Have you ever been bewildered by the hordes of different search advertising networks exhibiting at interactive advertising tradeshows?

More than 100 search advertising networks all claim to have "millions of searches per month" for you to advertise on. How can this be, you wonder, when the big four -- Google, Yahoo, MSN and Ask Jeeves have sewn up more than 80% of all searches conducted on the Internet?

And, more to the point, are any of the so-called networks worth advertising on?

We interviewed Andrew Beckman, President SearchAdNetwork who's tested more than 40 search engine networks for dozens of campaigns over the years. "Tier B would be Miva, LookSmart, GoClick and 7Search, among others," he explains. (See link below to longer list of Tier B networks.)

Where do these networks get their much-bragged-about millions of searches?

"Mostly parked domains, URL triggers and 404 (error) pages," explains Beckman. But some is straightforward contextual advertising -- text ads placed on content sites related to the topic of the ad. That's not exactly search, but Google and Yahoo both have similar offerings, and they're best known as search engines so, ipso facto, contextual may be called "search" even when it's not.

Doesn't matter, says Beckman. Just watch your results like a hawk and make judgments based on that.

Key: Clicks may be cheaper, but they'll also come in lower volume and convert at a lower percentage. If you need the click volume and you've already pushed the boundaries of the major search engines as far as they can go, Tier B is your next logical step.

More Tier B search marketing tips

There's an ugly underbelly to Tier B. "Some low-end search networks are gold sponsors at all these shows, but I've run tests and I would never do business with them again," says Beckman.

"At the end of the day I think the only way these guys stay in business is that most of their clients just don't have proper tracking in place. There are some deals. I just scratch my head."

Tip: Beckman says he's had great results for particular campaigns with the following Tier B engines -- SearchFeed, Information.com, Roar and Miva, among others. But no Tier B works for all of his clients' offers. Some only perform well for financial services, and others stink in that arena. It all depends on where their traffic is coming from.

Tip: If you get serious about Tier B, you'll need bid management tools that make it easy to handle accounts across multiple engines. "It takes you the same amount of time to set up a campaign on GoClick as it would on Yahoo."

Tip: "I'll run tests for deals I have a long history with, and I know what my conversions should look like on Tier B networks. If leads are much higher than other networks, I know we've got a problem and shut it off immediately. Then I'll look at the traffic and see where it's coming from."

Tip: Run an initial test for $100 worth of clicks and ask the network to kick in an extra $100 worth as a good faith gesture. If the test works, then roll out carefully. Never, ever spend more than $750 up front with a network you have no prior test results for.

Tip: Routinely monitor referring URLs for incoming clicks. You don't know where your ad is appearing in the network, and sometimes dubious sites that might hurt your brand (or just might be responsible for fraudulent clicks) creep in. "I'll poke around at URLs," says Beckman.

Shopping engines

Together, the major shopping engines' (Shopzilla, Shopping.com, etc.) total traffic accounts for roughly 8.5% of total searches. That rises fairly steadily by a percentage point or so every year as more consumers discover the engines during the holiday season and then get addicted to using them for shopping-specific search.

Contrary to popular marketing belief, consumers using shopping search aren't limited to bargain hunters. Only 20%-25% are motivated by price as their main conversion factor. Most others decide based on reviews, brand name, shipping speed, clarity of landing page, etc.

Consumers using shopping engines tend to be further down the sales cycle -- more likely to be ready to buy now. Which explains why clicks convert at a slightly higher rate than major search engine clicks. However, clicks are also more expensive.

You have to work harder to shopping search to make it pay off. Three tips:

Tip: Update your listings every couple of weeks, if not more often (even if little's changed). Shopping engines often show results based on the assumption that fresher listings are better, more accurate, listings.

Tip: "You gotta get reviews, positive consumer reviews," notes Beckman. Reviews not only help with clicks, they also influence your shopping engine ranking.

Tip: Deeplinking is critical for conversions. Shoppers who've already used an engine to drill down to a particular product or highly specific category will be highly disappointed if your click link leads them to a more generic page -- perhaps the landing page for a broader category. (For example, a searcher for "women's red sweaters" landing on your generic "women's sweaters" page.)

Shopping engine users are less likely than general search engine users to stick around and try to find content on your site related to their search. They're too far down the buying cycle to be satisfied. This is a problem if you're running grouped campaigns with lots of different keywords all funneling to one group landing page.

Local Search: why it's not taking off (yet)

If you listened to the hype, local search was supposed to be the big growth area for 2005. Yet, we've hardly seen any local campaigns running, and only a tiny fraction of the marketers we've surveyed are planning to run local in 2006.

According to Beckman part of the problem is that, although 20% of all searches are local in nature, once you niche down those searches to the local level, it's a tiny number of searches to try to get clicks on. The volume isn't there to get the actual local advertisers all excited (even if they had Web marketing departments, which most don't).

Retail chains, franchise networks and other national companies with hundreds or thousands of local outlets definitely could take advantage of local search. But most haven't yet often because centrally managing a local campaign for multiple locations is a lot of work.

Beckman predicts when easier campaign management tools hit the market, local could take off. "But it's not going to lucrative right away." Local is a long-term play.

Tip: 91% of consumers who've recently moved house use local search.

Pay Per Call: quick overview and five tips

Just like Tier B search networks, the Pay-Per-Call arena is rife with resellers and affiliates. Which means there are actually relatively few providers, but a confusing number of sales reps from different companies you never heard of who are all trying to land your business. It also means there's room for fee negotiation.

Pricing per call starts at $2 minimum on services such as Ingenio, Miva and Verizon. (Compare that to a .05 cent minimum per click cost on Google.)

In addition, you'll probably want to pay for tracking. Pricing for tracking is generally by the minute (aka, the telephony model) and it's all over the map from .08 -.09 cents per minute to .35 cents per minute.

Tip: Per minute pricing usually means rounding up in the carrier's favor, so if your call lasts for 61 seconds, you're charged for two minutes. It's one of the secrets to profitability in telephony.

You can negotiate lower rates based on how many minutes you're willing to guarantee you'll buy per month. If you go over, no problem; the low per minute rate still holds. If you go under due to either low response rates or shorter calls than expected, you'll still have to pay the guaranteed monthly total.

So you're taking some risk in exchange for the lower rate. Networks start getting very interested in working with your budget when you hit at least 10,000 minutes per month.

You'll be assigned a range of phone numbers for your search campaigns, each trackable to a particular keyword. Yes, you can request local numbers for local ads.

Tip: Ask how "clean" the phone numbers are. You don't want to pay for incoming calls for the person who previously owned that phone number.

Tip: Don't use a vendor that relies entirely on VoIP (Voice over IP) instead of traditional phone lines to conduct calls. VoIP is (still) not quite ready for prime time, especially when a dropped or mangled call could mean missed business for you. A combination of VoIP plus regular telephony lines is best -- combining the strengths of each.

Tip: Prior to testing Pay-Per-Click you should have already have a well-trained (and measured) inbound call center for your offer so you know they'll be able to convert good leads as well as accurately judge lead quality overall.

Tip: Consider putting toll-free numbers in your Google AdWords ads for a rehearsal period prior to running a Pay-Per-Call campaign on other engines. Google is the only major engine that currently allows you to run phone numbers in your text-ad copy, and there's no additional charge for this at present.

Tip: Pay-Per-Call response volume can be very low. Instead of considering it a main campaign, you might use the tactic for market research instead.

Got any search terms on regular PPC campaigns that you feel should be converting higher than they are? Pop up a Pay-Per-Call campaign and take some inbound calls to discover which potential objections are stopping conversions.

Useful links related to this article

MarketingSherpa's Search Marketing Benchmark Guide 2006 http://SEMGuideB2C.MarketingSherpa.com

List of 100+ Tier B networks at Pay Per Click Analyst: http://www.payperclickanalyst.com/others.htm

SearchAdNetwork, a division of Location 3 Media http://www.searchadnetwork.com