Reviewing my Adwords Strategy

[EugeneF]

Before I started working on Organic SEO, I did Adwords.  Being that I am entirely self-taught with adwords, I am concerned that about what I might not know.  From time to time I have looked for outside advice, but I never got very far.  It seemed that I would be talking to a salesman, someone who knew less than me, or someone who was horrified that I had 7,500 tracked words.  So I am presenting my methods here for comment.

Each keyword I have used is placed into an exact match campaign, phrase match campaign, and broad match campaign.  Exact match gets the highest bid, broad match the lowest.  Every week, I cull new keywords from phrase and broad results.  I also create negative keywords each week based on searches that are obviously not looking for my products.  Over time, I get very few searches triggered from broad matches.  Keywords are kept in ad groups that generally have less than 6 keywords.  Acceptable conversion costs are decided based on product category.  I have made an assumption that first position adwords results are too expensive (idiots overpay for them).   I automatically reduce my bid 10% on any ad that has an average position of greater than 1.5.   I automatically reduce my bid 10% on any ad that has a conversion cost greater than my target.   I delete keyword that have a CTR below 1%.  I generally require 300 clicks to make a determination.

All ads have been endlessly split tested to the point I don’t split test them much now.  I judged ads for both click through rate and conversion rate.

End result:  After 4 years of doing this, conversion costs are below target costs.  Not sure what else I could do to improve it, but looking for ideas.  Management of the system is pretty easy now, as well.

Anything I am missing?

[Justin_Vanning]

Hey Eugene! Yeah that is a common problem with Retargeting campaigns...it can be tricky to set them up properly so that you are removing people from your cookie pool once they convert. I'd say only about half of all advertisers do this correctly so that is why you continue to see ads even after you convert.

With keeping that in mind, I still believe that the benefits of retargeting far outweigh the negatives. I can tell you that the majority of the people who sign up for free trials with us every week have at least been exposed to one of our retargeting banners. Not all of them will directly convert off of the banner but the brand impression has a value that shouldn't be underestimated.

-Justin

[EugeneF]

Thank you for your reply.   I gave a more extensive feedback to the previous message.  I have seen the retargeting options and had not tried them.  I will look at these closer.   I always must remember that the way I search is not the way the average person searches.  I get retargeted by ads and usually find it funny that I am being marketed for services I have already received.  I personally make buying decisions within 48 hours of my initial search.  Targeting ME after that is a waste of time.

[EugeneF]

Thank you for your reply.

I have generally found much more cost effective conversion rates when I am not trying to get the first position.  I still GET the first position a lot, I just reduce my bid on it until I average below 1.5

It is an interesting idea to me to delete keywords based on quality score instead of CTR.  I had not done this because I have had some low quality score words that were converting well.  I understood that CTR effected quality score and that overall account CTR can have an effect.

I have not done much with conversion optimizer.  I seemed to have trouble with this as I had so many keywords that getting statistically significant data was tough.

I have only use GeoTargeting for the regions I sell to.  I have not used day parting.  I guess it is possible that behavior changes based on the time of day, I am just not sure why it would do so.   I have tried display campaigns, but only using contextual keywords.  I will look at this again.  I n the past, I was not able to get good conversion costs on contextual ads.  I have wondered if display ads can create a branding effect that is not measured.

[Justin_Vanning]

I agree with most of John's points. I always hesitate to start thinking in absolutes when dealing with PPC campaigns since they really should be something that you are constantly managing and tweaking. Remember that searchers can experience fatigue if you are no longer testing ads and may ignore your ads over time if they just see the same thing over and over. I like to constantly be testing to make sure that you are running the best possible ads.

One other thought...are you doing any retargeting/remarketing? If not, I would definitely recommend it. A retargeting campaign that is set up properly will be a great compliment to a PPC campaign. It gives you a chance to get your display ads in front of non-converting clicks from your PPC campaign.

Hope that helps!

-Justin

[john4math]

First, have you tested whether first position ads convert well enough for you?  You described that as an "assumption", which you could test to see if it's true or not.  Even if the cost per conversion is higher, the larger volume of customers may make up for it.  Generally, if there are people competing for the top spots, they are making money off the ads.

I'm a litle wary of having hard and fast rules for changing bids and deleting keywords.  For changing bids, I like to look at a fair amount of historical data to see how the conversion rate has been trending.  It could just be a bad week or a bad month?  After doing this for 4 years, you probably are aware of any seasonality that exists in your business that may apply to different keywords or ad groups.  For deleting keywords, I look a lot at the quality score.  There are some queries that users for whatever reason don't like to click on ads, so CTR might not be the best indicator that a keyword is bad, if it has a decent quality score.

There are a lot more features to Adwords you may or may not be taking advantage of.  Have you tried any geo-targeting, language targeting, device targeting, or day parting?  Have you tried conversion optimizer in any campaigns?  Do you have any display network campaigns?  There are lots of different targeting options for display campaigns beyond contextual keyword targeting.