AdWords is now Google Ads. Our new name reflects the full range of advertising options we offer across Search, Display, YouTube, and more. Learn more

5.3K members online now
5.3K members online now
Improve your Google Ads performance and boost your ROI, CTR, and Quality Score
Guide Me

What is the correct way to optimize a GDN campaign?

Follower ✭ ✭ ✭
# 1
Follower ✭ ✭ ✭

This might be a more data science question.. but since I'm running GDN campaigns, i figured I should ask here.


Suppose I have a campaign with following results.


Banner 1 = 2% CVR
Banner 2 = 1% CVR

Placement 1 = 2% CVR
Placement 2 = 1% CVR

Landing page 1 = 2% CVR
Landing page 2 = 1% CVR

Day 1 (say Friday) = 2% CVR
Day 2 = 1% CVR

Time 1 (say Friday) = 2% CVR
Time 2 = 1% CVR




Now, I am looking for the highest converting combination (banner, LP, placement, device, time, demographics, etc.).



If I take the winning variations (looking at them solely w/o relation to each other - i.e.. w/o one dimensional analysis and aggregating them together).. and optimize by saying:


Winning Combo = Banner 1 x LP 1 x Placement 1 x Day 1 x Time 1 x (etc....).  IN other words.. imagine a venn diagram, and taking the intersection of all the variables. At least, that's what i'm attempting to do.



Couple of questions..


1) I'm pretty sure the answer is.. "this is wrong". Is this the WRONG way to optimize a GDN campaign?

2) If not, what is the right way?




2 Expert replyverified_user

Re: What is the correct way to optimize a GDN campaign?

Top Contributor
# 2
Top Contributor
There's nothing wrong with this approach... in theory...

However, we must consider the end game here... if you take this approach to its logical conclusion you will be running one single banner on one single placement between 3 and 4pm every Friday... OK - that's extreme, but you get my point.

My approach (and it may well be flawed) is to look for those general trends that work in the first instance. So, for example, do my ads work best during the day or during the night? Weekdays or weekends?

Starting with broad brush strokes we begin to get a picture of the market which we cannot get if we get too close. Or we get such a close view of it we misinterpret it.

And there is something else to bear in mind - and that is the volumes available. You might have a killer combination of factors which results in an incredible ROI - but if you only get one visit a month it's not going to be truly profitable.

We are not looking to find the killer combination (although it would be nice if there was volume to accompany it - no questions there) instead we are looking to get a positive ROI across our account. Consider this - if you had to choose between having 1,000 visitors that had a 100% ROI or 100 visitors with a 1,000% ROI which would you choose?

In terms of return they are the same, but it's probably much easier to increase the volume and get 2,000 visitors of the former kind than 200 of the latter.

It's a little like the marketing question - How do you make a $100 bottle of wine look cheap? The answer - put it next to a $250 bottle... this is similar - just because one campaign is "only" generating 100% ROI it looks to be underperforming next to the campaign that is giving a 1000% ROI. But if it was the only campaign you were running you might be quite content and be looking to nurture the campaign to get incremental improvements....

On the other hand, if you can find that perfect storm you are looking for, and there is enough volume, you have found a true winner!

Re: What is the correct way to optimize a GDN campaign?

Top Contributor
# 3
Top Contributor
Hi taewoo,

Steve gave you a great answer. My view point on this is there's not enough data presented here to formulate a plan. We don't know the number of impressions, so there's no way to calculate the significance of your data. We don't know the ads' performances on the specific placements (Does banner #1 perform better than banner #2 on site #1?). We don't know if 1% is profitable for you. We don't know the CTR. The landing page has little to do with CTR, but it can affect your conversion rate. That could be actionable data.

After all that, I agree with Steve. It's about maximizing profits. If the lower rate of return is still profitable I don't think I'd cut it off because it's less profitable than the higher rate. I'd probably design a new banner ad and see if I can beat the 2% CVR, or at least beat the 1% CVR.

Do you have more than one size ad? Are both banners the same size? This could affect the placements. Not all sites accept all sizes. Are you also running text ads along with your banners? That's usually recommended as a fallback option if the site does not accept any of your banner sizes.

The best way to optimize your campaign is to test, evaluate, take action, test again.

Best of Luck!

petebardo -- Deadhead doing AdWords