Plunge in Google Visitors Reveals Quality in Bing/Yahoo Results

[webcentric]

Well, a fortuitous (or "in-fortuitous" depending on how you look at it) shift in the winds at Google have helped to expose a very interesting pattern which I haven't really paid attention to until now. Because a high percentage of my traffic has historically come from Google SERPS, it's been easy to not get too analytical about the quality of traffic from Yahoo and Bing but here's what I'm finding out right now while Google is taking a little vacation where my site is concerned.

My pages per visit is almost 4 times higher when the mix of traffic is

60% Yahoo, Bing, other engines, direct traffic and links
40% Google

As opposed to

20% Yahoo, Bing, other engines, direct traffic and links
80% Google

I expect direct and link traffic to be more engaged generally but didn't really expect Bing and Yahoo to be sending far more engaged visitors than Google.

Couple of insights, Bing and Yahoo send visitors to more general pages (like a category page) where Google sends people to more targeted pages (items in a specific category). Perhaps it's the direct approach of Google (sending people to exactly what they want) that prevents further investigation of the site whereas Bing and Yahoo are sending people searching for a broader range of subject matter.

Anyone else notice this kind of distinction or have any further insights on this?

Bottom line is, I'm glad to have been able to get some traction in these other engines and to have developed a bit of a following as those sources are picking up the ball that Google keeps dropping and kicking around like a tin can.

[webcentric]

I'll add to this that the bounce rate is significantly lower under the above scenario which seems like a no-brainer but better said than assumed.

[Steven Davis]

I've seen the same. My RPM is 2.5 times higher with Bing/Yahoo visitors as opposed to Google visitors, and my bounce rate is reduced by 40%. I'm starting to believe Bing is starting to outpace Google in overall search quality. I think too many algorithmic shifts on the part of Google are starting to produce less focused results. Google is breaking a cardinal rule of life and business - If its not broken don't fix it.

[vandelayweb]

That seems to be in line with the report published by Shareaholic that showed higher engagement (time on the site & pages per visit) and lower bounce rates with Bing and Yahoo versus Google. There is an article on Search Engine Land for anyone who wants to explore this further.

I agree with you Steven. I spend more time on Google because my job dictates it, but a lot of the time when I am searching I'll go to Bing. Google seems to be putting a greater emphasis on the big brands which is generally fine, but doesn't always deliver the best results for what you are looking for.

[Itanium]

Well, you can view this in two ways. Maybe Google is sending people to the exact thing they were looking for. No need to explore other pages on the site. I'd argue, that this is better for the overall user experience.

[webcentric]

I agree with your premise Itanium. On the other hand, if the actual search is more general, then the user is also getting what they asked for by landing on a more general section of the site. Bing seems to be indexing my site from the ground up whereas Google has many more of the leaves on this tree indexed. I'm sure this accounts for the volume of visits from Google generally while diluting the the engagement results at the same time. Bing on the other hand seems to give my category levels more weight and when someone finds one, they seem more engaged because they were searching generally to start with. I might find that as Bing gets deeper into the site, engagement from it's visits will also decrease. It may just be the way different sections of the site are being targeted in the search. I'm gonna watch this more closely in the future though.

[Steven Davis]

@ Webcentric and Itanium

You both posit a strange notion namely; that high bounce rate is a sign of highly targeted/quality search results. I note those comments as strange since here and at most other forums for years Webmasters have declared that High Bounce Rate is a sign of a low quality search results. In fact, here in Webmaster World I've watched as many of you have advocated that a high bounce rate should be a negative search signal, but now High Bounce Rate is a sign of a highly target query that was well satisfied by the presented content.

A truly novel notion -- Should then High Bounce Rate be used as a positive search signal?

[aristotle]

All sites get some mis-matched traffic. Sometimes the search engine just mis-interprets the query. But most mis-matched traffic, in my opinion, is due to the search engine's failure to give enough weight to relevance.

So if one search engine sends better-matched traffic, it might be because of a greater emphasis on relevance.

[EditorialGuy]

Could demographics play a role? Bing and Yahoo are minor players outside the United States, while Google is international.

[webcentric]

@Steven Davis - Strange indeed. I've always thought of bounce rate more in terms of a page's ability to engage a visitor but as we all know, a bounce isn't so bad when the user leaves on an ad click for example. And yes, now that you mention it, what is said above does lead in the direction you put your finger on. Maybe a bounce, in some cases anyway, is a sign that you did your job perfectly and gave the user the absolute best experience possible. And maybe the SERP that delivered that bounce did exactly the best possible job it could in leading the visitor to exactly the information they were seeking. How to quantify and separate one positive reason for a bounce from a negative reason (e.g. this site sucks, get me out of here) is the real challenge. Of course this might be where an honest assessment of the quality of your site/information comes into play. In this age of secure search, it's harder to look at the actual query and guess about the user's intention and whether they got what they were looking for or not.

[tangor]

Or... the search was so targeted that the user still didn't get what they wanted and bounced out. I think time on page is necessary to make that determination. What we all really want (and can't easily get) is the user's actual happiness/unhappiness with the page/link. We can make some pretty fair guesses, but there's nothing provable.

[JD_Toims]

As far as conversions go, historically, I can't remember reading, hearing or finding anything in my stats to indicate anything other than both Bing and Yahoo! are *much* superior to Google in that aspect.

[tangor]

I'm wondering, do we know the demographics of who uses which SE? I mean good numbers for age groups, etc. It's been my experience that older folks, usually more interested in real research, not just answers, use Bing or Yahoo... the yahoo folks because they have been there "forever" and the Bing folks who aren't wide0eyed over G.

[webcentric]

Could demographics play a role? Bing and Yahoo are minor players outside the United States, while Google is international.

I block traffic originating outside North America (well some proxy stuff probably slips through but not much). This site is targeted at America and that's where it's traffic comes from except a trickle from Canada.

I think time on page is necessary to make that determination...

It only take a few seconds to write down a phone number or dial it. On the other hand, if the searcher is looking for a recipe, it could take several minutes to absorb. In both cases you could have given the user exactly what they were looking for and in both cases they could bounce off. This is where understanding the quality of your own content comes into play but it's still guesswork to some degree when you don't know what the original search term was. This conversation has me less certain of the meaning of bounce rate now than ever.

As far as conversions go, historically, I can't remember reading, hearing or finding anything in my stats to indicate anything other than Bing, Yahoo! and NerdyData are superior to Google

My gut is agreeing with you but my brain isn't exactly sure what to hinge that intuition on. Is it better targeting? Google's shotgun approach to indexing the web vs Bing and Yahoo's more careful, measured approach? A matter of what pages each engine elevates in any given query situation?

So if one search engine sends better-matched traffic, it might be because of a greater emphasis on relevance.

Or dumb luck. Either way, gauging the quality of traffic from certain historic indicators such as bounce rate (or even pages per visit) doesn't seem so cut and dry as it once did. Did someone visit 10 pages and end up totally frustrated that they felt close the whole time but never found what they were looking for? I may be happy they saw 10 pages and clueless to the fact that they'll never visit again.

Added: and tangor adds another interesting twist with SE user demographics.

[Itanium]

@Steven Davis

I think it depends what you classify as "bounce rate".

User who are closing the page or going back to the search results after 10-15 seconds are people who didn't find what they were looking for.

However, people who just looked at one page, without checking out other pages on the site, might have found exactly what they were looking for. So there was no need to check other pages.

One can only differentiate both groups with proper on page analysis (like a heatmap). Bounce rate as a stand alone metric is pretty meaningless in my eyes.

[Martin Ice Web]

A direct bounce could also be produced by the site itself. If i reach a site i donīt like from the first look i go back without taking a second look. Unfortunatelly google servs a lot of these sites i donīt like.

But have a look at googles panda. Before panda we had conversions rate from about 25%. After all the panda quality updates this conversion rate is down to 5%. If i take this as a measure then google produces unrelated results, leading to a high bounce.
Sometimes when google has a good old time for one hour conversion rises back to 25%.

bing visitors generally lead to high conversions and most of them have more then one item in basket, while google visitors are one item shoppers.

biggest problem with google is their "similar" keyword annoying serps. Searching for Keyword A, produces results for Keyword A, google fantasy Keyword B / C / D.

[netmeg]

As far as conversions go, historically, I can't remember reading, hearing or finding anything in my stats to indicate anything other than both Bing and Yahoo! are *much* superior to Google in that aspect.

As far as PPC, Bing's conversion rates have been far far higher than Google's for years now, across a wide variety of accounts both B2C and B2B. And not just for me. I only wish they had the traffic. Wouldn't surprise me a bit if Bing's conversions on organics were better too, but I couldn't explain why.

[Steven Davis]

I love the front page of Bing today a Panda munching on Bamboo, anyone else thing Microsoft is taking a jab at Google?

[tangor]

Hey! You gotta love a panda! (All the "folks with agendas" tell us we must! :)

[turbocharged]

A lot of scrapers/bots scan through Google's search results and hit the destination sites to harvest data. That's why, in my opinion, Bing/Yahoo converts better per 100 visitors as compared to Google.

[webcentric]

A lot of scrapers/bots scan through Google's search results and hit the destination sites to harvest data.

Which is where the blocking of server facilities comes in handy, oh and certain countries as well as known bots and bot behaviors. I like to filter as much of that noise out as possible.

[EditorialGuy]

I love the front page of Bing today a Panda munching on Bamboo, anyone else thing Microsoft is taking a jab at Google?

To paraphrase Freud, "Sometimes a panda is just a panda." :-)

[Brett_Tabke]

I think it depends what you classify as "bounce rate".

User who are closing the page or going back to the search results after 10-15 seconds are people who didn't find what they were looking for.

What if Bounce Rate were something else to Google?

- user clicks top organic link.
- user does not find info - clicks back button
- user looks closer at ads and user clicks an Adword
- ad satisfies search.

Google wins, and trains users to use AdWords.

[webcentric]

Or maybe lazy user lands on a page with exactly the information they're looking for but has no interest in reading through or even scanning a 2000 word article to find it. Too much work involved so hit's the back button and clicks on an ad on the SERP page.

Google wins, and trains users to use AdWords.

There's more than one way to skin a cat.

[EditorialGuy]

Google wins, and trains users to use AdWords.

Is that old Webmaster World macro still working? The logic behind it certainly isn't. Consider:

1) Many search queries yield SERPs with few or even zero ads.

2) The only site owners who can be "trained to use AdWords" are people who are actively selling goods and services. (The ROI just isn't there for non-commerce sites. It never was, except for a brief period when AdSense "click arbitrage" was a fad that Google slapped down.)

Getting back to bounce rate, it's worth noting that even "bounce back to search" is only one metric for gauging user satisfaction with a search result. There are others, too.

Example: The "Engagement" numbers in Google Analytics. Hardly anyone ever talks about these, but they probably should. What percentage of a site's page views come from visits of more than 30 minutes or of 20+ pages in a session? And how do those percentages compare with the percentages for similar sites? (Think of those numbers as a basic quality check: If the numbers are low, that's probably to be expected, but if nobody is engaging with the site while similar sites are attracting at least some interested users, that may raise questions about the site's quality.)

[webcentric]

The "Engagement" numbers in Google Analytics. Hardly anyone ever talks about these, but they probably should.

Don't use GA any more but what I do use does a pretty darn good job of letting me know how people are engaging.

Engagement metrics for a user landing on a blog category page is going to be different than that of a user who lands on an article or a user who lands on a directory listing or a user who lands on a product page. If I'm comparing product prices, I can tell in seconds if your product is less expensive than the last site I visited.

So, expectations kind of have to match the landing environment where further engagement is concerned. Did they add the item to their cart, check out another category or two, spend time reading the article before bouncing off? All are measuring engagement in one way or another. Each type of page has it's own level of engagement effectiveness and you have to think about each page type when setting goals and developing strategies for getting that next click. I guess my point is that site-wide engagement numbers such as page views per visit can be deceptive unless you break that information down by page type. Reasons for engagement are myriad and depend on the page being viewed as much as the goals of the user viewing it and the SE's ability to interpret those goals (if and when the visitor came from a search engine that is).

[JD_Toims]

Is it better targeting? Google's shotgun approach to indexing the web vs Bing and Yahoo's more careful, measured approach?

To the best of my knowledge and based on what I've seen: Yes to both questions.

[JD_Toims]

What if Bounce Rate were something else to Google?

- user clicks top organic link.
- user does not find info - clicks back button
- user looks closer at ads and user clicks an Adword
- ad satisfies search.

Google wins, and trains users to use AdWords.

They would never!

I mean that would mean if the top result didn't satisfy the query they would make more by *not* promoting the best result to the top of the SERP for a given query, and that's unconscionable -- Er, uh, maybe now that's just considered business these days.

Or maybe lazy user lands on a page with exactly the information they're looking for but has no interest in reading through or even scanning a 2000 word article to find it. Too much work involved so hit's the back button and clicks on an ad on the SERP page.

--

There's more than one way to skin a cat.

Yup, as Google, you can't contradict the "content is king" mantra of old too much especially if it makes you more $$$ to have a page with "value added" content rather than giving a "concise, definitive, easy to find" answer at the top of the results, can you?

Gotta wonder a bit about their motives and results, personally.

I think, compared to the info we got years ago from the Google reps who posted here most of what we get anywhere today is nothing more than FUD, which is usually best ignored while concentrating on what *works* instead of what they say, because unfortunately for us, what they say and what works seem to be two totally different things lately.