DO NOT pause or remove ads without statistically significant data

Avoid conversations with this guy about PPC ads. At my first paid search job I would frequently get in arguments about what ads were “better” and what people liked. That is a subjective argument you will never win. Enter your best friend in PPC, data and testing. You no longer have to have these discussions with anybody once an ad shows a statistically significant higher click through or conversion rate compared to the other ad or ads in rotation. There are multiple tools available to plug your data into and determine which ad has a statistically significantly higher click through rate or conversion rate. I prefer this tool from Web Share. Adwords also is making progress in the area of testing within their platform. Seer Interactive offers a great tutorial on how to implement ad testing within Adwords. Let data and performance make these decisions for you.

DO NOT track more than one ad element at a time

This is a VERY difficult one to practice. I will be writing ads and have multiple new elements I want to test. Once you add multiple variables to test your results become difficult to analyze. Elements in paid search usually are one of these:

• headline
• description line
• display URL
• punctuation
• capitalization

There are multiple elements to test but PPC ads have a limited number of characters to work with. If you are testing only one element at a time it simplifies the process of telling what’s working and what’s not. These learnings can better inform writing ads in the future to improve results.

DO NOT have more ads in an ad group than allows you to get testing results in a reasonable amount of time

A good rule of thumb is your ad group should receive a minimum of 200 clicks a month for rotating two ads in the ad group. So, if your ad group gets 300 clicks it’s OK to rotate three ads. Most of our clients do not have the volume for more than two ads per ad group in rotation. It’s important to ensure ad groups have enough volume for you to reach statistical significance in your testing. If you have too many ads in an ad group based on the volume you are unable to make data informed decisions with your ads.

DO document learnings (winners, losers, and detailed notes)

Keep some type of testing document or spreadsheet. Keeping detailed documentation of your ads allows two things:

1. You know what is working and you can expand those ads to other campaigns and ad groups
2. You also know what is NOT working and you can learn to not repeat that ad copy in the future

After pausing or removing ads I try and do a summary and analysis of what I think was working and what was not. I prefer to have a spreadsheet with two tabs. One with new ads along with the published date. The other tab includes removed or paused ads. Each tab follows the standard columns in Adwords with-headline, description line 1, description line 2, display URL, and destination URL. Here’s a screen shot of one of the tabs:

After determining winning and losing ads I will update the new ads tab by highlighting the winning ads in green and the losing ads in red. I also will have a column for notes where I will add comments of what the conclusion of the test ad was. I’m sure there a many ways to do this process but what’s important is you document it and have a system to learn and optimize your PPC ads.

DO Set up a document or place for brainstorming and stage future ad tests

Writing ads is one of the more creative areas of paid search. Many times when I’m writing ads I will have many good ideas but then have to go back to my earlier resolution of only testing one element at a time. Take all these great ideas and place them in a document to use next time you have to write ads. I’d also suggest you provide detailed notes on the logic behind the element. I’ve documented ideas only to come back to them later and don’t understand my cryptic notes or vague references. Putting all of these ideas in one place also allows you to have an area to stage and prioritize future tests and elements. Because we collaborate in a team environment we typically use either a Google Doc or Basecamp for this document.

What tips do you have to share about an effective PPC ad testing process? Share in the comments.

This past month was difficult to find blog posts that were NOT about Google +. But don’t fret, I’ve found the best the web has to offer in PPC and listed them below.

The biggest theme that stood out to me from this month’s playlist was testing. Ongoing testing and optimizing is critical to the success of any PPC campaign. So put on your lab coat, read up, and get er’ done.

Amplified Gold Posts

• 5 ways to increase CTR with your PPC ads.
• Brad Geddes provides a quick way to identify ways to increase exposure of you PPC campaigns.

Amplified Platinum Posts

• Choosing a PPC Agency: 10 Questions to Ask A great list of questions you should be asking. I wish I had this list several years ago when I was on the other side of agency PPC management.
• Data is a powerful tool when used properly in PPC. This post shares 5 powerful PPC tips to make sure you use data correctly to improve your campaign performance.
• Remember the days PPC conversions were cheap and competition was non existant? I do. If you are in a semi competitive niche this is usually no longer the case.  Here’s a good post on why “You Can’t Buy More Conversions“.  Bonus is there is a pic of a terrifying baby after the jump. I’m pretty sure I’ve seen that face from past managers.

Amplified Diamond Posts

• I LOVE THIS POST! Bryan Eisenberg outlines a simple set of 3 guiding principles to lift conversions and click through rate . Writing PPC ads can be  the most difficult part of a PPC analyst job. Use these three rules and make your life easier.
• Another gem from this month. Wordstream’s blog is a must read for PPC peeps. Check out this post that provides estimates on average PPC bids ranked by industry. This explains how Google can justify continued failed attempts at social networks like Buzz and Wave. Let’s hope Google + doesn’t go that direction. But if it does I’m sure it won’t be Google’s last try.

Chime in the comments with any posts that you thought were required PPC reading in the last few weeks. Or if you are looking for PPC management services give us a shout.

In the spirit of the playlist theme I’ve queued up a YouTube playlist of 3 songs that have got me chair dancing in the last month. There’s some new stuff with a classic at the end. I recently added a Beastie Boys station to Pandora and have been loving it. I’ll plan on queueing up new songs for each playlist post in the future.

Image by shutrbug72

Several years ago I had a friend majoring in accounting and had started a book keeping business on the side. I once had a conversation with him where I asked “How do you enjoy working with numbers all day?” and his response was ” I love working with numbers, there is so much to be learned” and I thought to myself “what a nerd I have for a friend and what a boring career choice”. It’s funny to look back on that conversation and think how much my mind has changed and also grown to embrace the power of numbers and data. A quote that resonates for me from the blogosphere for this playlist comes from Tom Demers, “Many paid search practitioners (myself included) take a very data-driven approach to managing PPC accounts”. Data is a powerful tool when used properly and seemed to be a theme for this playlist.

Amplified Gold Posts

• “Our ability to participate and profit from ad auctions is dependent on a deep understanding and manipulation of a wide variety of signals.” Alex Cohen wrote a great post on using paid search signals. The post has 43 signals, which led to quite a long post. How about 10 signals next time?
• A great post on understanding search user intent . Using query intent is an effective way to guide bidding, ads, and landing page strategy.
• Testing landing pages to optimize performance in campaigns is not always ideal for a variety of reasons. If you or your client are looking for ways to optimize conversion rate without testing, here’s 5 easy ways to improve website conversion.
• In the years I’ve been doing PPC it seems there is always a spirited debate regarding to bid or not to bid on branded keywords. I for one say bid. Although I don’t agree with everything in 7 reasons why nixing brand keywords is not the fix , it has some good points. George gets into part’s of the post I don’t agree with over here.

Amplified Platinum Posts

• Rimm Kaufman’s blog is usually a good one. Using data to manage PPC is always a theme over there and I concur. I use averages constantly but it’s important to not get carried away. Read Average’s Lie for great info on when to NOT use averages. As a bonus the post uses the word “wagawaga” multiple times.
• Ad testing should be a regular part of any paid search campaign. So, you’ve tested and the test ad has a higher click through rate than the current ad but a lower conversion rate. What now? Hello Profit per Impression , a great way to select the best ad when ad testing.

Amplified Diamond Posts

• Doing everything necessary to launch a new PPC campaign or client account can be time consuming and filled with anticipation. What do you do when you launch and the results are not filled with unicorns and rainbows? Check out a couple tactics to improve new PPC campaigns.
• Everybody has an opinion on landing pages. Sometimes when consulting with clients on landing pages I feel like this . With all the elements and different opinions it can sometimes be overwhelming on what elements to test on a landing page. Keep it simple and start with 3 low hanging opps to increase conversion rates .

Chime in the comments with any posts that you thought were required PPC reading in the last few weeks.

Here’s Yahoo’s announcement: Yahoo! Transitions organic Search Back-End to Microsoft Platform

And some more in-depth coverage from Search Engine Land: Yahoo’s Results Now Come From Bing

—————–

8/17

Today Bing announced that they will be powering Yahoo’s organic search results for the English language in the United States and Canadian markets as early as this week.

This move has been anticipated since Yahoo & Bing announced their intent to merge their search capabilities in  mid-2009.  The transition is set to take place as early as this week and means that Bing will now own approximately 32% of the search market share (5.2 billion monthly searches), a large leap forward from the 13% they previously owned. Google still owns the majority of the search share with roughly 65% however search marketers/site owners can no longer ignore the impact a top listing in Bing could have on their bottom line.

This transition does not come as a surprise as the 2 companies announced their intent to merge in mid-2009. There is still no firm date for the transition of the PPC results – the initial intent was to make the transition prior to the 2010 holiday season but the alliance may postpone the transition until early 2011. To learn more about the merge visit the Yahoo & Microsoft Search Alliance site.

Related Post:

• Yahoo & Bing Team Up to Take on Google

Previously, this feature was an “opt-in” feature where you had to manually go and set this feature up. Now, Google is turning it into an “opt-out” feature.

Why this is something to take note of?

1. It influences your perception of search engine rankings. When this feature is on, actual search engine rankings become distorted due to your search history and previous search behavior.
2. This being “opt-out” means, more than likely, more individuals will be using this feature (whether knowing about it or not). With Google serving personalized results, it’ll be even more important for your website to show up from the onset so that a user’s personalized results doesn’t shut your website out of their results just because they’ve clicked on your competitor’s websites.

It’ll be important to know whether this feature is turned on in your Google account settings so you can accurately depict unbiased search engine result pages.