Position Bidding vs. CPC Bidding: A Comparison Of Two Bid Optimization Methods
There are two competing approaches to bid optimization that I’ve found are most commonly used by top PPC analysts. There is the approach that I primarily use, which determines optimal bid levels to reach an ROI/volume goal. I’m going to call this ‘Direct ROI Bidding’. The other approach determines optimal position to reach an ROI/volume goal. Then bids are adjusted to get to that position. I’m going to call this ‘Position Bidding’.
If both of these approaches get you to the same goal, what’s the difference? Is one method any better than the other? Let’s break each strategy down and compare.
What Is Our ROI Goal?
I often feel a little silly asking a client what their advertising goals are. Almost every advertiser has essentially the same goal – drive the maximum profitable revenue for their ad dollars. But there are a lot of different paths to get there, which is what I’m really trying to understand. A consumer packaged goods advertiser might need to run a branding campaign optimizing display ad impressions so consumers will be swayed by brand recognition next time they’re in the grocery store. An ecommerce advertiser would be looking to optimize tracked revenue returned from ad spend. ROI could be any number of metrics based on the business model of the advertiser. For the purposes of this article, I’m just going to talk about optimizing to a CPA. The same concepts will apply whether we’re talking about optimizing for CPA, ROAS, or even traffic.
Approach #1: Direct ROI Bidding
There are two ways we can optimize bids directly to a CPA goal – a rules approach and a formulaic approach. The rules approach simply increases bids up or down when CPA is below or above target. A workable method, but it doesn’t provide much guidance on how much bids need to be raised or lowered. So I’m going to focus on the more robust formulaic approach instead. The formulaic approach looks at historical conversion rates in an effort to predict future conversion rates. Once a future conversion rate has been predicted, analysts can determine the highest CPC they can pay without exceeding their target CPA by multiplying their CPA goal by the conversion rate prediction.
Max Allowed CPA X Predicted Conversion Rate = Calculated CPC Target
(Don’t forget to factor in a bid gap)
See Wijnand Meijer’s bidding article for one example of this method.
Assuming your conversion rate prediction is solid, this approach will put your ad in the highest position possible without exceeding your target CPA. I know this because ad rank is determined by multiplying your max CPC bid times your quality score. And ads are positioned based on their relative ad rank against other bidders. So position may change from day to day based on the competitive bidding landscape, but you will always be getting to the highest position possible based on your conversion rate prediction and your target CPA.
To illustrate this, consider an advertiser wanting to drive maximum conversions without exceeding a $10 CPA. A high volume keyword has been consistently converting at 5%, which will become our conversion rate prediction.
Max Allowed CPA: $10.00 X Predicted Conv Rate: 5% = Calculated CPC Target: $0.50
As you can see in the table below, competitors bidding up and down on the same keyword will move your relative ad rank and position up and down the page. Click and conversion volume will move up and down right along with position. But assuming the conversion rate prediction was accurate, we are always driving the maximum number of conversions without exceeding our Max CPA.
An increase in max CPC to attain a higher position during any period would result in exceeding your max allowed CPA. A decrease in max CPC will lower your CPA, but miss out on volume. As you can see, the direct ROI bidding method essentially removes the need to think about position. Who cares if you’re in position 2 or position 8 as long as you’re getting the most conversions possible at your maximum allowed CPA.
Direct ROI Bidding Advantages:
- Easy to apply to high volume keywords.
- Easy to automate with simple bulk sheets and Excel.
- Naturally adapts to changes in the competitive landscape. Position fluctuates while ROI remains in line with goals.
- Does not factor in variances in conversion rate by position. . . to be discussed below.
Approach #2: Position Bidding
The position bidding strategy assumes there is a particular position where each keyword will convert best, providing the best ROI. In this video post by Sean Quadlin, he shows how performance data by position can be analyzed to determine which position has historically driven the highest conversion rates and lowest CPA.
To summarize this strategy, an analyst could report on conversion rates and CPA by position for a group of keywords as shown below. These numbers were pulled directly from the video and rounded.
Assuming we have a statistically significant sample size (i.e. click volume), we can predict that position 5 will lead to the best CPA. An analyst could then setup a bid rule to adjust bids towards position 5, and let the position define the CPC he pays.
One advantage of position bidding is that it can identify conversion rate variances by position. Direct ROI bidding operated off an assumption that conversion rates will be relatively consistent across positions. But if conversion rates vary by position as shown in Sean’s data above, the direct ROI bidding strategy could lead you to grievously poor results by comparison.
That said, I must point out a couple issues with the data used in the video post. We should note that the date range of the original report spanned a six-month timeframe. If the competitive landscape changed significantly in this time period, it would go unnoticed. Second, average CPC is lower in position 2 than in position 3 and 4. This indicates that either a different concentration of keywords made up most of the position 2 clicks or there was a significant change in competition over the report timeframe. Sean was just using this data for demo purposes, but we should all watch out for such issues when running an analysis like this.
The biggest disadvantage of position bidding is it is extremely difficult to apply at the keyword level. For a single keyword, if I look back at 30 days of performance data, I only see a single average position. Not helpful. So, we might start segmenting the data into days to get something like this…
You’re probably already seeing the first problem. Unless you’ve been changing your bid significantly over the date range, you’re probably showing in a small range of positions, like the above example where position is only ranging between 2.4 and 3.1. The only way I can determine if position 1 or position 4 might perform better is to change bids and test these other positions. And that takes time and lots of bid management effort when working with thousands of keywords. Better have some serious bidding automation in place!
Another issue is the increased need for volume. A keyword with 500 clicks over the last 30 days has enough volume to make a good conversion rate prediction using the direct ROI bidding method, but now volume is segmented into days and positions. The same keyword might only have had 10 clicks in position 5 over the last 30 days, making our margin of error at this position too large to suffer. And lastly, if I have to look back over 90+ days of historical data, am I really identifying the best performing position for how the market is behaving today?
In short, Position Bidding is not a good optimization method at the keyword level.
Position Bidding Advantages:
- Addresses variances in conversion rate in different positions.
- Extremely difficult to apply at the keyword level.
- Must proactively test at different positions to create a workable dataset.
Bidding to Maximize Profit
There is one situation where position bidding decidedly trumps direct ROI bidding – maximizing profit. Say I have an ecommerce campaign where I’m passing back profit margin instead of total sale value in my tracking script. The advertiser’s primary goal isn’t really to drive as much profit possible at a specific target ROAS. The advertiser wants to drive the maximum profit, regardless of ROAS. And if you test profit levels at varying ROAS levels, you will invariably get a trend like in the graph below.
At the lowest ROAS levels, we will always see low profits. Certainly the further we drop below a 100% ROAS, the further your profit value will drop below zero. At the highest ROAS levels, we will also see low profits because the low bids necessary to drive a high ROAS will eventually reduce our conversion volume to nothing. Somewhere in the middle will be a peak where we are maximizing total profit dollars. You’re thinking to yourself, “Wow! That’s right! Why aren’t we all doing this instead of targeting a specific ROAS?” Simple answer. . . because it’s hard. It requires us to predict the rate at which click and conversion volume changes at different bid levels (and/or position). This rate is neither linear nor consistent. It will invariably involve deriving a trend line from a scatterplot of inconsistent CPC-to-Clicks relationships. Furthermore, this is a rapidly moving target as your competitors change their bids and thus your CPC-to-Clicks ratio.
I will be discussing profit maximization techniques in detail in a future post. But for now, let me just point out how position bidding has an advantage here. Position bidding already relies on testing performance (including the necessary click & conversion volume) at different positions. Therefore we can at least see that testing performance by position will get us closer to identifying this point of maximum profit than the direct ROI bidding method.
As would be expected, my comparison came down decidedly in favor of my preferred direct ROI bidding method. But that doesn’t necessarily mean it is the best approach. I’m hoping that position bidding advocates are reading this and are ready to poke holes in my comparison. Please poke holes! What are other shortcomings of direct ROI bidding? Are there effective ways to overcome some of the disadvantages I identified in the position bidding method? What other advantages might we need to consider with position bidding. Healthy debate will lead to a better understanding for us all.
Browse By Category
Social Media Demographic Strategy Tips for 2017
Don't miss out on paid social opportunities. Get tips to guide your social media demographic strategy for 2017!
Using Data to Create a Killer Business Story For Your C-Suite
In this new live webinar, Hanapin’s Jeff Baum and Optmyzr’s Frederick Vallaeys will show you how to formulate the perfect business story for your C-suite, and how to explain the vital need for PPC success in your company.
A bi-weekly newsletter packed full of resources and strategies that will help make you a better PPC expert.
Hanapin Marketing | The PPC Agency of Experts Behind PPC Hero
How To Think Straight When Your PPC Account Is Going Sideways
Learn the good starting points when troubleshooting a PPC account that suddenly went off the rails.
How Storytelling can Up your PPC Reports
Learn how to use storytelling to increase the effectiveness of your PPC reports. Start streamlining PPC reporting and save time!
Pitch to Speak Now: Hero Conf LA is Calling!
The deadline for speaker pitches to Hero Conf Los Angeles is midnight Friday, December 2. If that's enough to remind you to get writing, then head on over and pitch now!