Forum Moderators: buckworks & skibum

Message Too Old, No Replies

Opt Out Geotargeting is the way to go

         

ogletree

3:31 pm on Nov 19, 2014 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



If you want to Geotarget one ore more cities or metro areas the only way to get all or most of that cities traffic is to use "Opt Out" targeting. "Opt Out" targeting is when you target a local area and the entire state while excluding out everything in that state except the area you want to target.

For some reason a lot of people don't seem to notice this and even some argue it is not true. I have personally seen it in action and can say for certain it is true.

I have a very old campaign that has been running for many years that targeted Texas as a whole. I made some changes that increased our spend way too much but had great results. I found that the vast majority of our business was coming form 3 metro areas. I have always had Texas selected and last year I added the major cities in Texas. This way I could watch average position in each city and use bid modifiers to make sure my average bid was the same in each city.

Now that I knew I could spend all my money in those cities and pay a much lower CPA and get more business I decided the rest of TX needed to go. I removed TX and every other city other than my top 3. The next day my traffic tanked. When I compared my traffic from those 3 cities from the past my traffic had gone down 90% when it should have not gone down at all. The account should have gone down about 35% not 90%. I could understand a few percentage or even 10% but not that much.

I called Google and went on the Adwords community forum. Nobody could seem to understand what I was saying. I then found an article that mentioned that Marta Turek had spoken about this at SMX in 2013. I was so glad I was not the only one who knew about this.

I switched to Opt Out by adding Texas back in and then excluding out everything but the areas I wanted. (that was a pain). A week later I realized that I still wanted the TX traffic I just did not want to pay much for it. I then removed all the excludes and set the state with a negative bid modifier. Both ways brought my 3 main cities traffic back to previous levels. The exclusion method still brought in a very tiny amount of traffic throughout the state. The reduced bid method brought a lot more traffic at a much lower CPC and did not affect the average position very much. It only reduced my average position by 1.

My account is on steroids compared to what it was before all this.

earlpearl

5:50 pm on Dec 3, 2014 (gmt 0)

10+ Year Member



The research and article are very compelling. Over time, I've done some opting out. I hadn't really thought through it, but was simply eliminating some far flung areas in wide coverage, (similar to those described in the research). I don't have a city to state comparison...but over time by opting out of some widespread geo areas I've haven't lost much in impressions ...simply eliminated clicks and costs from areas that weren't generating conversions into sales.

I'm very comfortable with the strategy.

Sabrala

12:27 am on Dec 5, 2014 (gmt 0)

10+ Year Member



I feel this question is related if you would give me your best guess- I want to promote a local news website in a rural California area.

Many local IP addresses are from out of area and even assigned to Texas.

I still feel limiting my Adwords to the local counties would be best. We want to promote our real estate and health info- traffic more than clicks we are not converting anyone to subscribe we are ad revenue based- more traffic is the goal but we want people who will come back of course... our SEO is fine (webmaster tools and site maps everything submitted to Google News regularly etc)

Current ads are set to the entire US in 4 different campaigns- I have no data on just CA and everyone is advising against limiting it.

Should I select all of California and Nevada and then *opt out* all counties far north and south of here? There has to be case studies on California? We wouldn't mind letting SF know we are here but LA could care less. Sacramento and Tahoe are debatable but Mendocino, Redding, Los Vegas and Southern CA will not need our content...until they do and then they will find it other ways.

Are their industry regions already established to define the very large state of California? I will do the testing but I do not want to re-invent the wheel!

Sabrala

12:53 am on Dec 5, 2014 (gmt 0)

10+ Year Member



Ok sorry I found the DMA areas I want 3 of the 14 in California. Should I opt out of 11 or opt in the 3? Thank you!

earlpearl

5:25 pm on Dec 5, 2014 (gmt 0)

10+ Year Member



Sabrala: The best response I could give is to test. The presentation that was highlighted at an advanced search session, tested in Colorado, then limited a test to Denver, then readjusted to all of Colorado and opted out of areas way outside of Denver.

Results were far better by Including all of Colorado and opting out of far flung areas.

But that was one test. If you were able to set this up 2 ways for test periods; and if you didn't think you were going to get impacted by seasonality I'd try a test.

I'd spend a couple of weeks with a more narrow geo focus...and then a couple of weeks including the entire state while opting out of various far flung and remote areas.

See which method gives you preferable results.

One caution...when you switch to opt out and are in the process of excluding counties, dma's etc...be sure at the end...you're overall campaign is still geographical in nature and has not defaulted to a general non geo campaign.

My experiences are similar to the printed experiment. I start with a wide area and opt out of narrow, far flung, and areas that don't give a return.

What I don't have is data on the differences of a deliberately narrow geo campaign compared to the wider geo campaign wherein far flung areas are opted out. The data from the test was very very interesting and suggested the opt out strategy was dramatically preferable.

Man, if I had the luxury I'd test it for myself. Really interesting results.

Sabrala

5:31 pm on Dec 5, 2014 (gmt 0)

10+ Year Member



Thanks for your encouragement it is not easy to promote the idea of testing in my business but I believe I have enough of a reason to insist we try. Thanks again

GabGoldenberg

9:12 am on Dec 11, 2014 (gmt 0)

10+ Year Member



Going to try this out - thanks for sharing Ogletree.

Question: If my negative bid adjustment results in a super low bid for the traffic I don't want, does that defeat the effectiveness of that method? Should I then try to exclude every other location?

Related question: How would you apply this method in a case where you know that many ad networks' data on geo targeting is off. I live in Jerusalem, Israel, and get shown ads and landing pages that think I'm in Tel Aviv, Petach Tikva or elsewhere in Israel.