Bidding and Segmenting with Facebook Ads
Sitting in the shadow of Google as the secondary PPC tool often leaves Facebook neglected and misunderstood as a PPC medium. The active users on search engines are a world apart from the passive audience, and while ad copy strategy is the immediate distinction between the two, the management of profile targeting and ‘ad burnout’ isn’t hard if time is invested.
So how does Facebook’s ad auction work?
An individual CPM or ‘cost-per-impression’ auction happens with each eligible page view. Although ads are charged on a CPC basis the cost is calculated on a CPM basis – the more impressions you need to generate a click, the more you’ll pay. So, like Google, click-through rate (CTR) is King.
The calculation for costs is:
Max CPC Bid x CTR = Max eCPM
and highest Max eCPM wins
Below are shown four advertisers who are all eligible to display for the same user, but it’s pretty clear some never had a chance of showing their ad.
So an advertiser has 3 choices:
1. Increase volume by increasing bid
2. Reduce CPC by lowering bid
3. Increase volume and reduce CPC by increasing CTR
Of these, option 3 obviously provides the biggest impact for improving performance.
Like the first advertiser in the above example bidding against ‘all users’ if you fail to segment your bids you’ll repeatedly fail to win the auction due to the likelihood of lower click-through; targeted ad copy aimed at specific demographic groups is more likely to receive attention and interaction. Generic targeting and poor ad copy will leave you winning only the dregs of the targeting, which seasoned advertisers consider lower value. Segmenting your campaigns for different demographics allows you to tier the process and bid according to performance and the assumed value of the demographic.
Ignore Facebook’s ‘Suggested Bids’ as these vary by the minute. Instead use a constant bid across a demographic to test ad variations. The true value of Facebook’s targeting isn’t in purely market research and demographics, but in bid optimisation for those particular demographics. Watch the volume of impressions you receive in each segment closely and ask whether you are winning too few auctions – can you afford to increase the bid?
3. Reaching targets
Once you’ve run a test to establish your conversion rate you can calculate the target CPC required to reach your CPA goal.
Bespoke ads for each audience increases CTR, which mean more auctions are won and the max bid can be kept lower.
targetCPC = conversion rate x targetCPA
(£0.40 CPC = 20% Conversion Rate x £2.00 CPA)
Is your targetCPC ridiculously low? Optimise your conversion rate by testing content and placement variations to reduce friction for the user. A higher conversion rate will give your targetCPC more flexibility.
Make sure your segmentations are tracked separately in Google Analytics so an individual conversion rate can be applied to each demographic breakdown. You may find that one demographic’s high conversion rate is supporting the whole campaign, and the bids should be raised to try and win more auctions.
Once targetCPA is being reached the daily budget can be increased. Eventually this will saturate the market, lowering the CTR and causing you to win fewer auctions and drive less traffic.
However, keeping your ad copy fresh and using segments to test copy strategies will stop audiences becoming saturated and push up the CTR, meaning more auctions won and more traffic.
Just increasing your daily budget will do more harm than good in the long run – the key to driving more traffic is the CTR.
Be careful not to have overlap in your campaign and demographic targeting – you may for example, display different ads to a woman aged 21 (‘female’ campaign and ‘age 18–21′ campaign) and end up paying for impressions and clicks on multiple ads, skewing your conversion rate and forcing you to raise your average CPC.