An important part of pay-per-click campaign manager’s job is to find the best bid for a specific ad, keyword and budget combination. This is a subtlety of PPC management that is often overlooked or misunderstood. People spend a considerable amount of time doing ad optimization and building negative keyword lists but they may not test incremental bid levels to see how they affect click volume.
Bigger (Bidding) Isn’t Always Better
Frequently I see very high bids set where they don’t make sense; generally people tell me they set them that high “to get the most clicks.” The rationale is that if you bid higher your ad will be placed higher and you’ll get more clicks. This may be the case for some but it’s rarely the best idea. If you have a limited budget and you bid higher, you may get fewer clicks because you will exhaust your budget faster, like happened with this campaign mid-July. (Note: I didn’t do this ;) )
Testing bid changes can result in huge savings if done properly. In most cases you have a specific budget to spend and need to get a certain level of return on ad spend. Whether your desired result is traffic or conversions, you want to get the most for the lowest cost possible. If you have a bid set high enough to allow your ad to display in the first few positions for relevant keywords, undoubtedly you will get some clicks. But, at a lower bid you may be able to get more clicks if your budget is limited. Or, with a lower bid you may be able to reduce your cost per acquisition (CPA) without dramatically decreasing your conversion volume. The principle is simple but the execution isn’t.
The Sweet Spot
The “sweet spot” is the ad position where you get the most value. If you are optimizing based on conversions, this will be the position where you get an acceptable level of conversions for the lowest CPA. If you bid higher, you may get more conversions at an acceptable CPA but the CPA may be high enough that those additional conversions aren’t worthwhile.
For instance, let’s say you sell a product that costs $200 and you have a 2% conversion rate*. You may increase your bid and get additional clicks and that are still profitable for you. The increased revenue may be tempting:
However, those extra 5 conversions actually come at a very high price:
If you knew you could get a high volume of conversions at a $75 CPA you may decrease your bid. You would optimize your campaign to maximize the number of clicks you get at that lower CPC to increase the conversion volume. But, if you don’t know that you can get a large percentage of your conversions for a much lower cost, you may leave your bid at the higher level. This is why testing is so important.
To find the “sweet spot” where your cost per click is low but you are still generating an acceptable level of clicks you need to test many bid levels to find where you get your desired actual CPC and conversion rate. The best way to do this is using AdWords Campaign Experiments (ACE) so you only impact a percentage of your impressions with the bid changes and so you can easily undo the changes if need be. Once you’ve created your experiment you need to determine how much you are going to reduce the bids. There are two possible strategies:
- Start with small bid decreases to minimize the chance of a major shift in traffic. This tactic will require longer to test as you’ll need to run a test on the small bid change, report on the results then run another test with the bid decreased slightly more and report on those results and continue repeating the test until you find the bid at which you are generating the most clicks for the lowest CPC. This is the best strategy if you can’t risk a major decrease in click volume.
- Start with a large bid decrease to see an immediate impact, then if click volume decreases dramatically begin testing progressively higher bids until you find the sweet spot. This is the best strategy if you have a higher tolerance for changes in click volume and want to see faster results.
Things to Consider
- CTR. One concern I hear regularly when I recommend decreased bids is that the CTR will go down resulting in a decrease in Quality Score. AdWords normalizes CTR by ad position, recognizing that CTR will be lower in lower ad positions, therefore you are not penalized as long as your CTR is in line with what others are getting in that position.
- Impression Share. If you decrease your bids too low you may end up not getting any clicks because you aren’t bid high enough to participate in the auction. Keep a close eye on your impression share to make sure you aren’t losing too much with your decreased bids.
- Ad Copy. Different ad copy works better in different ad positions. You will want to test different ad variations in different positions to see what works best in your market.
- High Ad Postions Don’t Equal More Conversions. I have experienced ads in the top 1 or 2 positions having lower conversion rates, albeit higher traffic overall, than ads in lower positions. This may be due a host of reasons including people just clicking the first result without really looking at what the ad is for. As you test ad position you need to keep a close eye on conversion rate and bounce rate to see which position is optimal for your site.
If you’ve been following my posts on PPC you’ve no doubt noticed a trend which is testing, testing, testing. You should always be testing something in your ad campaigns and you should test every facet of your campaign to make consistent improvements in performance. Never be satisfied with “good enough” just because you are meeting your goals. When you meet your goals, create new more difficult goals and optimize until you reach them.
*For the sake of keeping the math simple we will assume a constant conversion rate although it does fluctuate and should be tracked along with CPC and position changes.