There’s sometimes a start-up period for a new adWords campaign when ads aren’t shown for all keyword matches and so the traffic is lower than it “should” be.

I covered it in this post on what to expect from a brand new PPC campaign.

My assessment of the reasons for lower traffic has been that it’s because a new campaign has no history and therefore the keyword quality scores are predictive and probably pessimistic. Google doesn’t show all variations of broad matches for keywords (accounts?) with a low quality score hence the waiting period for the full quota of traffic due. When actual data comes into the equation a high performing campaign can have superior quality scores and traffic will start to increase when ads are shown on broader variations of keywords.

There’s another interesting theory about the adWords “sandbox” on the RevenueWire PPC super affiliate blog. This is that Google puts a “go slow” on new campaigns until they have established that the advertiser can pay the bill.

Does landing page quality score bridge the gap between PPC and SEO?

Landing pages are now assessed for relevance so keyword optimisation is important. Pages that have little or no content are penalised and currently this manifests itself by a jump in the minimum bid required to show an ad for that keyword.

So there is now a crossover between SEO and PPC with on page optimisation important for both organic rankings and paid search. It has always made sense though to make the landing page relevant to the keyword and the ad. It’s important from a usability and conversion perspective whether or not it plays a part in the quality score.

The outstanding questions I have relate to whether other SEO factors will also play a part in contributing to the quality score. What about links, anchor text , age of domain, quality of website etc? I’ve no reason to think that these factors are currently taken into account but I don’t actually know for sure or whether there are any plans to include other factors besides on page optimisation.

PPC has always been effective for new websites that don’t have any history so don’t have many links and other trust indicators. It’s also good for one-off promotions and offers that aren’t meant to be a permanent part of other product or service offerings. In these cases you don’t want the page indexed by the search engines at all. PPC also allows testing of short copy versus long copy, navigation and links out versus a stand alone sales letter etc. Having to take into consideration what will be good for a quality score will affect decisions on what works best.

SEO is competitive and web pages can’t be equally relevant as they are listed in an order. Subsequently the algorithm for determining relevance is not trivial. In contrast the “show quality score” column in the adWords report has three values “great”, “ok” and “poor”. Presumably the Google internal scoring is something different but it’s not known whether it’s along these lines or something much more complex in line with the SEO ranking algorithm.

Currently the landing page isn’t part of the quality score that determines ad rank but only the quality score that determines the minimum bid required. Here is the explanation of quality score and landing pages from Google adWords help.

Google seems to be targeting particular types of advertisers and the majority of campaigns aren’t affected. I had a look at the new quality score column when it was first introduced but now have it hidden because I can’t see that it offers any benefit at all. If anyone can point me to some benefits I’m happy to hear them but so long as my ads are running it doesn’t appear to make any actual difference regardless of the score. If it affected ad rank I’d pay a lot more attention obviously.

More information on quality score for both the current situation and likely changes for the future would be nice. However, uncertainty and change are part of the nature of search engine marketing.

So are PPC and SEO coming closer together? Possibly but not to any great extent just yet. I’d welcome any challenges on this though.

E-consultancy have an excellent 12 page report on paid search including the latest statistics (from the UK) and the market trends in paid search. It’s well worth downloading and is free after registering. Thanks to Danny Sullivan for including it in his “day in search” round up post.

Here’s a summary of the market trends that were identified. This was the output from a roundtable which had 5 UK organisations in attendance.

Market Trends

1. Ongoing concern about click fraud
2. Increased sophistication of bid management tools
3. Paid search becoming more like SEO due to the importance of landing pages
4. Increased use of paid search for brand building
5. Marketers more savvy about ROI and starting to exploit longer keyword phrases that are cheaper than generic terms
6. Search is being built into overall marketing plans and there is more recognition of the need for paid search to complement SEO and other marketing channels (online and offline)

Robert Scoble blogs about Google showing fewer ads per page. It’s speculation based on a conversation with a Google employee. He gives a couple of possible examples and mentions Google research that people trust advertising more when there are fewer ads. Li Evans covers the story saying that it’s one that every online marketer should pay attention to as it will affect their budgets.

It’s certainly true that Google doesn’t display all ads. The blurb on the adwords home page says:

“When people search on Google using one of your keywords, your ad may appear next to the search results”. (The bold is mine)

The most common reasons why an ad doesn’t show are:

1. The daily budget for the campaign isn’t high enough to cover all possible impressions
2. The quality score is low resulting in the keyword being inactive because the bid is less than the minimum bid required
3. The quality score is low so the ads aren’t showing for broad matches. From Google’s help file: “A keyword’s eligibility for broad match variations is evaluated continuously based on its performance, so if the keyword performs poorly, it will be eligible for fewer variations. Conversely, if it performs well, it will be eligible for more.”

So, in addition to these scenarios, is Google also moving towards limiting the number of ads displayed in total?

In the example Robert Scoble gives, for the search “San Francisco Sushi”, he points to the fact that there are only 5 ads in total and that the Google employee claims there used to be more. Well it may be the case that advertisers have dropped out due to the quality score changes racking up their minimum bids to a level that makes it unprofitable for them to participate. Li Evans appears to be drawing the conclusion that Google is capping the total number of ads with her question “what happens to all those other advertisers that used to be in positions 4 on out to 100 (or however high they go)? “

I can’t see any logical explanation for reducing the total number of ads however maybe there’ll be a reduction in the number of ads per page. Currently up to 11 ads can show on page one, 3 in the blue box and 8 down the right hand side. Maybe some of these ads will get pushed off the first page and on to subsequent pages.

I’m assuming this would result in an increase in keyword costs as (generally) preferred positions are on page one and also above the “fold”. A reduction in the space for paid ads on the first page would increase the value of those spots and drive up the costs for advertisers.

I haven’t seen any changes in campaigns I manage in Australia but it’s interesting speculation nevertheless.

The changes to the quality score went in overnight and with it a major bug! I’m glad I saw reports of the bug before I looked over all the campaigns I manage as it has affected most of them.
The adWords official blog doesn’t have an update yet but here’s an overview from Danny Sullivan at Search Engine Land.

Edit: All fixed - no major dramas in the end

Two major changes affecting the adWords quality score are imminent.

1. There will be more transparency regarding quality score with a new column in the adWords reporting interface. This is scheduled to be available as early as tomorrow.
2. There will be changes to the quality score algorithm in the next week or two. The full details aren’t known but Nick Fox, Senior Business Product Manager for Ad Quality reported that the “bulk of the impact will be on keywords not driving traffic.” So I take this to mean that the click through rate will play an even more significant role.

Information from Andy Beal.

Edit - I read it on Andy Beal’s blog first but the official adWords blog has the updated info. Looking through all the blogs I read it seems as though a lot of people are covering this today. Hardly surprising given the importance that quality score plays in adWords campaign success.

There are reasons why a brand new pay per click campaign will take time to start performing at an optimum level.

1. There is no click through history so Google assigns a predictive quality score rather than one based on historical performance. You will start with a low quality score which will mean a period of “buying in” (with a relatively high bid price) to get a good ad position. Expect that experienced advertisers in competition for your keywords are probably paying less if they have earned a higher quality score.
2. If you have keywords set up using the broad match option, you will probably get a lower volume of ad impressions at first than you will when the campaign has been running for some time. Your keywords will not consistently be broad matched until your quality score improves.
3. When a campaign is new there is no data to analyse and draw conclusions from to improve the campaign performance. Nothing has been tested yet so you don’t know the best ad position, the best time of day to run ads, the most effective ad copy and how well the landing page converts. Over time a systematic programme of testing each of these variables will give valuable information on what works the best. Experience with other campaigns gives a head start to knowing what to try first, but nothing beats the actual data from a campaign that has been running for some time.
4. Targeting is one of the key principles of PPC advertising and the narrower the targeting the better. If someone searches for “vintage heuer watch” then an ad title “vintage heuer watch” and targeted ad copy will get more clicks than a generic “ladies watches” ad. It’s essential to organise the campaign into targeted ad groups so that the ads can then in turn be properly targeted. This can be done effectively right from the beginning but when the campaign has been running for a while there will be some keywords that will be worth putting into a new ad group to further improve performance. You can’t always predict which keywords in advance.

What does this mean for new advertisers / new campaigns?

• Perseverance will pay off. Google rewards good advertisers by giving them higher ad positions for a lower cost. This creates opportunities for those who know what they are doing and who put the work required into an ongoing cycle of testing and targeting.
• Be careful not to jump to conclusions too early and start making changes that may be counter productive in the longer term. I expanded on this here.

There are only two ways to pay less for Google adWords clicks.

1. Lower the bid
2. Improve the ad rank

Lower The Bid

This gives immediate results. You pay less for each click in the short term. It can seem like the only viable option especially if the campaign isn’t giving a positive return.

It will also have the effect of lowering the ad position. The most likely effect of lowering the ad position is to lower the click through rate which over time will lower the quality score which will then cause an even lower ad position.

Improve the Ad Rank

The quickest and easiest way to improve the ad rank is to raise the bid. This is very counter-intuitive when the goal is to pay less for clicks. However, paying for a higher ad position will almost certainly improve the CTR which will increase the quality score which will result in an higher ad position. At this point you can lower the bid and still retain a spot in the top few positions.

Raising the bid isn’t the only way to improve the ad rank. Ad rank is max bid * quality score so improving the quality score will also achieve this goal. CTR is the predominant factor in quality score. Ad position makes a big difference but so too does effective, relevant ad copy.

Conclusions

• Beware short term fixes that can have a negative effect on longer term success
• Give a campaign time to improve. Click through history must be established before results of changes can be measured.
• Be aware of all the variables and test to get the optimum combination
• Keep in mind that although it’s good to pay less for clicks, its only one part of the profit equation.

Every keyword in an adWords campaign is given a quality score by Google. This score is extremely important because it determines:

1. How much you actually pay for each click
2. Whether your ads will run at all
3. Your ad rank (position)
4. Whether your ad will show for all broad matches of the keyword

The finer details of the quality score algorithm aren’t made public, however:

Quality Score is determined by CTR (click through rate), ad relevancy and landing page relevancy. At this point, the landing page relevancy only affects the minimum bid required for the ads to run at all. However a senior product manager at Google indicated last November that landing pages will at some point be taken into consideration when determining ad rank.

The quality score was introduced to provide a better experience for searchers by giving prominence to ads that are the most relevant to their search query.

The amount you pay for your adWords campaign doesn’t determine your position on the page. Good PPC advertising can mean that you pay less than a competitor and are also in a higher position if your quality score is higher.

Marketwatch.com wrote about a couple of businesses advertising in the US who claim that Google adWords campaigns are becoming prohibitively expensive. These companies are planning on reducing their online advertising budget this year.

One of the companies is ebags.com. They sell suitcases and handbags online and last year spent somewhere between $5 and $8 million dollars on advertising of which 75% went to Google. The “soaring cost of Google adWords” was attributed to increased competition driving up the price of keywords. Google displays paid ads in an order determined by how much each advertiser has bid for a keyword in conjunction with a quality score given to the advertiser by Google. Ebags.com say that for each keyword they now pay roughly 45% of the cost of the product. This is eating into their profit margins and becoming too costly.

So, the news doesn’t appear to be good and marketwatch.com, a stock market site, covered the story believing that if more advertisers follow suit then Google’s share price will suffer.

But is it true? Not in Australia it isn’t, in my experience.

• I’m often surprised at the lack of competition for targeted keywords in what are considered to be competitive online industries. I recently set up a campaign for an online fashion business and the cost per click averaged 16 cents in the first month. It’s nowhere near the 45% of the product cost claims featured in the article.
• The travel industry is one of the most competitive online and keywords are more expensive but still averaging well under $1 for the campaigns I run.
• The most expensive keywords I buy are those for my own search engine marketing business. These are search terms such as “online advertising” or “pay per click”. (edit: no longer true)
  
• There are some absolute bargains to be found all the time. I don’t want to give current examples but I’ll give one from a month ago. A major golf tournament (PGA) took place last month at the Hyatt Coolum. I was the only advertiser bidding on search terms such as PGA and PGA golf. There were 16,000 searches for the PGA tournament in one month and this phrase was extremely relevant for one of my clients who had tournament packages and accommodation on his website. The cost per click was 7 cents.

I suspect the marketwatch article is anecdotal rather than based on any meaningful stats. In any case if advertisers drop out of the market for particular keywords in particular regions then the price gets cheaper for those remaining until it’s cheap enough to lure back previous advertisers. And so it goes. We’re nowhere near this situation in Australia though.