Color us happily surprised. Despite rising costs, competition and click fraud, marketers say they are FAR happier with paid search advertising (PPC) results now than they were a year ago. Only 34% of 3,007 active search marketers that MarketingSherpa surveyed in July 2004 said their PPC ad results were "very good." 13 months later, 43% of 3,271 surveyed MarketingSherpa readers said PPC ad results were "very effective." That's a nine-point jump.

Search optimization (SEO) campaign satisfaction didn't fare nearly as well. Last year 31% of surveyed marketers said SEO results were very good. This year 33% say SEO is very effective. Yes, it's a gain, but only two points. (We suspect folks are getting a bit fed up with the never-ending Google dance.)

Even then, SEO came out far ahead of non-search tactics such as email marketing (25% "very effective" rate) and affiliate marketing (22% "very effective" rate).

So how is this affecting marketers' plans as they swing into the 2006 budget season? And, do marketers who outsource SEM to specialist agencies get better results? Read on.

How search marketing budgets are changing for 2006

Our 3,271 survey respondents told us that online overall was a huge portion of their budgets. It was 56% of total budget for B-to-C marketers and 40% of total budget for B-to-B marketers. While this is fairly normal for B-to-B marketers; the B-to-C data is skewed by our high readership of marketers in ecommerce/multichannel retail, travel/hospitality and financial services.

Here's how they plan to spend on SEM for the next 12 months:

SEO Big Increase; PPC Big Increase; B-to-B Marketers 23% 20%; B-to-C Marketers 29% 22%

SEO Small Increase; PPC Small Increase; B-to-B Marketers 36% 36%; B-to-C Marketers 35% 43%

In addition, roughly 30% of marketers in both camps said they'd hold spending "steady." Only a 4%-5% said they'd decrease SEO spending. Paid search didn't fare as well. 12% of B-to-B and 8% of B-to-C marketers were planning to cut PPC budget.

Overall, according to MarketingSherpa and Merrill Lynch data, US total SEO spending rose 177% in the last 12 months as it is. We expect that rise to continue, although perhaps not as steeply. Only 21% of surveyed marketers say they are not doing any SEO at all. Last year 32% weren't doing any SEO.

In comparison, US paid search ad spending rose 67%. (But then, paid search spend is already measured in billions while SEO spend probably won't crack one billion until early 2007.)

Quick stats on clicks, costs, conversions ... and fraud

Here are overall survey results for you. These stats are broken down in useful detail by marketer-type and search engine in our Benchmark Guide. See link at end for info.

2004 2005 PPC Clickthrough Rate 3.0% 2.6%; Google Cost per Click $1.29 $1.61; Conversion PPC 6.6% 3.6% Conversion SEO 6.1% 4.2%

We didn't ask about click fraud last year, but this year only 7% of marketers said it was a "big problem already."

For us, the most interesting quick stat of all was the overall increase in keyword volume for marketers' keywords PPC campaigns. Here's an average number of keywords surveyed marketers said they were conducting PPC campaigns under per month:

Sept 2004 keywords 9,100; March 2005 keywords 14,700; Sept 2005 keywords 17,314

That's an overall 90% increase in campaigns over the past 12 months. This means marketers are spreading their budgets far more broadly than they had in the past. No wonder there's such a surge in search campaign management and landing page content management tools. Handling search has turned into an almost inhuman challenge without them.

BTW: What's the best way to discover new search terms? 48% of agencies said segmenting website log files looking for highly converting keywords produces "great ROI." So, we predict a sunny future for site-side programs that enable this.


Tests that work tests that don't work, and tests nobody's bothering with (yet)

Aside from trying out a bazillion more keywords, as noted above, what else are marketers testing?

Creative rules! Turns out search success is all about old-fashioned copy and graphics testing. Here are the top three testing tactics in-house search marketers said work well: #1. Landing page copy 89% #2. Text ad headlines 87% #3. Text ad body copy 82%

What's not worth testing? 14% of marketers said local search ads "work but are not worth the effort."

Plus, 26% of in-house marketers say daypart and weekpart testing (where you only allow your campaign to appear on "best" hours or days of the week) isn't worth the work. This is initially surprising because we've heard a lot of data supporting the fact that clicks convert far differently at different times of the day and week.

On the other hand, perhaps it marks professional maturity, as marketers are less likely to rely solely on immediate conversions for success. We all now know due to latency, search conversions can take place days or weeks after the initial click.

A startlingly huge number of marketers are *not* testing second tier PPC search campaigns -- basically anything outside of Google's AdWords and Yahoo/Overture. Depending on the specific engine (we asked about 12 second tier searches and five shopping engines), anywhere from 50%-83% of marketers had never run even a teeny test on them.

Those who had tested second tier tended to rate them "somewhat effective." Given how much cheaper these clicks can be, we suspect there may be a surge into second tier search as long as marketers have access to easy campaign management tools that enable them to handle multiple campaigns over multiple engines without a nervous breakdown.

How do search agencies and consultancies differ from in-house search marketers?

You guessed it, agencies overall prefer to test a variety of campaign factors more than in-house marketers do. On one hand, they have more time to focus on search tests, on the other hand they also need to find ways to justify their fees.

For SEO campaigns, agencies claimed to be performing vastly better than marketers relying on in-house optimization. We might not trust this data if it were not for the fact that hundreds of agencies and thousands of house marketers participated in this survey, making results fairly statistically significant.

Plus, everyone started at roughly the same level. Agencies said their clients' site traffic prior to SEO was about 74,000 monthly visitors. In-house marketers said their pre-SEO traffic was about 71,000 visitors per month.

However, overall site traffic lift six months after optimization was:

Agency optimized 110% increase; In-house optimized 38% increase

On the PPC side, agencies also reported higher results in terms of campaign effectiveness. One reason may be the fact that agencies were far more likely to aggressively measure results and tweak campaigns based on data. For example, only 35% of marketers were tracking delayed online conversions while 50% of agencies were.

One last fascinating fact -- we saw real differences in search test aggressiveness between newer and more experienced agencies. So many newcomers have flooded the market that we wondered if it's worth paying for experience. Answer = probably.

Note: Want more data on search, including special reports on affiliate SEM, B-to-B search and all-new eyetracking "heatmap" studies of top engines? See link immediately below.

Thanks to everyone who participated in this year's survey!

Useful links related to this article

Get your own copy of MarketingSherpa's entire 2005-2006 Search Benchmark Guide http://www.sherpastore.com/store/page.cfm/2166 or call (877) 895-1717

Results data from our 2004 annual survey on search marketing: http://www.marketingsherpa.com/sample.cfm?contentID=2780