|Google Direct Answers: overrated threat?|
Quite a few site owners have complained about Google's "Direct Answers" (a.k.a. "Knowledge Graph" boxes), which they think will deprive of them of traffic by answering searchers' questions on the SERPs.
In some cases, at least, that threat may be overrated.
Here's an example (and I'll use a real one, just to test the temporary change in this forum's TOS):
I did a search on "doughnuts" a few minutes ago, and the resulting SERP had a huge Direct Answer box with a photo of doughnuts, a dictionary definition of "doughnut," and a long list of "Nutrition Facts" for "doughnuts, with chocolate" from the USDA. For the searcher who might be looking for information on making or buying doughnuts, the Direct Answer box was essentially a waste of space. It seemed to exist mainly a technical showpiece, as in:
"We've figured out a way to mash up photos, dictionary definitions, and USDA information in food searches. Isn't that cool?"
Obviously, there are cases where Google Direct Answers may take away casual traffic (people wanting to know the capital of North Dakota or whatever), but for at least some searches, Direct Answers and the Knowledge Graph are little more than fluff.
I'm trying hard not to sound like I'm wearing a tin foil hat, but Google's m.o. is to roll things out in the least controversial way they can...and then turn the knobs later. You start the lobsters in cold water, and they don't notice you're boiling them.
Examples that progressively took more screen space over time (not related to "Direct Answers", but related to their historical approach):
- flights from nyc to miami
(nothing organic above the fold, at all)
- compare credit cards
(3 results, at best, above the fold, on a big monitor)
Direct Answer queries that seem to be siphoning "non fluff" traffic:
- nhl standings
Sure, not much commercial intent, but it's live, relevant, not fluff, and that traffic used to go somewhere.
- syndey weather
I bet I know who isn't happy about that big widget
- best actor oscar 2014
Perhaps fluff, but that traffic did used to go elsewhere
-AA flight 101 gate
AA misses this opportunity to interact with a customer
I can see this extending out to the point that it's a measurable cost. That last one, in fact, I know pulls value from the airline sites.
The point I was making is simply that Direct Answers aren't likely to be traffic siphons or threats to publishers for all queries. For the example I gave, there's no lobster to boil: Most people who search on "doughnuts" probably want something more than (or at least different from) a dictionary definition and nutritional information. They're more likely to be interested in recipes, where to buy doughnuts, pictures of cinnamon twist doughnuts like Grandma used to make, etc.
If anything is likely to divert traffic from the organic results, it's the carousel with names and addresses of local doughnut shops (which is probably just fine with the shop owners, who want prospective customers to find their brick-and-mortar stores).
|Most people who search on "doughnuts" probably want something more than (or at least different from) a dictionary definition and nutritional information. |
Yes. So what they are presenting for "doughnuts", at the moment, isn't impacting recipe sites, etc.
Unless, they decide at some future point that it will. I tried to present some niches where they've already made that decision by presenting widgets that go well beyond dictionary definitions. Some of the examples are very clearly taking traffic away from real sites.
At the end of the day, google.com is google's website and they can do whatever they want.
If they want to replace the search box with a funny cat video and never show a search result again, they can, and no one can stop them from doing it.
In the not too distant future, google's search results will look more like About.com / wikipedia than a traditional search engine. And you know what, people will still use them to find information, and website owners will still complain that google owes them something...
|google.com is google's website and they can do whatever they want. |
To some degree. When they started, for example, scraping Yelp! reviews and using them in Google Places, they eventually capitulated to legal pressure.
There's also the issue of the symbiotic relationship. Some of the sites they scrape earn most of their money from the organic traffic Google sends their way. If Google scrapes and presents the content in such a way that less people visit the actual source of the data, it's only natural to assume that some sites will shut down.
You can see that trend right now. Look at historical traffic for the top "dictionary" type sites. Also interesting: wikipedia page views are down in year-over-year comparisons for the first time....ever.
Google Direct Answers is starting to make its way into my area. In some cases, they use my sites for the answers, in other cases they don't. Because I deal in events, it's important to convey correct information, and a lot of the "Answers" aren't correct - in many cases showing, for example, an About.com answer about a 2013 event when the query is for a 2014 event (that is updated for 2014 on my site)
They're only showing very limited information, and usually I'm the top result (so why are they scraping other sites?) So for now, I have to make sure my page titles and meta descriptions make it clear I have the most accurate and comprehensive information. I'll lose some people who only want the date, but that's life*.
Oddly, my traffic is way up this year on almost all the sites, since the events started showing up in the box. Wish I had a way to tell if I was actually getting any traffic from them, but as far as I know, there's no way to do that.
* stuff like this, (Plus smaller Facebook reach) is why I think everyone needs to have their own email list
Every time something new populates the top of the search page it pulls traffic from organic results below that topic. It takes traffic.
It might take a tiny bit of traffic, it might directly "answer" what the searcher wants and they'll never search any further...but it takes traffic from websites.
The more traffic google organically diverts from websites the more they will need to rely upon paid results.
The knowledge box is ultimately an organic traffic killer, whether large or small.
Ask the NHL (insert TSN, ESPN, etc) for example if they care that me (and others) just enter search "nhl scores" and I don't need to go anywhere. I don't need to click to go to a website. I don't need to see their websites or their promotions. I see clean scores, past and present and even upcoming games. All that without dealing with a website. It's Google, not a website.
I could list many other examples, but this is how it works for me. The affected sites wouldn't see a ripple or would be so clueless as to any traffic fluctuations that they don't care. At this point at least.
I would just say for anyone to be ignorant of the trend, I send my regards to you.
MrSavage: So what's your proposed solution? That Google and other search engines refrain from giving direct answers when searchers want simple facts such as NHL scores, flight-arrival times, the current temperature, or the value of euro in USD?
If your site deals in simple answers, then maybe your business is going the way of DOS memory managers, third-party Winsocks, and $99 hard-drive defraggers. That's a tough break, but change happens. Google, Bing, and Yahoo! aren't going to turn back the clock.
I run a historical data-oriented site, and this does scare me. Although data is certainly factual in nature, the fact of the matter is that Google is profiting off the heavier lifting done by others.
My site has been cited by Wikipedia as a primary source for tens of thousands of pages. It is safe to say that without the information I have compiled - over the course of 20 years - Wikipedia would not have as much information as they do. But I'm now no longer #1 when people search for information that I originated on the internet (via offline research). I'm behind Wikipedia. And now Google may just make my page irrelevant? How does that help the ecosystem?
A really good example of the end game is Google images. I used to get a ton of traffic from people searching for logos of widgets. I get almost none now, because Google has taken those logos from my site and displays them in their image carousel. They're displaying full resolution copies of them. Google's design eliminates the need for the click-through - when I search for a logo, and my site is the only source of this logo, and I click on it, I get a Google page that shows my full-size logo - hotlinked!. If I click on the hotlinked logo, I'm directed to another Google redirect page with my page's link on it. I have to click yet again to get to my site. Almost nobody clicks three times.
Google couldn't deliver that content without me, but now I'm cut out of the picture. What choice do I have? I can block Google from that page (effectively depriving people of the content) or I can allow Google to show it to people in such a way that it makes people think that Google got it for them, not me.
That is pretty clearly the future for Google.
I should point out and ask for predictions for when searches involving "release date" will simply be answered on Google. Video games, movies, products, I'm sure those will be included. I'm simply asking what people here are predicting for when that day comes. Every time I'm doing those searches now, I'm expecting that answer box to appear. It's not yet, but perhaps other have different findings.
I'm not just pointing at Google here because this is all part of this trend of Siri type answers. If they all move together, there is no one target or one bad guy as it were. If everyone is doing it then there isn't anything wrong with it.
Summary wise, I'm just looking at eradicating release date related information from any of my projects. Seems futile at this point. I mean can anyone here argue against this information being provided directly? No need for my website page. I'm looking for a counter viewpoint disagreeing with my views on it. Any takers?
|I should point out and ask for predictions for when searches involving "release date" will simply be answered on Google. |
Isn't that likely to depend on markup data provided by the people doing the releasing?
Interestingly, Schema.org doesn't seem to have markup for "releaseDate," although there are some other properties that might work if the search engines interpreted them correctly in the context of a movie, game, etc. (such as "startDate" or "datePublished").