Return to site

How to tame the FB bid

Case Study: Hopeless 2

Do you find FB & IG traffic to be high quality, but expensive?

Would it be ideal to manage the bid, over and over, win the auction, and generate the maximum number of installments? Below we unveil our secret, but let’s have a look at the game itself – Hopeless 2.

We at Aitarget implemented a FB and Instagram campaign, targeted towards men, 18 - 34 y/o. We articulated the CPI (cost per install) should be no higher than 60 cents. We were puzzled when Facebook’s algorithm recommended a CPI estimate of $2.14. For placement we chose FB and IG Newsfeed and Audience Network. We used our tool to fix the parameters of the CPI by creating three basic rules, which adjusted the bid to our requirements.

Our goal was to balance the bid between 10 cents and $1, with flexibility of 10 cents, while simultaneously managing the CPI between 50 cents and 60 cents. We set up an initial condition which had to be followed; a reach of more than 15 installs. (The initial condition needs to be taken into account in developing the strategy, in order to obtain the minimum valid statistics for optimization).

As you can see from the graph, immediately after campaign launch, the CPI increased to the bid amount. A few days later, the CPI cost fell and then went on to plummet. Therefore, we recalculated the CPI over a longer time interval (the ratio of spend to the number of installations). The CPI eventually reached a minimum point of $0.32, and the core source of installments was Instagram.


Navigating the waters of Facebook ad buying can be complex and intimidating. Even when advertisers set budget limits, sometimes the Facebook algorithm will return over budget results. Our AITarget Optimization tool allows advertisers to create the Corridor Strategy for the FB bid, and acquire all possible traffic at fixed prices. In addition, the tool lets advertisers create strategies for managing ads, which are based on the KPI of an advertiser - CPA, ROI, ARPU, LTV and others. Our tool also enables automation of any actions: changing the bid/budgets, turning on/off adsets, etc. All in all, our tool streamlines the process, makes the complex manageable, and empowers advertisers to reach the right audience on budget.


In the case where you’re targeting an audience with specific interests with a small reach, this audience could ‘burn out’ quickly. This means that the frequency of one ad per one user will increase, and the Facebook algorithm will quickly show the ad to all chosen users. Users who had shown interest in the product or service, and were likely to go to the AppStore via the ad, as well as those who had no interest and would complain about its irrelevance would be included in the audience.

An advertiser has to think through the strategy, and broaden the target to non-overlapping audiences, that advertisers can work with in the same way. It’s also important to be able to go back to the previous audience as a benchmark, and to gather learnings over time.

Here is an instruction guide to creating a campaign with AITarget.

All Posts

Almost done…

We just sent you an email. Please click the link in the email to confirm your subscription!

OKSubscriptions powered by Strikingly