|5 Pay Search Engines Tested, Section 1 (Research)|
|Monday, 05 March 2001|
Topic: 5 Pay Search Engines Are Compared In An 8 Month Study
Word Count: 2130+
Focus: 8 Questions
Ezine Promotion - We Test 45 Lists, Directories, And Review Sites
For this experiment, we worked with Brian Alt, consultant and editor of the highly popular daily, "Ezine-Tips", to craft a comprehensive ezine promotion campaign. We registered with 16 directories, 17 Announcement Lists, and the 3 leading review sites. Our objective? To discover our true cost per subscriber.
How can you invest $35 in a pay search engine, advertise for 8 consecutive months, add no additional funds, somehow end up with $65 in your account, and STILL lose money?
How can you invest $.06 per hit in the phrase "ecommerce software", advertise for 8 consecutive months, maintain a #2 ranking, and gain just 8 paltry click-throughs?
Plutarch said, "Research is the act of going up alleys to see if they are blind." He must have been anticipating this project. While testing these search engines, this researcher has found himself in more than one "blind alley".
Here is a quick synopsis of the experiment.
We invested in 5 leading search engines, for a period of 8 months, to compare 3 elements: response rate, number of clicks, and ease of use.
We tested a combination of the words:
> ecommerce consultant
We used this description (or a shorter variation):
"CoBuild Your Store With An E-Commerce Expert - At CoBuildit.Com, live e-Commerce Experts are standing by to help you build a powerful, customized e-store. Get real help from real people."
We tested theses services:
We invested the following in labor:
> 1 hour - Listing preparation and research
We invested $400 in setup fees and final deposits, and we prepared to invest an additional $4600 depending upon preliminary results. Our total (actual) cost was estimated at: $1550.
Here are the estimated click-throughs:
> FindWhat - 21
So which engine cost us the most; was it GoTo, at $2.60 per click, or Win4Win, at $.01 per click?
You might be surprised...
We placed (the princely sum of) $35 at Win4Win, in an initial deposit. It remains, there, safely ensconced, today - As according to their own stats, we have NOT received a single click-through.
So Win4Win banked our original $35 plus they have generously added a credit of $30 bringing the grand total to $65. This money, however, might as well be maintained in a foreign currency. For it is inaccessible, unless we get clicks.
And so far, we have none.
GoTo won our largest cash infusion. Nearly 4 times what we spent on the other engines, but what was our true ROI (return on investment)?
Here is a simplified breakdown.
> For GoTo:
We adjusted our bids on at least 3 occasions, so these numbers are not precise, but they clearly infer an (all too common) problem: We ran out of money... FAST.
Getting bleary eyed? Don't "check out" yet.
Before these dull stats consign this report into information oblivion, please compare them with those from the next search engine. The difference could be worth the additional brain RAM.
> For Sprinks:
At Sprinks, we never got higher than 12th place in our rankings; we never spent more than $.38 per click, and yet we nearly DOUBLED our response.
What's more, we spent $200 at GoTo, ran out of money in 30 days, and got only 60 clicks. We spent $50 at Sprinks, advertised for 8 months, and got 100 clicks. And if that wasn't enough, somehow we ended up with a balance of $11.62.
Does this mean we now favor Sprinks? It does not. It just means that Sprinks was more cost effective for this search term.
In fact, we found the Sprinks manager interface to be quite ponderous. We estimate that it took us approximately 50% more time to complete a listing change in Sprinks than in GoTo.
So what about FindWhat?
Here's the mercifully quick answer: Smooth, fast interface - weak, tepid response: just 21 clicks.
After 8 months our account balance ($42.25) was higher than our starting balance. Apparently, like Win4Win, they graced us with a bonus. But if you add the labor cost to the setup cost... it still wasn't worth the investment.
At Kanoodle we bid $.06 for the term "e-commerce software and it's variation "ecommerce software". We won a ranking of #1 and #7, respectively. We maintained our ranking for 8 months, and at the end of this time our account boasted a balance of $49.52.
Why so high?
The answer is belied by the number of click-throughs. Our grand totalwas just 8.
Soooo... what will $400 and a lot of "hurry up and wait" at a pay search engine get you these days? It's hard to say, but here's what it got us:
186 responses at an average cost of $8.33@click. And even, if you "back out" the labor charge, the cost remains $2.33@click.
Somebody said "Numbers are like people, torture them long enough and they will tell you anything." We are committed to accurate answers, so we've strived to be exceedingly gentle with these numbers. Still, for all the tender loving care, the best they can proffer are some tentative, inferential points of interest.
Section 2 (Continue...)