AdWords
2.1K members online now
2.1K members online now
Dive into advanced features like Remarketing, Flexible Bid Strategies, AdWords Editor, and AdWords Scripts
Guide Me
star_border
Reply

Using Google Dynamic Remarketing Alongside Criteo

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Hi everyone!

 

We have been using Criteo for around 2 years across all our stores in 7 different EU countries. We recently decided to trial Google Dynamic Remarketing (GDR) in one of these countries and when setting this up with our Google account manager we were advised to turn off Criteo when turning this live as running both would affect the performance.

 

It's now been running for 3 weeks and the performance is real poor in comparison to what Criteo used to provide. We have lost 70% revenue from remarketing  compared to the previous period. Criteo used to achieve a conversion rate of around 2.5% whereas GDR only manages 0.45%. CTR is also down from 0.9% to 0.4%. 

 

We don't want to give up on GDR and I'm sure that it will improve over time, but we are now wondering if we could run both side by side?

 

Would they be competing against each other? Would they compliment each other? Has anyone got experience of running both at the same time?

 

I've tried to find articles online but I can only ever find the typical "Adwords vs Criteo" articles so any advice would be greatly appreciated.

 

Thanks for your help!

3 Expert replyverified_user

Re: Using Google Dynamic Remarketing Alongside Criteo

Rising Star
# 2
Rising Star
Hey Alex, how are things?

I've had a similar scenario like the one you've mentioned.

I think it's possible to run both services with good results, but GDR I found it takes a little longer to optimize compared to Criteo. Also, Criteo has an advantage of not being restricted to Google Display Netword and can connect with so may other services, that's why they can deliver better results depending on the segment.

As you said: "I'm sure that it will improve over time", that's is the most important thing right here. A lot of advertisers don't OPTIMIZE. They just run campaigns hoping for results to appear, and whey they don't appear, they say "this doesn't work" when in reality they need to improve, tune and optimize in order to get results.

One thing you can try to do is to prevent Criteo from using GDN for a while and see how it goes, because running GDR and Criteo at the same time on Google, can create conflicts on your results.

That's my view. Hope this helps.
_

Leandro Martinez | Basta1Click

Marked as Best Answer.
Solution
Accepted by topic author Alex C
April 2016

Re: Using Google Dynamic Remarketing Alongside Criteo

Top Contributor
# 3
Top Contributor

Hi Alex;

This is a nice question.... And I don't think much data is available about comparing the two.
Technically, you can run both. In practice, since Remarketing is the issue, the question is which brings additional conversions. If you are in a competitive verticals in which tailored ads are significant to attract  the user, to my recollection dynamically tailored ads are not an option on Criteo (Correct me if I am wrong  on that...). So, you are left with Google.

 

Moshe, AdWords Top Contributor , Twitter | Linkedin | Community Profile | Ad-Globe
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’

Re: Using Google Dynamic Remarketing Alongside Criteo

Rising Star
# 4
Rising Star
Hi Alex,

If your goal is to compare the performance of remarketing campaigns between Criteo and GDR then I strongly recommend not running them concurrently with the same targeting settings (i.e. potential for users to see ads served by either network). You will be unable to properly measure the performance if you don't have a way to isolate visitors into two distinct pools so that they only see ads from Criteo or GDR.
The reason is because when users are retargeted with ads from both networks it muddles the conversion data. For example, say a user leaves your site and is hit with 10 GDR ad impressions and one Criteo ad impression. If that user returns to the site via a channel that’s not display and subsequently purchases, both Criteo and GDR get credit for a view-through conversion. In this case the performance for Criteo will look much more favorable as conversion rate is higher even though GDR was likely much more impactful in that conversion. Similar issues arise even if you are only optimizing to click conversions.
This scenario also leads to double counting sales. An attribution platform will show a deduped view but it could make reporting more difficult since the same orders could track in both the Criteo and Google platform.
One solution would be to run GDR in just a couple countries while using Criteo in the others to minimize the aggregate negative performance impact and keep network crossover do a minimum.
I would also recommend being very intentional about further optimizing GDR. If you’ve optimized everything and performance doesn’t improve in a couple weeks (and you’ve allowed the GDR remarketing audiences to fully populate) then I would recommend switching back to Criteo. I’ve ran tests between the two networks for a number of clients and Criteo tends to win, especially if the goal is to maximize click conversions. The expanded inventory of Facebook and specialized remarketing algorithms are the big value props for us.

-Jim
Jim Vaillancourt, AdWords Rising Star, LinkedIn
Was my response helpful? If yes, please mark it as the ‘Best Answer.’ Learn how here.

Re: Using Google Dynamic Remarketing Alongside Criteo

[ Edited ]
Top Contributor
# 5
Top Contributor

 I slightly disagree with @JimVaillancourt on one aspect:

>>"one solution would be to run GDR in just a couple countries while using Criteo in the others to minimize the aggregate negative performance impact and keep network crossover do a minimum.

 

Testing both platforms in different countries / regions could (statistically) bias the results. I would rather test  in different time periods (avoiding seasonality as much as this could be done... Though statisticians would most likely to say this also could be biased... Smiley Surprised ), but at least you'll test the same targeting audience.

 

Moshe, AdWords Top Contributor , Twitter | Linkedin | Community Profile | Ad-Globe
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’

Re: Using Google Dynamic Remarketing Alongside Criteo

Rising Star
# 6
Rising Star

I believe that MosheLTV and I are proposing similar solutions. My recommendation is to test GDR in certain, but not all, geographic regions. In these regions it would be a serial (on-off test).


With any serial test there will be variables that can’t be controlled, but need to be accounted for when analyzing results. On an ecommerce site these would include promotions, competitors, one-off marketing campaigns and seasonality - just to name a few.


If GDR is tested in every geo at the same time with Criteo turned off then total website revenue will need to be used to normalize the data for these variables. If GDR is used only in certain geos then the Criteo display revenue over the same time period can be used to normalize the data. This will allow you to analyze display performance before versus after GDR launched in the test geos (serial) but also compare how Criteo revenue changed over those two periods for the geos where GDR wasn’t active.For example, if a test geo saw a 20% decline in display revenue after GDR launched but over that same time period a geo not using GDR saw a 30% decline in dispaly revenue then that would indicate better performance from GDR. Make sure to choose test and control geos that have a high correlation in daily sales.


If you believe that GDR will outperform Criteo in some countries and Criteo will outperform GDR in others then you should test GDR in every country, and it would likely have to be a serial test. If this is the case you could still do it in a staged approach where you aren’t testing every country at the same time. I believe that if GDR underperforms in the geos that it’s tested in then you won’t need to test in the rest as results should be consistent. Not testing every country will also minimize lost revenue, which was mentioned as a concern.

Jim Vaillancourt, AdWords Rising Star, LinkedIn
Was my response helpful? If yes, please mark it as the ‘Best Answer.’ Learn how here.

Re: Using Google Dynamic Remarketing Alongside Criteo

Visitor ✭ ✭ ✭
# 7
Visitor ✭ ✭ ✭
I work at Criteo so I may be a bit biased in my response, but Criteo does specialize in dynamic, tailored ads.

Alex, obviously I hope you stick with Criteo and it sounds like you had great results with us, but whatever you choose it probably is best to just use one vendor at a time. Splitting the cookie pools limits your reach with both vendors; if both run simultaneously with the same cookie pool, then you're bidding against yourself, which just costs you more money in the long run.

Let me know if there is any way I can help with the decision, and I hope you get the retargeting results you want!

Re: Using Google Dynamic Remarketing Alongside Criteo

Visitor ✭ ✭ ✭
# 8
Visitor ✭ ✭ ✭
Hi,

Thanks for the help everyone.

It turns out that some of the tagging was not set up correctly by the guys at Google. I have fixed this now and the campaign has improved a lot. Revenue is still not quite at the level that criteo was but it is certainly improved the ROI. We also have to consider that the reporting is a bit different now so going to use analytics to get more comparable data - Criteo used count all conversions after it was clicked whereas this won't count if they come back through another channel.

I think we are going to turn Criteo back on with the GDN excluded though, that seems like a good idea, hopefully that will bring in some extra revenue.

Thanks again guys!

Using Google Dynamic Remarketing Alongside Criteo

Visitor ✭ ✭ ✭
# 9
Visitor ✭ ✭ ✭

Which type of conversions and revenue are you comparing? Keep in mind that Criteo reports all conversions by default, including every assisted conversion. Analytics only reports last click conversions by default. In AdWords conversions are attributed to the last AdWords click.

 

This can be confusing sometimes, and make it look like Criteo is outperforming when comparing conversions, CPO, revenue, etc. from Criteo's reports with Analytics and/or AdWords. I would suggest comparing performance in Analytics for all sources, and taking a look at conversion paths, assisted conversions and Model Comparison tool under 'Multi-Channel Funnel': https://support.google.com/analytics/answer/6148697?hl=en

More info on difference between Criteo and Analytics data: https://accelerate.criteo.com/hc/en-us/articles/209917585-What-are-the-differences-between-Google-An...