Matt Cutts talks about what makes a good quality site

[Whitey]

Top interview between Eric Enge and Matt

Eric Enge: We always speak to our clients about focusing on activities that are brand building.

By doing things that help build your own reputation, you are focusing on the right types of activity. Those are the signals we want to find and value the most anyway.

Does that make sense?

Matt Cutts: Yes, it does. By doing things that help build your own reputation, you are focusing on the right types of activity. Those are the signals we want to find and value the most anyway. Just promoting your site on a spammy blog network that no one would ever choose to visit is not a good strategy.

It’s wild to see some blog networks just repackage the same spammy sites and services and have the nerve claim that their content is “Panda and Penguin compliant” when the quality of the network is clearly not at the level that even a regular person would choose to read it
[stonetemple.com...]

Plus many more points. Well worth a read and some feedback comments.

[realmaverick]

If Google search worked in the way they suggest, we shouldn't have to consider Google. We should be building websites that visitors love to use.

There are so many considerations now, that often usability is sacrificed to please Google.

[indyank]

It is interesting that Matt cutts suggests Hipmunk as an example. Yes, according to him it has great visualizations. All I can agree with him is on its visualization but it is too geeky for a normal user.

If I remember right, he had tweeted about how he likes hipmunk about an year ago. I then tried to see why he liked this site. Then I suggested this site for flight & hotel search and booking to several friends, colleagues and others and then made sure I got their feedback on how they found the site.

In that test, several of those users came back saying it looked different and colorful with charts, etc, but they found it too confusing to use. They didn't find it user friendly and weren't pleased with their experience.

All I could say is it looked geeky as someone here suggested but was too geeky even for the geeks. For example when you navigate to this url [hipmunk.com...] from google, you don't get to see the book button by default and a lot of things on that page takes time to understand for a normal user. The book button pops up only after a few clicks which an ordinary search engine user might not even be aware of.

Another interesting thing about this site is its pages that you find on google's index. It would be completely different from what google has in its cache.Take the above url for example. All those charts and maps are only for real users while the search engine gets a page of pure text!

[arikgub]

Hipmunk is a nice example of how one can present the most simple data in the most counter-intuitive way. Putting flight data on a gantt chart? Well.. I definitely prefer visualization in the form of the industry-standard table, but it's a matter of personal taste I guess.

What's important for this discussion though, checking on search ranking data aggregators - hipmunk is not even getting 0.1% of Google traffic that Kayak is without any fancy visualizations.

[indyank]

What's important for this discussion though, checking on search ranking data aggregators - hipmunk is not even getting 0.1% of Google traffic that Kayak is without any fancy visualizations.

The reason is quite obvious, isn't it?

[Whitey]

The reason is quite obvious, isn't it?

@indyrank - i need help - blinded. What is obvious to you.

You know what, unique content alone means nothing.

Plenty of dialogue with other's and personal experience with $100k's of unique quality content, on unaffected and affected sites. Neither have responded in response to Panda's supposed quality requirements. Absolutely, no long tail response.

Now I'm not saying it's a bad thing in the long run, but in a "crowded space" you might be wasting priority resources heading in the wrong direction with flawed support of congested kw terms and longtail to surface in the SERP's.

With the prevelance of "brand" success in congested results, you have to really ask yourself, how is what Matt and Eric are saying linking into this?

[aakk9999]

What's important for this discussion though, checking on search ranking data aggregators - hipmunk is not even getting 0.1% of Google traffic that Kayak is without any fancy visualizations.

The reason is quite obvious, isn't it?

@indyrank - i need help - blinded. What is obvious to you.

Obvious reason: hipmunk is not ranking for top keywords related to flight searches whilst Kayak is.

[indyank]

Hipmunk's Information architecture is very poor and they don't have proper indexable pages for search engines. They have made a half hearted attempt in getting a few pages indexed in Google and other search engines but they do it through cloaking! What is presented to the visitors and the SERPs for the same URL is different.

Last but not the least, users don't have the same pleasant experience that Matt the geek is having on hipmunk.

To summarize, hipmunk looks like a site put together by a few innovative college folks for a geeky web competition. They don't seem to have a strategy for search engines.

So I am really surprised that Matt quotes this site as an example not once but twice (or probably many times).

[indyank]

May be the hipmunk founders (who are quite popular) don't want to depend on search engines for traffic but were more interested in making and promoting a popular mobile app that users would straight away use without doing any search.

[claaarky]

It must be hilarious being Matt Cutts, knowing how things work, trying to convey that to people and then watch them misinterpret everything.

I think the concept of how to rank well in Google is really quite simple. Martinibuster made some really great points earlier in the thread (and Tedster of course). If it's all about how the majority of humans actually react to your site then, when you're considering boilerplate content, authority, brand, and so on, just think how 'most' people will react to your content and you have your answer.

You can't fool all the people all of the time - that's why Google's ranking mechanism now relies heavily on how 'people' see your site these days (the user experience). It's like a constant voting system that we all participate in every day, leaving a trail of clues on what we thought about each site we visited, just by how we reacted to each site without even realising it or consciously thinking about whether the site is a brand, authority, trustworthy, useful. We just react instinctively and that tells Google everything without even needing to know if you have boilerplate text, unique or duplicate content, and so on.

I don't think it matters what you do these days, as long as your users love it. Quite liberating to think that way.

[indyank]

It must be hilarious being Matt Cutts, knowing how things work, trying to convey that to people and then watch them misinterpret everything.

There is no misinterpretation. Hipmunk doesn't get the kind of traffic that their competition get from Google search and a few other have mentioned this here.

Remember that information architecture and link structure is vital and there cannot be any genuinely good UX without them. After reading your posts here on a few threads, I do know your obsession with UX but remember that algorithms only take into account the data they have and not individual's perceptions. Not even Matt's.

[edited by: indyank at 9:19 am (utc) on Jul 19, 2012]

[indyank]

Anyway, do keep us posted about your recovery story.

[Whitey]

I'm not overly convinced about these comments regarding UI and content improvements in the crowded spaces, per my earlier post. In fact I'm convinced it has little effect versus brand.

I'm seeing brand sites that simply don't cut it in terms of quality that dominate the top ten organic listings in a crowded vertical. Seems to me, in verticals that Google considers high traffic, they have eliminated anything non brand. It's almost like a pure manual consideration.

This interview kinda confirms the importance of establishing a reputation, but if you're up against the big players, you may waste a lot of effort in the wrong direction. That's what I am suggesting.

[indyank]

It's almost like a pure manual consideration.

surely that plays a major role in all these quslity algos with names of animals :)

[indyank]

Google's quality army consists of people on their payroll and those working for some outsourced companies. Quality algorithms are still separate runs and are manually triggered as well. The underlying original ranking algorithms are still in place and the quality algorithms are run on their output. These are supposed to be using signals designed by the Google folks based on the quality army's input.

[claaarky]

indyank, the way I see it, Google's research prior to Panda was all about trying to gain a more detailed understanding of statistics they already had which reflect human perceptions of websites. They then realised those statistics were a reliable way of identifying all sorts of problems with a site (duplicate content, boilerplate, low trust, too many ads, etc.) and implemented Panda to demote or promote sites according to where each site sits in the quality scale relative to other sites it is competing with.

Whitey's comment is interesting about brands that don't cut it in terms of quality but rank well. If people trust a brand more, they react better to it, so what may appear to be low quality sites are actually valued sites by most people who use them. It doesn't matter what they look like, it's whether people trust and value them more than other similar sites.

Google is giving people what they want. Almost anything that doesn't make sense these days can be explained simply by looking closely at the site from the point of view of the people who would use it and considering why most of those people would like that site.

[Whitey]

Matt Cutts: Google does try to mirror the real world. We try to reflect the real-world importance of things as we see that reflected in the web. Brands sometimes are an indicator that people see value, but it isn’t the only way that people see value. There are many other possible indicators that something is important and worth surfacing in the search results.

A brand could be potentially useful, but it’s certainly not the only lens to interpret the world. There are lots of signals we use to try to find the results that bring the most value to users. And whether or not someone is an advertiser does not matter at all.

One of the great things about the web is that it still offers up-and-coming businesses opportunities to build their own reputation online. This can enable them to succeed even though other companies may have large advertising budgets.

.... yes, really I'm not seeing what Matt says in these crowded areas. and I'm having problems with consistency across this and other commentaries by Googlers. Do I have a problem with my personal interpretation in these communications..... need help.

On the one hand he says anyone can compete, then on the other hand he says you cannot compete in crowded areas because it's all been done. So how does this make sense in the same interview?

If you're talking niche, and Matt seems to encourage this in his interview, OK - i agree. But then there's no competition. Is he saying quality doesn't matter in crowded spaces occupied by brands because the brands will rank anyway?

I mean that's what we're seeing isn't it, in the major verticals, with an exception here and there.

[claaarky]

Whitey, I think he's saying that a small business can't go head to head with massive companies right across the board, but if you pick an area to specialise in then you have a chance at doing a better job at that than most big companies. By specialising you create value which people appreciate and the signals that produces will help you out rank the big companies in those areas.

However, if ALL the bigger companies see you're doing well in a niche that's important to them and they're losing business to you, they would easily be able to focus some resources on improving their sites/business to beat yours and get back above you (assuming they know what they're doing online of course - some don't if they started as bricks and mortar).

Of course, it would take ALL your competitors reacting in that way or lots of newcomers appearing in the same niche to give you a problem, so niches are a better bet for small businesses. As your business expands you can specialise in more niches and eventually maybe reach the size of the big companies. Then your well known name and great reputation will allow you to take them on head to head right across the board, because you'll be on a par with them.

Just like the real world.

[netmeg]

On the one hand he says anyone can compete, then on the other hand he says you cannot compete in crowded areas because it's all been done. So how does this make sense in the same interview?

Sure anyone can compete, but anyone can't compete on *anything.* I'm not going to compete with Amazon on books, for example. I might compete with them on other products, though.

Is he saying quality doesn't matter in crowded spaces occupied by brands because the brands will rank anyway?

Not quality, trust and authority. Brands get some amount of trust (that the transaction will be legit) and authority (that they are an authority source for the product) simply by virtue of being a brand. That's one of the reasons you want to work to become a brand. Most people would rather give their money to someone they've heard of or have reason to trust.

In order to compete, you need to give off as many trust and authority signals as the brand. Just being yet another seller of the same products won't do the trick. You're likely not going to out-authority Amazon on Books, Best Buy on electronics or Zappos on shoes.

But take it down to say, men's mountain hiking boots, or Thai cookbooks, or whatever - some subcategory - you might be able to find a way to wedge yourself in.

This is not a Google issue, believe it or not. This is a human type marketing (and business) issue.

[diberry]

This is not a Google issue, believe it or not. This is a human type marketing (and business) issue.

Great point - it's exactly the same ofline. If you want to make it as a restaurant, you can't just do exactly what Applebee's does, because they'll always have more money and be able to do it better. But if you have your grandmother's terrific recipe for [whatever], that single item might be all you need to ensure you have all the customers you can serve. It can build your reputation so that people come to you rather than Applebees for that item. Play the rest of your cards right, and you can do very well.

All things being equal, the company with more money is almost bound to win. So you have to make sure all things are not equal - that your site is better in some way, however limited.

[netmeg]

Or start small (or narrow) and work your way up (or wide).

[Whitey]

Any evidence that Google's learnings on Adwords quality scores may be part of what Google considers "quality"?

Wouldn't you think the principle would be similar ? Thoughts?

[netmeg]

Not entirely the same thing. AdWords quality scores hinge greatly on CTR and relevance to the ad.