Smart Geographic Segmentation & Bidding With Enhanced Campaigns
New PPC Best Practices In An Enhanced Campaign World was one of SMX Advanced Seattle’s most engaging sessions; and, as a speaker, I have received a great deal of feedback and questions about those bid location modifiers. As it turns out, location bid modifiers are one of the (only) benefits of upgrading to enhanced campaigns, […]
New PPC Best Practices In An Enhanced Campaign World was one of SMX Advanced Seattle’s most engaging sessions; and, as a speaker, I have received a great deal of feedback and questions about those bid location modifiers.
As it turns out, location bid modifiers are one of the (only) benefits of upgrading to enhanced campaigns, as they can be applied by location in a very granular and simple way. I thought I’d recap a couple of thoughts about geo-segmentation and location bid modifiers.
Segmenting Your Paid Search Account By Geography
The segmentation tips below can help improve the performance of your enhanced campaigns.
- Segment by country for language targeting/landing page purposes – this pretty much goes without saying. Note that some folks – more specifically Marta Turek from Mediative — are testing a different way of geo-targeting wherein you exclude everything but a specific location, as opposed to targeting a specific location. Early findings show some interesting results; however, this workaround might need some more in-depth testing before any general conclusions can be drawn.
- Segment by time zone for day-parting optimization purposes. For instance, in the US, it makes sense to segment your top campaigns by time zone to be able to set accurate day-parting modifiers and leverage those peak hours locally.
- Segment by physical store for custom messaging purposes (if applicable). This is a time-consuming task, but necessary if you want to have full control over the ads served in specific locations.
- Leverage location bid modifiers at the campaign level (they are not available at the ad group level). That will allow you to bid more aggressively in the best-performing locations — and, conversely, save money in poor-performing locations — without altering the actual account structure.
Leveraging The New Enhanced Campaign Location Bid Modifiers
In a previous post, I explained how to determine your location bid modifiers in general; and, while that information is still relevant, you might want to leverage multiple levels of geo-segmentation depending on how granular your statistically significant geographic data is. In AdWords, you can set your location bid modifiers by state, congressional district, city, postal code, Nielsen DMA region, and a couple of other levels.
Often, the main challenge here is to find the right balance between geo-targeting granularity and data dilution. In other words, you should be able to set location bid modifiers in a very specific manner as long as you have collected enough data. This is going to happen more in bigger cities but not so much in more rural areas.
So, you can expect your location bid modifier map to be similar to a population/city map. In the below example, we’d potentially have state-level bid modifiers as well as city-level bid modifiers for all major cities.
When using multiple levels of geo-segmentation is that in the case of overlap between multiple locations (cities within states, for instance), Google will use the most granular location and bid for that particular location.
[Note that if you have other parameters that overlap (such as a bid adjustment for mobile devices, and another for the location of Los Angeles), Google will add the adjustments to come up with the correct bid for a mobile device in Los Angeles. So, if your starting Max CPC bid is $1, and you set a +20% adjustment for mobile devices (bringing it up to $1.20) and a -50% adjustment for Los Angeles, then your resulting bid would be $0.60 — with the mobile devices raising the bid and the Los Angeles location lowering that adjusted bid.]
So, assuming you want to set location bid modifiers at the city or state level, here are my recommendations:
- For those states with a statistically significant amount of data, you can determine your state-level bid modifiers using the formulas from my previous post.
- For those cities with a statistically significant amount of data, you can determine your city-level bid modifiers.
- For those cities without enough data, you can let the state-level bid modifiers do the job.
Example: For a given campaign targeting the US where 100 clicks are considered to be statistically significant (see below screenshot), we saw:
- Campaign-level: 48,779 clicks and a CPA of $1.49 in the US
- State-level: 8,603 clicks and a CPA of $1.85 in California
- City-level: 1,433 clicks and a CPA of $2.87 in Los Angeles
- City-level: 464 clicks and a CPA of $2.19 in San Francisco
- City-level: 94 clicks and a CPA of $2.09 in Long Beach
As a result, below are the location bid modifiers you might want to use:
- California: (Campaign CPA/California CPA)-1= ($1.49/$1.85)-1= -19%
- Los Angeles: (Campaign CPA/ Los Angeles CPA)-1= ($1.49/$2.87)-1= -48%
- San Francisco: (Campaign CPA/ San Francisco CPA)-1= ($1.49/$2.19)-1= -31%
- Long Beach: since there were not enough clicks in this location, you can let the state-level bid modifiers take care of the bid adjustments.
This logic can be applied at any other geographic level (metropolitan areas, zip codes, etc.) — and, as usual, it can be automated based on a geographic report from AdWords.
Geo-targeted segmentation and bidding tactics are supposedly getting simpler with Enhanced Campaigns — as long as you know how to go about it.
Search marketers need to evaluate their needs and constraints from a budget allocation, local messaging and efficiency standpoint, then decide what’s the best editorial and bidding set-up for their accounts. Additionally, search marketers need to be cautious with the way enhanced campaigns’ hourly/mobile/location bid modifiers stack.
Editor’s note: This column has been changed from its original version to correct some issues that came to light after publication.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.