Recommended reports
Online Measurement and Strategy Report 2012
The fifth annual Online Measurement and Strategy Report, carried out in association with web analytics consultancy Lynchpin, contains a comprehensive analysis of issues affecting the web analytics industry and valuable insights into the use of analytics and business intelligence tools.
Online Media Report
The Online Media Report focuses on the online advertising landscape and how media buying has evolved in the last decade. The comprehensive report, which has a US focus but is relevant for a global audience, contains a detailed analysis of the current trends and issues affecting the marketplace, as well as an overview of emerging players such as demand-side platforms (DSPs).
Planning Director at Kitcatt Nohr Digitas
08 July 2008 12:38pm
Can anyone explain that mystery to me: several of my clients have arrival rates from the online ads we've produced for them in the range of 65-70%. I personally find that shocking: losing 30% of the CTRs is just amazing!
At the same time, neither clients nor their media agencies seem willing to address this issue, either through some sort of compensation, or technical means. Can anyone clarify what are the things that may influence arrival rates - why a recorded click on the ad may not result in the destination page being viewed (apart from the obvious reasons that there is no tracking in place, or that the destination server is down)?
E-Business Consultant at Dan Barker
08 July 2008 21:08pm
hi, Lazar, how are you?
That does seem very low. Here are a few possibilities & things to check:
- Page load time may be slow - this will always result in drop-outs, especially if your tracking code is at the end of the page
- Javascript tracking - you'll lose some amount due to that. Easy check is to compare your javascript numbers against your server logs.
- Visitors have your tracking company blocked. Admittedly a small amount, but there are tools out there (& other simple methods) for blocking google analytics, coremetrics, omniture, etc.
- Accidental ad clicks. I have an annoying habit of clicking my mouse round & about on pages. Occasionally I accidentally hit an ad, realise, & hit stop before the page loads completely.
- Fraud. Are these false clicks coming from the same sites?
Of all of those it's likely to be the first: If someone stops the page before your analytics code fires, it will show as a click, but not a visit/page load.I hope that's useful,
daniel
Planning Director at Kitcatt Nohr Digitas
09 July 2008 09:10am
Daniel, you're a star! Many thanks mate, this really helped, I know exactly what to beat the media agency (and the client) with now! :) Much appreciated and all the best.
Account Director at DoubleClick
09 July 2008 15:06pm
Daniel covered some great points, what type of analytics solution are you using, is it tracking url referrer or is it a 1x1 pixel? If it's a pixel, make sure it is as high up the page as you can get it.
Also would be worth doublechecking the implementation and that the url's are pointing to the right destinations. When tracking across multiple channels this can slip on one channel from time to time.
Dave
-------
hi, Lazar, how are you?
That does seem very low. Here are a few possibilities & things to check:
- Page load time may be slow - this will always result in drop-outs, especially if your tracking code is at the end of the page
- Javascript tracking - you'll lose some amount due to that. Easy check is to compare your javascript numbers against your server logs.
- Visitors have your tracking company blocked. Admittedly a small amount, but there are tools out there (& other simple methods) for blocking google analytics, coremetrics, omniture, etc.
- Accidental ad clicks. I have an annoying habit of clicking my mouse round & about on pages. Occasionally I accidentally hit an ad, realise, & hit stop before the page loads completely.
- Fraud. Are these false clicks coming from the same sites?
Of all of those it's likely to be the first: If someone stops the page before your analytics code fires, it will show as a click, but not a visit/page load.I hope that's useful,
daniel
Planning Director at Kitcatt Nohr Digitas
09 July 2008 16:12pm
Many thanks Dave,
the URLs are definitely right, but it could be the server speed issue, as the pages seem to be loading too slow (the pages themselves are simple, so thye are unlikely to be a problem).
The client and the media agency are investigating the thing in light of all your suggestions here, so hopefully it will be OK soon.
On the point of where to place the tracking tag, I think that having it at the bottom of the page could give us more robust conversion % - i.e. I'm happier to know that the conversion % is calculated from the total number of fully loaded pages (the full exposure) than from the total number of server requests, irrespective of whether the landing page is fully loaded or not.
What do you guys think about that?
Account Director at DoubleClick
11 July 2008 09:09am
Hi Lazar
Putting the tag at the bottom of the page is always risky. On one side you want to know how many people clicked and 'landed', and on the other you want to know how many people clicked and really 'arrived'.
I would get your webteams and agencys support and temporarily put a second different landing tag at the top of the page and check out both tags numbers after a week.
If you see that the top tag has significantly higher numbers, you've found your issue and can work with the web team to tackle it. If the two tags are reporting the same numbers, you know it's not page speed and you can take that temporary top page tag out (although your technology company will strongly advise to keep it in to avoid this type of issue). In that case the other points Daniel went into are still good to check out, let me know if that sounds like a good idea.
Dave
Publisher at 2N Media Ltd - ModernSelling.com
11 July 2008 10:54am
This is very interesting, particularly since it would appear to be a universal problem. We publish a B2B site carrying lots of banner ads, and are just switching to using an extension of Google Analytics to measure impressions and clicks, precisely because reconciling WebTrends numbers with anyone else's was a nightmare. The scary stat I saw that prompted this was that 75% of all web traffic, PPC, SEO, banners, whatever "bounces" - and Bounce Rate is a feature of Google Analytics. I later learned that it includes not just these 30% or so that last 00:00 secs on a page but anyone who leaves from the page they arrived on, be that 00:01 or 10:00 secs/mins later. Some of the experts in this forum might be familiar with this, but I reckon 90% of those spending their $billions online are not - and it would definitely become a factor in any comparable marketing spend where you actually pay for a "result" that really isn't - as in PPC/AdWords.
Planning Director at Kitcatt Nohr Digitas
11 July 2008 12:57pm
Yes, it seems that we have uncovered a very neglected area of evaluating online marketing campaigns!
Dave, the two-tag suggestion is great one and certainly something I will recommend to the teams. They are investigating the thing at the mo and I will get some results early week, so will pass them on. BTW, what is, in your experience, good loading speed for a simple flat-HTML page geared towards conversion (but not a checkout page)?
Neil, just to understand the 75% figure better: is that the % of all the traffic that is reported as 'bounced' in Google analytics, or is it something else?
Publisher at 2N Media Ltd - ModernSelling.com
11 July 2008 14:07pm
Hi Lazar - yes it's a standard measurement on the main dashboard of Google Analytics - which is reportedly now being used by over 60% of all web sites that have any form of traffic measurement in place. And it does refer to all traffic, but can also then be analysed on the subsections of your traffic. For some sites, like ours, the fact that someone came on and just read a single article and then left might not be so bad. But, for many, the fact that 75% of their visitors get no further than that landing page might be food for thought. And then, also, it might make people look at which traffic sources do, in fact, stay and browse more than others. Our incentive has been to note that people clicking through our banners are getting on for twice as likely to hang around (not bounce) as traffic coming from generic Google or even AdWords - so what value a click?
Planning Director at Kitcatt Nohr Digitas
11 July 2008 14:11pm
Thanks Neil, a very good point.
Does anyone know if there has been any challenge to the media agencies and ad networks on this issue yet? By clients, or even trade press?