|AdWords clickthrough disparity|
At odds with that reported by traffic tracking service
| 4:05 am on Sep 1, 2003 (gmt 0)|
All PPC landing pages have tracking script on them so I'm puzzled as to why there is this big difference (Google shows 1,800 click thrus & Hitslink shows 1,200).
I wondered if it was maybe because of the time frame period used by Hitslinks to define a user session. I figured if it was too long maybe Google traffic clicking in & out quickly wasn't being registered. However my testing shows that Hitslinks seems to register the shortest of visits, so I guess that idea's dead.
Can someone please suggest what might be the cause of this disparity? Thanks.
| 12:26 pm on Sep 1, 2003 (gmt 0)|
I don't know how your tracking service works or anything about it, so hopefully someone who has experience with it can offer relevant comments.
I use another service and I have found it matches Adwords results very closely. If I do find significant differences in results or suspect click spam, I copy and paste the relevant stats from my tracking service report into an email to Adwords support and let them sort it out.
If your tracking service supplies detailed reports I suggest you do the same. Adwords Support is very helpful, especially if you give them quality information eg: timestamps, referring urls, IP's etc.
| 2:59 am on Sep 5, 2003 (gmt 0)|
hmmmm, I'm using hitslink as well. in the beginning I was watching very closely and the stats matched. But, since you raise the question I've looked again. Over about three weeks the Google click counter shows 18.5% more hits than hitslink, though my data set is much smaller than yours. And, when I look at my actual logs, they appear to match the hitslink stats. I've written to adwords support. Now let's see what happens next.
| 10:47 pm on Sep 11, 2003 (gmt 0)|
If you read on this board their are a other messages that indicate similar observations and there is really nothing you can do unless they show you how they count and you know they would never do that.
Andrew Goodman said in his book "its a sleazy industry". They will make you pay what you are willing to pay and you just cant do anything because they hide things very well.
| 2:51 am on Sep 14, 2003 (gmt 0)|
Advertiser123, I sincerely hope that you are quoting me out of context. :) Haven't re-read the report in awhile.
There is of course a history of sleaze in PPC here and there, but overall, is it sleazy today? I would say quite the opposite given the tighter relationship PPC advertisers have with ROI data - and the insistence by the major clickthrough providers (Google & Overture & FindWhat) that advertisers should expect a strong, measurable ROI and should track their results.
I would say that advertising in general in the past 50 years or so should bear most of the "sleaze" tag for its willingness to oversell something so nebulous. Here in Canada a government-linked ad agency (typical stuff in politics where friendly ad firms get juicy contracts to promote government services etc.) has just been involved in a scandal involving billing for promotion that never happened - "creative" exposure at sporting events and a bunch of other stuff they figured would be difficult to verify. But that seems to point to a larger problem: the benefit (or lack of same) of advertising is often difficult to verify - and big ad agencies have the big-time profits to prove it!
So in that context, search-based PPC advertising is pretty accountable by comparison.
I am trying to learn a little bit more about reporting disparities myself. It's clear that a click is not an indisputable, measurable "event." The interpretation of what "counts" as a click can vary depending on who is measuring. Many of the metrics services will give you a list of possible reasons for disparities, as will Google if asked. I suppose my rule of thumb is that disparities of 5%-10% are normal enough given the steady volume of lukewarm, uncommitted clickers who may click away very soon after clicking (without waiting for a page to load, etc.). But when you get big disparities clearly it represents a problem or anomaly of some sort. I don't think those problems are necessarily attributable to bad reporting or to click fraud in all cases, though. It's difficult to generalize. Sometimes an investigation is needed, but on the whole, I believe we're going to have to allow for considerable slippage for the time being.
Given the increased scrutiny the major and 2nd-tier PPC's face now, "sleaze" is virtually impossible to get away with, although there is still certainly a willingness to divert attention from uncomfortable issues such as click spam, impression spam, and unevenness in quality of partner traffic, etc.
Some companies (correct me if I'm wrong) make it more difficult to see what's going on than others. I find it difficult to get meaningful reporting from LookSmart, for example. I can track ROI of course using a tracking URL, but certainly they don't provide the kind of reporting detail offered by "true" PPC providers. Anytime there is a chance you're being sent clicks you don't understand, there is a problem IMHO.
| 8:41 am on Sep 14, 2003 (gmt 0)|
We use Gotoast.com to track Google Adwords performance. The result is the same as that of Adwords.
| 6:26 pm on Sep 14, 2003 (gmt 0)|
I am not sure why you are giving an excuse for these "miscalculations"(for ANYONE in the industry). If you read these posts most people complaining have used 3rd party software to track their clicks (unbiased) and are finding that about 20% of the clicks google shows are not being delivered (I would not use gotoast as a fair tracking clicks company since they are brokers working for google's adword's program and present what google shows otherwise their service would be useless).
Computers are nothing but a precise mathematical tool and here we are talking about simple math. You have google with 100 Phd's in either computer science or math suppose to be running complex algorithms for their search engine and cannot track with precision?
It is very easy to track double clicks, triple, etc. and everyone with some programming experience knows that. They say if you complain and you show that your stats are showing more clicks they will most likely pay you back so they can try to keep you "happy".