Often, the results of an online advertising campaign are similar to those of a good Parisian demonstration! 150,000 according to the police, 500,000 according to the organizers!
So the question comes up almost every time... Well, not to such an extent, but how is it possible to observe certain statistical discrepancies between my different tracking tools? Or between my tool and that of my service provider? In fact, it's not uncommon to observe discrepancies in the statistics for traffic generated by advertisers' online ads.
What data are we talking about? Essentially impressions and clicks, which are rarely the same from one tracking tool to another... KPI's on which the value and success of a campaign depend... and therefore, financial stakes that can be a source of tension between the various parties involved...
The discrepancies observed are often due not to the tools themselves, but to different different calculation methods applied.
Here's an example to illustrate this fascinating subject. One tool will count a click when an Internet user clicks in the e-mail. Another tool will count a visit to the advertiser's site (as a click). Any redirection problem (error in the URL, slow loading of the site resulting in the user abandoning the site, redirection to an external page, widespread power failure due to an emerging nuclear conflict, etc.) could therefore create a performance gap between the two tracking tools.
At Dataventure, we're no slouches when it comes to having fun, but when it comes to statistics, we're extremely rigorous!
We won't solve the problems between demonstrators and the police, but we can help you understand these statistical discrepancies and help you minimize them...
A reminder for the two at the back...
Distinguish between a click and a session.
- A click: simply the action of clicking on a tracked link in an e-mail.
- A session: this is defined as all the interactions carried out by an Internet user between a browser and your website over a given period (for example, for Google Analytics, the session ends after 30 minutes of inactivity, but you can also determine this in your analysis parameters).
For many reasons, therefore, there may be several clicks in a single session (especially if a visitor clicks several times in your e-mail during the same session).
At Dataventure , we talk about unique clicks ! If, for example, we see several clicks from the same e-mail (including all tracked links), we only count one. Why make advertisers pay for any bugs caused intentionally or unintentionally by a web surfer?
How can I identify and exclude robot clicks?
Robots, in digital as in Black Mirror, are a real nuisance! With the explosion in email traffic worldwide (3.9 billion at the end of October 2019, with a forecast of 4.3 billion by 2023), the emailing sector and its players have reacted with an arsenal of protection measures (firewalls, anti-spam filters, anti-virus software...) that are increasingly firm.
This isn't The Matrix, but the fact is, robots are putting up a fight, and it's taking its toll on your traffic stats. Their behavior is now very similar to that of humans, and it's becoming increasingly difficult to protect against them and exclude unwanted clicks.
"It is estimated that up to 10% of traffic is generated by non-human clicks on an advertiser's site.
"Welcome to the real world...
Since R2D2 isn't about to heat up his Visa Infinite to take advantage of Black Friday, it's essential to sort things out!
It is therefore possible toidentify these cyber-clicks in real time or a posteriori by probing click logs. By analyzing IP addresses (owner, geolocation...), UserAgent or HTTP header (browser-specific)... you can sort out malicious clicks and exclude them from your campaign performance.
For the past 6 years, Dataventure takes pride in honor to ensure the quality of its tracking system. It's an unwavering commitment and a guarantee of confidence for our advertisers. Our approach is simple: too much is better than too little! That's why we push restriction and control to the maximum, by implementing numerous verification, filtering and exclusion scripts for any behavior that might prove dubious.
In this way, we continuously analyze data incoming traffic for each of our advertisers. And if it turns out that we've been too strict, and that some clicks were actually of human origin, we can set up IP re-whitetrackingscripts... Mea(dventure) culpa!
And this rigor pays off! Our more stringent filtering ensures that our performance does not conflict with our advertisers' findings, so that our clicks match the number of visits to their sites.
Beyond a more realistic vision, the impact is also more ROÏste, since for an equivalent number of sales, our advertisers only pay for the unique clicks that have passed all our filters...
Advertisers and partners who place their trust in us benefit both from our expertise in email acquisition strategy and our technical know-how in tracking. This dual expertise enables them to achieve a high level of performance, with the guarantee of a highly reliable data .
We are convinced that our advertisers should only pay for "real" traffic and real campaign benefits. The rigorous selection of our partners is also part of this qualitative approach, while never losing sight of the need to optimize your KPI's...
To find out more about this fascinating subject of statistical discrepancies, not just in the wonderful world of emailing, check out the CPA (Collectif pour les Acteurs du marketing digital) fact sheets, which present 7 concrete examples on their October 2019 fact sheets.
