adwords to your worst competitor and you PAY for it!!

How insane are you?

almost 5 years ago

Brian Clifton

Brian Clifton, Author, CEO & Web Metrics Strategist at Advanced Web Metrics

Henrik - that is just silly

What you describe is fraud i.e. illegal. If you have evidence of this then forward to the legal authorities.

almost 5 years ago

Avatar-blank-50x50

草榴最新地

You couldn't be more correct!

almost 5 years ago

Comment
No-profile-pic
Save or Cancel
Daily_pulse_signup_wide

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Daily Pulse newsletter. Each weekday, you ll receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.