Case Study: Bidding strategy for a TIM, Italy campaign
When setting up a new campaign you should always have in mind that choosing a slightly different bid can change the traffic you’ll receive and how it reflects on the conversions you generate.
In order to make the magic happen, we’re going to show you each of the steps we took during a successful collaboration with one of our clients.
Our first step was to generate the exact same campaigns for three times in a row with 3 different bids, taking care of the following
- To choose the ad format - mobile pop-under
- To targeted all operating system
- To focus only on the TIM Italy carrier
- To target the whole network for adult publishers by using the RON option (Run On Network).
- To set the frequency at 1/24.
After taking care of the first step, we assigned different bids to the campaigns (low bid, medium bid and high bid), we assigned a budged to each of them and run them simultaneously to see which one had the best ROI.
For each of the campaigns the CPMs were set at certain limit, depending on the information shown by the company’s bidding tool (their in-house tool had the option of showing the top CPMs on a specific targeting):
- The low bid campaign started with a minimum CPM of 1.5 $
- The medium bid campaign started with a minimum CPM of 2$
- The high bid campaign started with a minimum CPM of 3$
The next step we took was starting the campaign on RON traffic. But because TIM is the biggest mobile carrier in Italy, we have limited the campaigns at a daily limit of 50$, just to make sure we weren’t taking the risk of spending a very big budget on running a RON campaign with top CPM within a short period of time.
The first results that came through were pretty much to be expected:
- The low bid campaign didn’t receive much traffic during day time (around 1500 impressions)
- The medium bid campaign received around 5000 impressions during day time
- The high bid campaign received around 17 000 impressions during day time
However, when it came to night time (between 2 A.M. and 7 A.M. to be more precise), on all 3 campaigns the traffic dropped close to zero clicks and the lower the bid was, the longer the time frame when the number of clicks was getting lower.
But what caught our eye even more was the distribution of the various operating systems. While the low and medium bid campaigns got approximately equal parts of Android and iOS traffic, the high bid campaign got around 3 times more iOS traffic. That helped us decide which bids generated the most conversions and conclude that iOS was converting much better than Android when it came to the high bid campaign.
High Bid Campaign
Medium Bid Campaign
Low Bid Campaign
|iOS||Conversion Rate (CR)|
|High Bid Campaign||1/5000|
|Medium Bid Campaign||1/12000|
|Low Bid Campaign||1/15000|
|Android||Conversion Rate (CR)|
|High Bid Campaign||1/10000|
|Low and Mid Campaign||No Conversion|
Looking at the numbers even closer we were able to generate a better conversion rate from the same operating system with a higher bid. And from here on, we identified the domains that were generating these types of conversions in order to find the most profitable sources.
|iOS RESULTS FOR VARIOUS WEBSITES: Table 1|
iOS RESULTS FOR VARIOUS WEBSITES: Table 2
In the end we were able to say we had found a campaign that generated a much bigger ROI allowing us to raise our bid and get even more competitive on the converting sources.
Furthermore, we’ve set up a new campaign using just iOS traffic and we also excluded the sources that were sending the most of the traffic but didn’t generate enough conversions to turn the campaign profitable.
At this stage, in order to optimize the ROI and generate new campaigns we played around with the settings and found the ones that were bringing the most profit.
It’s always worth to test different bids, and this example proves it. Keep in mind though, this is now the only bidding strategy you could adopt for any traffic source. Cause every traffic source works differently and traffic types can vary.
You just need to define what approach you’re willing to take when it comes to biding taking into consideration various factors such as:
- Your budget
- The auction model you can select
- What stage are you in when it comes to the optimization process and so on.
Your options are pretty straight forward: you can either go for a conservative approach (starting at a low level and increase step by step), go with an aggressive one (start on top to get the first impression and the highest quality for your placement) or choose staggered bids (set up multiple campaigns with different bids like we did in our study case).
In the end, it’s all about testing, going out and taking action!