AdWords filters allow you to make smart bidding decisions en masse on your AdWords account.  Unlike ad group breakouts, long-tail keyword expansion and proper ad copy tests, bid adjustments are a quick way to create a positive impact on your AdWords account.  Too many advertisers leave money on the table by leaving high-performing keywords in low positions and overspend on underperforming keywords in high positions.

What is your bidding strategy?

This is a post on using AdWords filters, not on overall bidding strategy.  That said, it’s important to clarify what we’re accomplish with bids.  Bids, in combination with Quality Score, determine your ad’s placement on the SERP.  Higher position represents more exposure, and generally more clicks resulting in more sales.  Lower position on the SERP represents lower position, less clicks and less overall sales.

While your exact bid strategy is contingent on goals, the overall goal is basic: use bidding to increase the exposure of high-performing keywords and decrease the exposure of poor-performing keywords.

How to define bid rules?

At a minimum, I would advocate using three levels of filters for bid adjustments.  For simplicity, I’ll call these the:

  • Baseline Metric: The baseline metric defines what is a high-performing vs. underperforming keyword.
  • Volume Metric: The volume metric provides a sample size control to allow you to dictate if you’re comfortable bidding up keywords that have converted once vs. requiring multiple conversions.
  • Position Metric: The position metric is an indicator of what you consider to be high/low position on the page.

Baseline Metric

If you are evaluating keyword bids within a non-brand campaign, then a good starting point is the average CPA for that campaign.  This will ensure that you’re only raising bids for keywords performing above average relative to that campaign.

If you are an eCommerce advertiser that is tracking revenue in AdWords, I would recommend using an ROI-metric (ie – conv. value/cost) to evaluate keyword performance.

Using the campaign’s average CPA or ROI is helpful if your goal is to decrease average CPA or increase average ROI per keyword.  However, if your goal is to maximize conversions or revenue above a minimum profitable CPA or minimum ROI, then you should input that figure.  For example, if a tennis racket retailer has a profitable ROI of 4.0, but they’ve been operating at an average ROI of 5.5, they may still want to bid up all keywords that are in low positions at an ROI of 4.0 or higher.

Volume Metric 

For the volume metric, it makes sense to set a filter so that you are only evaluating keywords with multiple conversions in the given time range.  This is to ensure more reliable data and a larger sample size for the keywords on which we’re actually going to make bid changes.

If you are working within a small account with very few multi-conversion keywords or just wish to be more aggressive with your bids, you can consider evaluating all keywords with a conversion.  If you have a larger account with a greater number of conversions, you may also want to set a higher threshold.

Position Metric   

The position metric is just a gauge of where you want to set a cutoff point for positioning.  Often times I’ll concentrate on an average position that’s around 2.5-3.0.  You can influence a big change in volume by making sure your ads are showing up more consistently in the three spots on the top of the page.

Note that it’s also possible to segment so that you increase bids 10% on keywords that show up in average position 2.0-3.0 and 15% on keywords that show up in average position 3.0-4.0.

Results 

Using this example, here is a retailer who is profitable below a $12 CPA.  With filters, we are able to identify a ton of profitable keywords that could be bringing in more sales from a higher position on the page.

The same methodology can be used to lower bids on poor performers.  In the example above, the retailer might look-to identify all keywords with a CPA > $15 in positions 2.0 or higher.  Similarly, they could create a filter for keywords they would consider pausing with:

Conversions = 0

Clicks > 100

Avg. Position “better than” 2.0

For insight into how to make these bid changes efficiently, see this post from Google on making advanced bid changes in AdWords Editor: http://support.google.com/adwords/editor/bin/answer.py?hl=en&answer=47655