Loading ...« Weekly Wrapup - 01/11/2008 | Home | SEM Greenhorns - Google’s Online Marketing Challenge »
Related != Relevant - Really?
By Cameron Ferroni | January 15, 2008
I almost called this people are smarter than computers part 4, but figured you guys were getting tired of that theme. Plus it forced me to be more creative. In any case, today I’m thinking about relevance. Search engines are all about relevance. For years search engines have been trying their best to figure out exactly what the user is searching for, and then provide results that will hopefully give the user the answer they wanted. It’s no surprise then that online advertising is pretty heavily weighted towards relevance as well. You figure that one company controls most of the $$ around online advertising, and their background is in search, and voila - advertising + relevance = big $$. But it got me thinking - is that really the best way to maximize $$?
I’ve been toying around with the idea of “second order relevance” - in other words, what if rather than only showing users ads that were based solely on the keyword that the user was searching for, what if we also showed ads that were related to that concept. Would it be possible to make more money, and maybe, just maybe, provide a more useful experience to the user?
First the concept - let’s say I’m searching for a restaurant in Seattle on a leading search engine. On any given site I’m likely to see anywhere from 15-30 links for information on Seattle restaurants (a mix of paid ads + organic search results). As a user, how many links do I really need - and more importantly, how many am I going to pay attention to? Now, let’s say for grins that rather than showing 15 listings for restaurants, and 10 ads for restaurants, we showed 15 listings, plus 4 ads for restaurants, 4 ads for other local events/movies and 1 ad for a local limo company, and an ad for a local florist. Do we think this would be better, or worse?
I’ll start with the consumer - I could make a strong case that this is a better experience - I already know I want a restaurant, and there are plenty of authoritative results for me to choose from, so I don’t think we are taking anything away there. But now, I get some additional benefit as well - maybe it’s a romantic evening out, and buying flowers as well would be a great win. Or maybe I’m doing dinner, but now that I realize there is a cool event in town, I’m going to do both.
From the advertisers perspective, you get some of the same benefit - rather than being the 8th restaurant on a page with 30 restaurants, I can be the 1st florist. I have the opportunity to get my message out to a broader audience, and at a likely lower CPC than in other scenarios.
From the search engine perspective, obviously you would have to test the water, but it would seem that if it’s better for advertisers and for consumers, it would be better for you.
What’s striking to me isn’t so much that this isn’t being done yet, but that in fact the big players actually make this next to impossible to do. If the keywords that you buy aren’t directly related to the product you are selling, you can get your prices jacked, your ads dropped down the page or even blocked etc.
What’s also striking to me is that this is exactly how traditional advertising works, and has worked for year, and yet we are somehow moving away from it. Think about print or television. If I pick up a skiing magazine every single ad isn’t for skis. Sure most of them are, but I have ads for hotels, clothing, energy bars, airlines, resorts, SUVs, and, of course, skis. Or how about when I watch TV - if I go home tonight and flip on the Sonics game, I can be pretty certain that the ads that I see won’t only be about basketball and sporting goods.
“Well,” will say the skeptics, “that’s because on the internet the user has told us what they are looking for, so we are just trying to give it to them”. Really? That’s no different than the other outlets. If I’m watching a basketball game it’s pretty clear that I like basketball (the equivalent of typing basketball into a search engine - maybe even more directed) - but the advertising experts of the last 6 decades have figured out that just because I like basketball doesn’t mean I only want stuff related to basketball for the 2 captive hours. They know that other products - beer, cars, local news, television shows, new movies - are likely to appeal to me as well - and they make sure that among the other basketball broadcasts that they advertise and other sporting events that they make me aware of, that the rest of the advertisers get some play as well.
So what do folks think? Have we gotten too focused on the trees?
Topics: Local Advertisers, Local Search, PPClick, Advertising |
January 15th, 2008 at 3:49 pm
I think you’re onto something.
The question is, what are people doing when they reach a search results page? Unlike a passive viewer of a televised basketball game, they are, by definition, searching for something specific. An “in-kind” ad might just be exactly what the user is seeking, while a “second-order” ad is more likely a distraction relative to the user’s goal.
However, internet users are nothing if not good multi-taskers, so it’s completely plausible to imagine that someone finds the restaurant they want in the many good organic results that you point out they will find on a typical search results page, such that an ad for a limo might actually BE just what the user was looking for NEXT…without even realizing it.
January 15th, 2008 at 5:56 pm
I, for one, am not buying. Maybe some people will want this, but to me it’s a lot like when the salesperson tries to “upsell” me–”Do you want a tie to go with that shirt that you just picked out?”
No, if I did, I would have picked one out as I walked past the rack to bring the shirt up to the counter to purchase it.
Your example doesn’t persuade me; it distracts the user from their intent. A user watching TV or reading a magazine can’t be said to have any specific intent.