Close
Join 237,000 weekly readers and receive practical marketing advice for FREE.
MarketingSherpa's Case Studies, New Research Data, How-tos, Interviews and Articles

Enter your email below to join thousands of marketers and get FREE weekly newsletters with practical Case Studies, research and training, as well as MarketingSherpa updates and promotions.

 

Please refer to our Privacy Policy and About Us page for contact details.

No thanks, take me to MarketingSherpa

First Name:
Last Name:
Email:
Text HTML
Jan 30, 2001
Case Study

Test Results for Paid Search Engine Listings: Google Beats GoTo.com, Lycos and Sprinks

SUMMARY: No summary available.
CHALLENGE

Although WebSurveyor is a dot-com start-up, the company has a reputation for carefully watching the bottom- line. So, when its Marcom Director Meg Walker needed fresh sales leads for research tools ranging in price from $199- $4,000, she didn't splash out in a big print ad campaign or a fancy trade show booth.

Walker had heard search engine keyword placements could be highly effective, and she noticed some of her competitors were already using them. However, as she says, "It's a really, really easy way to spend a lot of money if you aren't careful."

CAMPAIGN

Walker decided to test keyword buys on four different search engines. She limited her budget on each and watched results like a hawk, quickly changing or eliminating things that didn't work out.

1. Google -- There are two keyword buy options with Google. Marketers can use the AdWords automated system to buy the small boxed ads that appear to the right of search results; or by contacting a sales rep they can purchase key listings which appear directly above the search results. Walker decided on the latter for greater impact.

She says, "I gave the salesperson a list of words and terms descriptive of our products. He told me how many times those words had been searched for over a period of time. I decided to test ten keywords, including 'survey' and 'survey software.'" She tested Google for three months. Pricing was on a CPM basis, averaging $43 per thousand impressions.

Walker had 24/7 access to her results, including impressions served and clickthroughs, via Google's Web site for advertisers.

2. GoTo.com -- Although GoTo.com's search engine listings are brief, text-only, ads that run directly above search results just like Google's, they are priced per click via an online auction.

Walker explains, "You set up a credit card account at GoTo.com and then you can look through search terms to see who else has bid on them. The price isn't set, you bid against other people for how high up in the results you want your ad to appear. You need to watch out because people are playing with their bid amounts constantly. One day you'll be number one with a bid of a dollar with number two below you paying two cents, and the next day you'll be at position five! Just bear in mind, you don't have to bid for the number one spot, especially if the other people with the same keywords are not competitors. Your listing can be in the second or third spots and still work."

Walker tested nine different search terms, many of them the same as terms she tested on Google, over three months.

She had one reservation, "GoTo.com doesn't tell you exposures so you never know your click through rate. They send you an email with a report occasionally, but by the time you get it, it's two weeks out of date. I like to make changes really fast if something isn't working, so this was frustrating."

3. Lycos -- Lycos' keywords sales system is similar to Google's. Walker worked with a sales rep to select keywords, and purchased them over three months on a CPM basis. She tested a total of four terms, including 'market research', 'survey' and 'Web survey'.

4. Sprinks -- The pay-per-click program on the About.com sites is very similar to GoTo.com's. You have a choice of a short auctioned listing or a longer, more expensive, listing. Walker ran a quick test for just $100 for a few keywords via the auctioned section.

For all four search engines, Walker's listings linked to a registration page on WebSurveyor's site offering a free trial. Unusually for a B-to-B site, all names were put through a double opt-in process. People who registered received an automated email message asking them if they really wanted to be registered. Only those who responded to that email were placed on the final leads list.

Walker explains why she insisted on a double-opt-in, although it lowered the number of names on her list, "Sometimes people will sign up friends or colleagues. Then those people get annoyed when our member services people follow-up with them. They think they are being spammed." About 75% of registrants replied in the affirmative to stay on the list.

Next all leads went through WebSurveyor's sales process which involves follow-up phone calls from member services reps and also several follow-up automated emails.



RESULTS

Google by far outperformed the other three search engines in every way, with GoTo.com coming in second.

For example, Walker's ad under the term 'survey software' at Google had a 3.4% click through rate. Of these 6.22% registered for the free trial; and, the accounts that closed from these leads were at a 3,900% higher dollar-per-sale amount than any other search engine's sales.

Walker wasn't able to track click through rates for GoTo.com due to their reporting limitations, but she does know that an average of 5.97% of click throughs then registered for the free trial, resulting in just eleven percent of the sales units she got from the campaign as a whole.

Lycos' results were partially hindered by the fact that far fewer users searched for the terms Walker selected. For example, she got 3,000 impressions at Lycos and 73,000 impressions at Google for the term 'market research' over the same time period. Lycos' click through rate was a low .9% and only 3.5% of the people who clicked through ended up registering for the free trial. None of these resulted in an actual sale.

As for Sprinks, Walker says, "We got no results from them, although if someone is marketing to a particular vertical and tries the longer description, it might be worth testing."

NOTE: These results were for a specifically business-to-business campaign. Business-to-consumer campaigns might get very different results. As always, we recommend you test everything before making final decisions.

Post a Comment

Note: Comments are lightly moderated. We post all comments without editing as long as they
(a) relate to the topic at hand,
(b) do not contain offensive content, and
(c) are not overt sales pitches for your company's own products/services.










To help us prevent spam, please type the numbers
(including dashes) you see in the image below.*

Invalid entry - please re-enter




*Please Note: Your comment will not appear immediately --
article comments are approved by a moderator.