Ask: Yet Another Play for a First Impression

November 6, 2008

Author’s Note: This post will not render correctly in Internet Explorer 7.0. I am looking for a fix.

My newsreader pointed me to “Ask.com Speeds Up Its Searches” on the IOL Technology Web site. The author was Rachel Metz. The article’s main point is that Barry Diller’s Ask.com has been tweaked to make it display results more quickly. The most interesting statement in the article was:

Ask.com, owned by InterActiveCorp, encountered the repeat-visitor problem after launching a version of its search engine, Ask 3D, in June last year. With Ask 3D, the site moved away from showing search results as a long list and sorted them into three vertical panels, one of which included photos and other multimedia content related to users’ queries. Ask 3D was well-received, chief executive officer Jim Safka said, but it was too slow at downloading search results. “A lot of people tried the site, but wouldn’t come back.”

A year, maybe 18 months ago, I had dinner with two people from a third tier consulting firm. One of the consultant’s comments lodged in my mind. The keen thinker said, “I think Ask.com is doing a very good job. I use the service because I find it more useful than Google.”

Quite useful for 11 year olds. Not so useful to me.

After this comment, I make Ask.com a regular stop on my swing through Web search engines. I come back to ask, so I am a repeat visitor. The problem is that I don’t use Ask.com, and it has zero to do with the interface. Speedier performance, related results, and skins don’t mean much to me. I learned in September 2008 that some middle school students find Ask.com a useful resource.

I don’t even know a middle school kid, so I can’t begin to think about an online search from that point of view. I made a couple of inquiries and learned that a middle school assignment is a personal narrative or a biography of an important person such as George Washington. I ran the query “George Washington” on Ask.com, Live.com, and Yahoo.com. I skipped Google because everyone I know uses Google for most searches. I wanted to see what the also-rans were doing to win me over.

I am running these queries in Denmark so the first hit is in Danish. The other results showed pictures and links to sites such as

I am reproducing the urls to illustrate that Ask.com has some interesting code behind characteristics. In this free Web log post, I am not going to explain what the “clean” interface masks. In fact, I don’t think middle school kids care. Advertisers may. I know for a fact that the reference to performance issues and optimization strikes deep at the way in which Ask.com is engineered. I don’t recall seeing any reference to this code behind issue in Mr. Metz’s write up which is understandable. I noted that the Ask.com executive quoted above jumps over this point as well.

The results from the Microsoft Live.com query were:

Microsoft makes extensive use of caching to speed results delivery to the user. Note the difference in the urls. Microsoft is lean and mean. The relevance of these results is good, but if I were a middle school student I think I might prefer the pictures and direct link to a biographical snapshot. Microsoft expects me to read.

The results from Yahoo were:

Yahoo presents image eye candy at the top of the results list. Yahoo also includes “see also” references. The relevance of the top three hits is good in my opinion. If I were an attention challenged middle school student, I could get what I needed from the Yahoo hits. Note that there is a bit of code behind magic going on, but the Yahoo code behind is lighter weight that Ask.com’s beefy, over burdened urls.

Let’s step back. Ask.com is struggling with performance. Ask.com is concerned about users who come once and unlike me, never return. Ask.com is saying that it thinks speed will solve the usage problem.

I disagree. Here’s my logic. Ask.com has had many chances since the late 1990s to make a first impression. The untenable “natural language processing” that haunted the firm in its early days is gone. Then AskJeeves.com acquired a company and integrated Direct Hit. That made no impact on my perception of the company. Then the company shifted to the Teoma technology developed at Rutgers University. I recall reading a white paper about Ask.com’s architecture and its scaling and performance properties. I looked at the diagrams and deleted the file. I thought the approach was derivative and a bit old fashioned compared to what open source information I had in my files about Google. Then the butler was killed. Hint: in the marketing department by the owner. Then we had 3D whatever. Now we have a Google like vanilla page.

Bottomline: I don’t think Ask.com knows what to do to build traffic. The company in the last 10 years has floated like a rudderless lake craft. Google on the other hand has been consistent in its focus on relevance and delivering search without search.

How can Ask.com get its rudder fixed and someone in the boat who knows where the dock is? I think there are three steps the company can take.

First, go back to the middle school example. Do that well. Then move to high school users. Do that well. Then move to junior college users. Do that well. Stop. Instead of trying to be Google, Microsoft, and Yahoo, focus. I don’t know what happened to the vertical focus on female users, but the student search sector makes more sense to me.

Second, look around for some AltaVista.com DNA and graft it into Ask.com. Not only will this approach open new service opportunities but Ask.com may discover that it can license an engine and concentrate on the content and services for its demographic. One happy outcome will be an approach to scaling and performance that is affordable even by an abstemious corporate baron.

Third, time is running out. Shift the outfit into high gear. If that means more management changes, make them.

Keep in mind that this is my opinion. Your mileage will vary. Keep in mind that most sites get one chance to make a first impression.

Stephen Arnold, November 6, 2008

Written by Stephen E. Arnold · Filed Under Business strategy, Feature, Online (general), Search, Technology

Comments

Comments are closed.

Search the site
Subscribe to Beyond Search
Feature archive
News archive

Stephen E. Arnold monitors search, content processing, text mining and related topics from his high-tech nerve center in rural Kentucky. He tries to winnow the goose feathers from the giblets. He works with colleagues worldwide to make this Web log useful to those who want to go "beyond search". Contact him at sa [at] arnoldit.com. His Web site with additional information about search is arnoldit.com.