Downtrodden Adwords Quality Scores -- Really?

[groovykarma]

Wow! 93% of our Adword keywords in our new campaign received a Quality Score of 4 or less.  That means most of the keywords aren't showing.

I received the quick answer from our Adwords advisor that, "Quality score is created from a variety of factors ... etc."  Yes, I know that by reading Google's documentation.

I dug deeper into the data.

When I looked at keywords dashboard for this campaign, what vexes me is that it's all about "Keyword relevance: poor". That is repeated time and again in the keywords hover, bubble pop-up in Adwords.

"Landing page quality: no problems".

"Landing page load time: no problems".

63% of keywords have quality score = 3

29% of keywords have quality score = 4

We have thousands of keywords that are electronic part numbers.  All keywords use phrase matching.  We use dynamic keyword matching in the ads.

I dug deeper.  I chose random keywords (and corresponding landing pages) from lower quality scores (1,2,3,4) and higher scores (5,6,7,8,9,10).  What is the difference?

Examples:

• Quality score 1 = http://www.usbid.com/parts/LM2901
• 2 = http://www.usbid.com/parts/BZX84-A20
• 3 = http://www.usbid.com/parts/MAX4796
• 4 = http://www.usbid.com/parts/TMP302A
• 5 = http://www.usbid.com/parts/LTC4267-3
• 6 = http://www.usbid.com/parts/CAT1161LI-28-G
• 7 = http://www.usbid.com/parts/DS1216C
• 8 = http://www.usbid.com/parts/R10S-E1Y1-J5.0K
• 9 = http://www.usbid.com/parts/208M822-19B11 145785-000
• 10 = http://www.usbid.com/parts/TP3-PNEU-0.250 243362-000

Notice URLs with score 9 and 10 have url-encoded space (%20) -- just pointing it out.

What is the difference between these pages that have such different quality score?

And, interestingly enough, the majority of example keywords in the Urls above (LM2901, BZX84-A20, etc) have zero impressions and zero clicks thus far.

Yes, the keywords have low traffic because this is exactly what people search for an purchase when making a B2B component buy.  It's all about the exact part number.

**I'd love specific suggestions of how to improve quality score of pages with a 3 or lower! ** Thanks kindly, Loren

[DavidKonigsberg]

What are your ctr's? How long Has your account been running? Also check search queries to see if the phrase match is bringing in relevant searches

[groovykarma]

Hello David,

The whole account has been running for 2+ years.

This particular campaign was started yesterday -- alongside several campaigns that have been running for months.

CTR average = 0.63% (pretty lousy)

For today's data, 187 distinct keywords had clicks.

28% had CTR >= 50%

34% had CTR 10% - 49%

37% had CTR 0.07% - 0.09%

The median is 20% so that means a lot of the CTR are at the lower end.

We are in a loooooong tail business (with lots of part numbers) so the higher CTR are for 1 impression total, 1 click total, etc.

Lots of items with impressions and no clicks.

Search results are relevant given these are manufacturer part numbers.

**Great questions.  Did I answer what you sought? **

I'd love additional thoughts from you.  Kindly, Loren

[DavidKonigsberg]

Do your old campaigns have bad scores also?

[groovykarma]

I love that you're helping me dig deeper.

I crunched quality scores (export and into Excel) for two longer-running campaigns that get an acceptable number of leads/conversions for the cost.

Older campaign 1 : average quality score = 4.8

Older campaign 2 : average quality score = 4.2

So those aren't great either (imho) yet better than the current campaign.

More questions to help you noodle on my situation or do you have gems to share on how to raise the quality score?

Cheers, Loren

[DavidKonigsberg]

I find it hard to answer these sort of questions without seeing  the account from my experiences new accounts sometimes show bad scores when you first start a  campaign. But if you see it across the account i would try to make sure the ads relate to what the people are searching for as best as possible etc this post might help you as well - http://support.google.com/adwords/bin/answer.py?hl=en&answer=2454010&from=10215&rd=1

[flowsimple]

Loren first off in your number crunching, I'd suggest you calculate a weighted average of your Quality Score, whereby you take into account the number of impressions for each keyword.

This is my general thought based on limited knowledge of your account:

The most tell tale factor is that you have an aged account of 2+ years with a lousy historical CTR. This is going to greatly impact the QS of any keywords you add to your account no matter how relevant (or how much better you've gotten at creating relevant campaigns). Your predicted CTR is terrible, so your QS reflects that.

There is an implied volatility, and what I haven't been able to figure out is, how much change it takes to impact your QS. I've noticed that if your future CTR is really good, and your keywords have high volume impressions and clicks, the QS can improve very quickly (sometimes within days). But if the new keywords you've added are low volume, it can take a long time before your QS improves to reflect the underlying relevance.

George Michie of RKG wrote a detailed blog (rant) post about Quality Score, and while it doesn't directly answer your question, it does educate you on what goes on in that "black box."