Last month while at Hero Conf, I had the opportunity to speak on a panel about bidding and budgeting. I learned some really unique Facebook bidding and budgeting strategies, along with gaining a deeper understanding of how automated bidding can improve account performance.
What I Talked About
I presented a case study about how Acquisio’s automated bid and budget management system delivered improved results for one of my clients. First off, here’s a brief rundown of how the bid and budget management (BBM) system works.
- Bidding is nearly real-time: Instead of making large bid moves 1x per day, BBM make small incremental changes 48x per day (every 30 minutes).
- Set Small Budgets: Budgets are set only 20% over what you intend to spend on a daily basis.
- Based on the budget you set, the BBM system will try to spend your budget while at the same time gaining as many conversions as possible within that budget.
Bids are adjusted every 30 minutes to compensate for the ebb and flow nature of Google auctions throughout the day. Based on data BBM collects, the system will bid down when auctions are less likely to convert and bid up when auctions are more likely to convert.
Client Case Study
The client represented in my case study services the higher education market with hundreds of thousands of dollars in ppc spend, dispersed amongst numerous locations. Each location has its own unique budget, making manual bidding and budgeting a very complicated and time consuming task.
This particular client also has a very distinct ebb and flow regarding their conversion activity. Very few conversions take place in the early morning hours followed by a sharp rise in conversion activity in the mid-morning to mid afternoon hours, followed by another slowdown in the early evening and a final uptick during prime time evening hours.
CPC’s in the education space are very expensive, so it is imperative that ‘click burn’ be kept to an absolute minimum. Overspending during a time of day least likely to convert unnecessarily wastes precious budget and sharply reduces profitability. BBM takes these factors into account and automatically allocates spend appropriately. Humans simply cannot efficiently adjust bid or budget targets often enough to prevent wasted spend as well as an automated bidding system.
Overall Performance
The first part of my presentation was an overall performance review. For all BBM campaigns combined, cost per conversion decreased 21%, while conversions increased 29%. This is in part to reducing budgets and pausing all non-converting traffic. Removing non-converting traffic allowed the BBM system to focus the budget only on programs that lead to a conversion. Below are some examples of how cost per conversion decreased while conversion volume grew.
Individual Location Performance
I also discussed how BBM performed on an individual location level and lead to a tremendous performance increase. One particular location saw both stronger than average CPL improvement (37%) and an above average increase in conversions (121%). This increase occurred for two reasons 1) Budget for this campus was higher than the average, allowing more room to operate. 2) This campus had a lot of conversion and click history behind it. The more data any automated bidding system has to work with, the smarter the system becomes in predicting when and where conversions will take place. The charts below illustrate how well this campus performed.
Key Learning’s
Key learning’s from the case study I presented at Hero Conf were:
1) When BBM was originally implemented into the account, we did not budget the campaigns properly. Setting to high a budget before an automated system has enough traffic and conversion history can lead to overspending and reduced profitability. Budget conservatively and increase only once positive performance can be verified.
2) Account structure is extremely important. Our original account structure was one campaign per location housing ad groups representing multiple programs. Although BBM was able to operate, the system couldn’t optimize to its full potential. Reason for this is each program’s keywords had vastly different CPC’s. A couple of clicks at a $50 cost per click was exhausting budget and causing other areas of the campaign to not receive any traffic. Segmenting each program into its own campaign allowed for more even distribution of spend because CPC’s were now similar throughout the campaign.
Conclusion
It was a wonderful experience not only to share this case study with the attendees of Hero Conf, but also to learn how other automated tools bring success to accounts. I am eager to run some split tests of BBM vs. conversion optimizer and other automated bidding tools/features and compare results. I am a believer that bringing automation to accounts can increase performance if done correctly. What are your thoughts? Have you split tested any bidding tools and seen a difference in your results?
