Paid Search Track
Decrypting Quality Scores - Life used to be so easy. You paid more than your competitor, and you ranked tops as a result. Now the perceived "quality" of your ads can make a dramatic difference in how much you pay and are positioned. Unfortunately, since the exact formula of calculating quality scores isn't revealed by the major search ad networks, marketers have some deciphering to do. This session looks at ways to increase the quality scores of your ads. Moderator: Chris Sherman, Executive Editor, Search Engine Land Q&A Moderator: Anton Konikoff, Founder and CEO, Acronym Media
Speakers: Andrew Goodman, Founder and Principal, Page Zero Media Jon Kelly, President, SureHits
Q&A Speakers: Mary Berk, Senior Product Manager, Microsoft Nick Fox, Director Business Product Management, Google David Pann, Vice President, Marketplace Design and Matching, Yahoo!
Andrew Goodman is up first.
Before we get into detail, think like a search product manager on teh side of the user trying to determine what a high quality ad is: Low quality is bad experience on site - sites or pages that mislead users. Also, low quality is misdirection or poor relevance - advertising on irrelevant keyword or keywords that are pretending to be relevant. See how simple it is? Nearly everything they're doing can be subsumed under those two categories. The old formula that only used CTR to rank ads (along with bids) isn't sophisticated enough to achieve these objectives.
Three generations of paid search ad ranking - Goto.com/Overture: pure bid for placement - AdWords .20: Max Bid X CTR - various approaches - AdWords 2.5, 2.6: quality based bidding.
How it works - there are two quality scores. Minimum bid decides whether the keyword is active at all. The other quality score affects ad rank.
There are two distinctive types of quality 1. The overall formula is composed of signals of relevance and quality related to your keywords (their CTR history and more). 2. A sometimes hypercritical assessment of your website and landing pages.
You can diagnose your situation with the Google AdWords tool.
People thought that Google invented this to avoid arbitrage. Clearly stopping the bad guys was a goal.
The sandbox effect on the organic side requires a lot of money. The "dig out" process: - Media company wants to buy traffic and launches new verticals. - Past high quality not helping - many QS starting low and can't figure out why - Also, slow ad activation on timely news items - Multiple theories for low quality scores - Frustrated - new ad experiments - Frustrated - tips from Google rep - Uptick - some history establishing itself. Why are we getting low quality scores? Promise of a review by the technical team or vague theories followed by something else...
What is it? - Did we build the right thing over time through testing? - Did we cause someone to look at it? Is it manual since we alerted our Google rep? Account-Wide effects? Think in terms of your account history - as it becomes more mature, there's a halo effect.
How to be careful when setting up new accounts - CTR is still your anchor How to be careful when setting up new accounts - Consider how much QS can tell you about how Google views you - Anatomy of a high QS kickoff - step by step * Narrow keywords * Hand built ads * Tailored landing pages that work * Geo-specific helps * Segmenting and microtargeting
Funny stories: watch out for popups!
Jon Kelly is up next.
Ad quality is essential to the economics of search engines. Because of that, it's essential to your campaign. If you have puzzling results, you're not alone.
Quality score affects ad rank and minimum bid. Let's look at ad rank - Quality score brings the value of the CTR back into the engine and builds an effective CPC value for the engine. CTR x Bid Price X 1000 = eCPM for the Search Engine ex: $10 bid, 5% CTR, 50cents/impression, $500 eCPM for the engine. Search engines have a goal to sell the space for as much as possible so moves the highest eCPM bid on top. Landing page quality maximizes long term eCPM.
Key QS factors - relative CTR by position, account history, and landing pages (minimum bid)
From a math perspective, beat expectation.
Total throughput management = your key metric is value per impression: Click Through Rate x Conversion Rate = Total Outcomes / Impression for any given ad Why? Your 2 key metrics are often at odds. "Qualifying" words raise conversion but may lower CTR. "Enticing" words may raise CTR but may lower conversion. Example: "cheap" is an enticing word. But is it cheap? Maybe you'll get a good CTR but will you get a conversion?
Throughput checklist - - Track total throughput (CTR & conversion together) for all ad copy variants - Make sure you isolate the variables - Track the $ sales only if creative changes impact product choice or quality - Creative should match the scope and quality of your offer - Match landing page with ad copy - Experiment with different grouping structure - Homogenize ad groups - Only exact matches count (regardless of what you have set up) - Use day parting to focus on high demand and conversion periods
How the engines can improve * More transparency is needed into the scoring method - other factors, for example??
Myth: you can't succeed if you don't have a major brand. That's not true. You can.
Q&A: Q: What do you do if you have an abysmal quality score? Should you start fresh? David/Yahoo: We've seen that this is because of keyword selection. Find relevant keywords and you won't be hurt. Do some AB testing. Nick/Google: Check from a user's perspective and see if it makes sense. We have an ad score quality diagnostic tool. Mary/Microsoft: That check is the best thing you can do. Scrutinize your site to see what the end user sees. A ton of advertisers don't realize that they have popups, paid links above your content, and simple things like that.
Q: How many clicks does it take to establish that initial quality score so that the campaign is considered mature? Mary/MS: It determines on the search engine. David/Yahoo: It depends if you get clicks. Tail terms need fewer versus other terms. We want to give all advertisers a chance. Nick/Google: It depends on the keyword and the situation. If the keyword has a history of being bad, it will be harder for you.
Q: What do you think about DKI into landing pages? Can this be flagged during a manual review? Jon: I've found success with this method. Nick: It's a great way to add relevance to what the user is looking for. But be careful of two things. Don't insert keywords that create nonsensical ads. The other thing is that you should deliver people to a page that is relevant to the ad text.
Q: How would quality score optimization vary for search versus content? Nick: Great question. You need to think as organic search differently from paid search.
Q: Should you create separate landing pages for each search engine? Jon: No. Andrew: It's a slippery slope. David: User experience comes first and foremost.
Q: What about the black box element with lack of transparency? Are there any tools that you recommend? Jon: Focus on your analytics. Nick: There are 3 Google tools. One project is called Shine the Light on Black Box. On your campaign management page, you can get more information on your campaign. (Nick is talking way too fast so I can't get his second tool nor can I really get what he was saying on teh first tool. He is also really quiet.) There's also a search query report to see what keywords are more relevant.
Q: An advertiser is unhappy with their quality score. How do they escalate the issue for human review? David: Use your account rep. Mary: If all else fails, contact our editorial team. Nick: Check if there's a landing page quality issue.
Q: How can I reverse engineer my competitors' Quality Score? Jon: Look at the ad rank - it's the ad and click through rate, not the landing page. You need to worry about ad rank.