Jump to content

Cre8asiteforums Internet Marketing
and Conversion Web Design


Photo

One Site Or 19?

seo rankings multiple sites

  • Please log in to reply
11 replies to this topic

#1 WPMuse

WPMuse

    Gravity Master Member

  • Members
  • 141 posts

Posted 15 September 2017 - 10:15 AM

I have a client that instead of having one comprehensive site, decided to put up 19 identical sites -- one for each city that they wanted to target.  

 

I had advised that having one site that was a one-stop-shop for their services with individual targeted city pages was the best way to go -- especially when we are talking WordPress sites that need to be updated and monitored.  But they are DIYers and DIY'd each site.  I loaded up the theme and plugins and they took it from there.  All the sites have blank pages (or coming soon), blank Blogs so no content is being added.  They are basically brochurewares.   Duplicate everything just with "cityname" changed.

 

Because they didn't pay attention to updating, the majority of their sites got hacked.  I cleaned all the sites told them how they have to approach maintenance, and provided recommendations (UI, content, blah, blah, blah) moving forward.   Nothing has changed.   They either don't have or don't want to spend the money and/or time.  None of the site are https either.   They didn't like hearing my advice on that.

 

So that's the backstory...

 

I'm here to get input from you guys about what is the best approach at this point in time from an SEO POV.   I believe moving to one https comprehensive site is easier to manage, optimize, track and keep secure.  And less expensive in the long term.  Then, point/forward the 18 city domains to each city's page on the new site.

 

They do not want risk any rankings they have for their city sites.  I asked how they had been tracking rankings and not surprisingly their answer was they search for themselves to determine where they rank.  Ugh.

 

These are nice folks and I would really like to help them succeed.   So next week I've scheduled a telecon to go over their options. 

 

How would you handle this situation if the client were yours?   Am I on the right track?  Thanks in advance for your advice!



#2 glyn

glyn

    Sonic Boom Member

  • Hall Of Fame
  • 3287 posts

Posted 15 September 2017 - 04:43 PM

I'd use subdomains for each city on the main domain and redirect the sites there.

#3 WPMuse

WPMuse

    Gravity Master Member

  • Members
  • 141 posts

Posted 15 September 2017 - 04:53 PM

Being these sites are WordPress we're then still going to have 19 separate installs to track, maintenance, optimize, etc., right?



#4 iamlost

iamlost

    The Wind Master

  • Site Administrators
  • 5488 posts

Posted 15 September 2017 - 07:27 PM

You bring up several points worth discussing, both here and with clients.
1. Google.
Google started as a disruptive SE that had an ad network appended; these days Google is the dominant online ad network with the dominant SE appended, which requires a mind reset (that few have managed).

With the advent of personalisation traditional 'rankings' have lost most meaning, only traffic matters as a solid metric. Again, a mind reset not generally happening, especially by those hawking tools. Unfortunately, a page receiving 10 Google referrals a day may be topping a minor traffic query or page three results of a major. A general distinction can be made, however, not by many DIY/hobbyists. And as you mention personalisation is a highly restrictive ever shrinking bubble, particularly for those signed in.

A critical point to remember with all shared Google data is that it is incomplete, worse it is randomly incomplete. This holds for Search Console data as well. Google is historically fond of bucket sorts: problem (1) being that webdevs are told neither the total number of buckets nor the number being shared at any given moment; problem (2) is that neither the data buckets displayed nor their number are consistent. It is an illusion of openness.

2. Wordpress.
All websites can be attacked. CMS's, especially WP as the dominant one, are regularly targeted. Locking down a WP site is serious business that takes time and requires knowledge, or money (to pay for someone else's). The more sites, the more required. Perhaps the second biggest marketing fail for a site is to have Google (and other SEs) mark it a hazard in query results. The first being customers being inconvenienced or taking losses because of one's site becoming an attack node.

Given your explanation of their business model I'd shrink it to one site. Certainly not 19 - they don't have the resources, period.

3. Combo Pack.
The problem with multiple sites pushing pretty much the same stuff except for specific local info is that doing so puts deciding what to ignore for duplicate content wholly in G's hands. It is a recipe for giving up control of one's business. And potentially making it look incompetent.

Whether subdomains or subdirectories would be best is a synergy of best architecture for best business goal(s).
One note that may or may not be pertinent: Let's Encrypt will not have wild card certs until the new year - this makes free HTTPS certs to cover subdomains a pain at best and impractical at worst.

One can certainly redirect all existing back links as best appropriate to minimise SE values loss. Given the amount of duplicate content inherent in implied architecture there may actually be a values boost as what is currently diversified is concentrated.



#5 test-ok

test-ok

    Mach 1 Member

  • 250 Posts Club
  • 430 posts

Posted 16 September 2017 - 02:51 AM

Personally I'd go with one domain with the different cities within the home pages content and else where where appreciate and maybe create sub pages for the different cities if relevant.



#6 bobbb

bobbb

    Sonic Boom Member

  • Hall Of Fame
  • 3396 posts

Posted 16 September 2017 - 10:57 AM

Perhaps the second biggest marketing fail for a site is to have Google (and other SEs) mark it a hazard in query results.

Just to add.. Not just SEs but some browsers and anti-malware give this warning while surfing. FF is a big red screen with a prominent "I'm outta here" button.



#7 WPMuse

WPMuse

    Gravity Master Member

  • Members
  • 141 posts

Posted 16 September 2017 - 02:10 PM

Appreciate the explanation, @iamlost and @test-ok -- backs up my initial instincts on how to proceed.   Since they were not tracking rankings, other than searching for themselves, what @bobbb states could have happened -- we don't know.  But that's a risk that take if they stay on their current path of 19 sites that they do not keep updated.

 

Thank you for taking the time to respond! ;)



#8 Grumpus

Grumpus

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 6612 posts

Posted 17 September 2017 - 08:04 AM

Oddly enough, there is actually a right answer to this question regarding the area targeting, but most folks don't know about it. So let's take a look... The malware issue has been dealt with above, so I'll just look at the fun stuff.

 

The answer is one site and one page to target all towns - probably the contact page or an areas served page. (I use a combination of both).

 

Basically what you do is go with some schema markup. I don't know the industry we're talking about here - so "Local Business" may not be right, but it's the most general one so I'll use it. Certain industries have their own markup categories and name spaces.

http://schema.org/LocalBusiness

 

I'm assuming that this is one physical location serving a large area. If it's actually got physical branches - you can use branchOf and make individual pages for each branch or break up the page into sections to cover each one. There are lots of different ways to describe the area you cover - and the most useful for you will be to list each of the 19 towns you cover and mark them up with areaServed as the town name.

 

The markup is great because you're describing a specific area that Google (and other similar services) can understand natively. if you're lucky enough to not have a lot of competition in an outlying town, this can get you to show up on map results even if your physical area is outside what it might normally show in a "something near town" search. If there is nearby competition, it's always going to show the three closest, so you're screwed anyway.

 

The markup sucks because there isn't a lot of concise and clear writing on the proper way to do it. I've been tinkering with it for about a year now and am still homing in on the best way to make it all chug along. I can say that I've never had a negative impact on anything by doing it, but some have worked great right from the start, and others I've had to tweak a bit and am still playing with. Typically, the amount of competition and the size/budget of the competitors web site play the biggest role. I'm not doing anything wrong, but I'm just not as "complete" or something.

 

Play with it, though...

 

There's a bit of a learning curve, but when done properly it eliminates the need for these individual pages targeted to an area because you've got individual areas marked up on a single page.

 

Some basic reading to get you started: https://www.schemaap...local-business/

 

I use the "Microdata" method since it makes for cleaner code. With JSON, you end up with blocks of redundant code. With microdata, you mark up your HTML. So your town list goes from:

<p>We proudly serve, Hartford, West Hartford, and Avon Connecticut.</p>

To:

<p>We proudly serve, <span property="areaServed" typeof="City">
    <span property="name">Hartford</span> 
  </span>, <span property="areaServed" typeof="City">
    <span property="name">West Hartford</span> 
  </span>, and <span property="areaServed" typeof="City">
    <span property="name">Avon</span> 
  </span><span property="areaServed" typeof="State">
    <span property="name">Connecticut</span> 
  </span>.</p>

These two sets of code look identical on the page, but it sends a clear signal to Google - both for organic search and local search, as to what you're talking about.

 

This markup is in its infancy, but it IS the future - and using it already helps a lot when trying to target specific things. Basically this single line of code: <span property="areaServed" typeof="City"><span property="name">Hartford</span></span> serves the same purpose as (and eliminates the need for) creating a separate page for each town.

 

You can also mark regions (and Google will figure it out) with zip code lists, geographic shapes on a map, counties, states, countries, or whatever. I still like to use town names because it also gets our targeted word onto the page (and usually into the snippet the searcher sees in organic results) but I've also used combinations of several of these ways of defining a region.

 

Important note: The above code is incomplete - you need to wrap the section in other code/spans to define it as a part of the local business or whatever other type you're using and so on. I'm just giving that as an example of the relevant section we're talking about. Inside your overall encompassing spans, you'll want to include your phone numbers, addresses, and various other relevant items (hours of operation?) etc.

 

Super Pro Tip: If you have your local business stuff set up and properly associated with your web site, you can update your site with holiday hours and that sort of thing and your listing on Google and Bing will update on the next crawl without you having to go manually do each one. This requires that the site is well spidered and that you set it up far enough in advance for a crawl and update to happen (usually 10 days or so in my experience - though on sites we update a lot, I've seen it as quickly as 48 hours).

 

Hope that helps. If none of your competition is doing this (which is quite probable in small business circles) you'll get AMAZING results. If you're competing against some juggernauts who already employ it, it's one relatively simple way to level the playing field a bit.

 

G.



#9 earlpearl

earlpearl

    Hall of Fame

  • Hall Of Fame
  • 2502 posts

Posted 17 September 2017 - 01:38 PM

Oddly enough, there is actually a right answer to this question regarding the area targeting, but most folks don't know about it. So let's take a look... The malware issue has been dealt with above, so I'll just look at the fun stuff.

 

The answer is one site and one page to target all towns - probably the contact page or an areas served page. (I use a combination of both).

 

Basically what you do is go with some schema markup. I don't know the industry we're talking about here - so "Local Business" may not be right, but it's the most general one so I'll use it. Certain industries have their own markup categories and name spaces.

http://schema.org/LocalBusiness

 

I'm assuming that this is one physical location serving a large area. If it's actually got physical branches - you can use branchOf and make individual pages for each branch or break up the page into sections to cover each one. There are lots of different ways to describe the area you cover - and the most useful for you will be to list each of the 19 towns you cover and mark them up with areaServed as the town name.

 

The markup is great because you're describing a specific area that Google (and other similar services) can understand natively. if you're lucky enough to not have a lot of competition in an outlying town, this can get you to show up on map results even if your physical area is outside what it might normally show in a "something near town" search. If there is nearby competition, it's always going to show the three closest, so you're screwed anyway.

 

The markup sucks because there isn't a lot of concise and clear writing on the proper way to do it. I've been tinkering with it for about a year now and am still homing in on the best way to make it all chug along. I can say that I've never had a negative impact on anything by doing it, but some have worked great right from the start, and others I've had to tweak a bit and am still playing with. Typically, the amount of competition and the size/budget of the competitors web site play the biggest role. I'm not doing anything wrong, but I'm just not as "complete" or something.

 

Play with it, though...

 

There's a bit of a learning curve, but when done properly it eliminates the need for these individual pages targeted to an area because you've got individual areas marked up on a single page.

 

Some basic reading to get you started: https://www.schemaap...local-business/

 

I use the "Microdata" method since it makes for cleaner code. With JSON, you end up with blocks of redundant code. With microdata, you mark up your HTML. So your town list goes from:

<p>We proudly serve, Hartford, West Hartford, and Avon Connecticut.</p>

To:

<p>We proudly serve, <span property="areaServed" typeof="City">
    <span property="name">Hartford</span> 
  </span>, <span property="areaServed" typeof="City">
    <span property="name">West Hartford</span> 
  </span>, and <span property="areaServed" typeof="City">
    <span property="name">Avon</span> 
  </span><span property="areaServed" typeof="State">
    <span property="name">Connecticut</span> 
  </span>.</p>

These two sets of code look identical on the page, but it sends a clear signal to Google - both for organic search and local search, as to what you're talking about.

 

This markup is in its infancy, but it IS the future - and using it already helps a lot when trying to target specific things. Basically this single line of code: <span property="areaServed" typeof="City"><span property="name">Hartford</span></span> serves the same purpose as (and eliminates the need for) creating a separate page for each town.

 

You can also mark regions (and Google will figure it out) with zip code lists, geographic shapes on a map, counties, states, countries, or whatever. I still like to use town names because it also gets our targeted word onto the page (and usually into the snippet the searcher sees in organic results) but I've also used combinations of several of these ways of defining a region.

 

Important note: The above code is incomplete - you need to wrap the section in other code/spans to define it as a part of the local business or whatever other type you're using and so on. I'm just giving that as an example of the relevant section we're talking about. Inside your overall encompassing spans, you'll want to include your phone numbers, addresses, and various other relevant items (hours of operation?) etc.

 

Super Pro Tip: If you have your local business stuff set up and properly associated with your web site, you can update your site with holiday hours and that sort of thing and your listing on Google and Bing will update on the next crawl without you having to go manually do each one. This requires that the site is well spidered and that you set it up far enough in advance for a crawl and update to happen (usually 10 days or so in my experience - though on sites we update a lot, I've seen it as quickly as 48 hours).

 

Hope that helps. If none of your competition is doing this (which is quite probable in small business circles) you'll get AMAZING results. If you're competing against some juggernauts who already employ it, it's one relatively simple way to level the playing field a bit.

 

G.

This is working for me also.  I've been a little late to applying schema to web pages but it works.  I'm in the process of adding more.  Specifically I'm applying it to websites that completely respond to the situation that Grumpus described:  Local businesses whose services are applicable in multiple nearby towns while being located in one town.

 

From my perspective I'm also finding the same thing as Grumpus;  some of the applications tend to work better than others.  I add pages to a site that describes the service and title them with something that includes the specific nearby town.  Instead of identical content on every page except for the town name we create individual content for each town.

 

One particular finding that might apply to the example Grumpus provided is I'm finding that I need 1 page per town, versus what appears to me to be a page that is servicing 3 towns, Hartford, West Hartford and Avon noted above.   The difference in rankings appears to me to mostly be a function of competition on the web, but time and experimentation will tell more.

 

Now too many of these might end up looking spammy to the human eye and the SE eye.  This is still to be determined.  I'm targeting certain specific towns for testing purposes, competitive purposes and focusing on the towns with the largest populations and the populations wherein the demographics are best for the different smb's.  Additionally my targeting is based on both traffic to the sites analyzed via G analytics geo data and G adwords geo data, plus our own long developed data bases that include the home towns of leads and sales.

 

Schema shows up!!!!



#10 WPMuse

WPMuse

    Gravity Master Member

  • Members
  • 141 posts

Posted 18 September 2017 - 09:24 AM

Thank you @Grumpus and @EarlPearl!!   Will definitely look into how to leverage Schema better!   

 

In this particular case they aren't really "local" -- they are virtual nationwide with no physical location.

 

Has anyone told you guys that you rock lately? :)


Edited by WPMuse, 20 September 2017 - 09:46 AM.


#11 earlpearl

earlpearl

    Hall of Fame

  • Hall Of Fame
  • 2502 posts

Posted 02 October 2017 - 02:51 PM

I was looking at this for one of our urban markets.  I still have a long way to go.   I went back through leads data by towns and city, then visits by geo area, and impressions/clicks by adwords for the smb's in this area.

 

I noted 1 major city and about 40 suburban towns that merit this treatment.  I have 4 done.   

 

Of the one's done it appears we get a relatively higher rate of actual leads to traffic/imressions.   It works.

 

hmmmm.   Got about 36 more pages to go.   ;)



#12 earlpearl

earlpearl

    Hall of Fame

  • Hall Of Fame
  • 2502 posts

Posted 05 October 2017 - 05:07 PM

So I am looking at this harder and I haven't added schema to anything in over a year.   As Grumpus referenced above there is a debate between using microdata and JSON.  I'm reading through the "arguments".  Haven't made a decision on which way to go forward.  I'm not much of a developer so any additional insights, Grumpus are much appreciated.

 

In the context of doing this I'm looking at max about 35 more blog pages specific to one major city and another 35 cities/towns in a metro region that are substantial draws to the business(es).  (two in the market).   Not sure if I'll get to each one.   I do want specific pages per town to better compete on the web for these "major" sources.  If I overthink it, I'll also realize that one of these pages can roughly impact 2-3% of the whole market....but that isn't the way to proceed is it?  Gotta just keep knibbling away.

 

One thing I've seen over a few years is an incredibly effective use of schema both on the web and in real world effectiveness using the "rating" schema that lends itself to acquiring reviews, getting them into your system and showing them on the web.

 

The best application I've seen is a custom built system for this garage door business that turns up some eye catching results for searches such as "Garage Door Repair Montclair" or "Garage Door Repair Caldwell NJ"

 

Of course Google , with its endless monetizing and breakup of the serps is making life harder.

 

On a desktop, under the ads and under the Local Pac/Map section are top results for a business called Precision Garage Door of NJ (or something like that).    Schema markup on ratings has been a part of creating highly visible serps wherein the following occurs:

 

For a roughly non competitive topic the town town name gets in the URL and in titles and optimized target phrase.  It helps generate higher rankings  (as might the schema for location).

 

The powerful visual element is the rating/review element that shows like a charm.   There is a lot more to it, as the operating business has figured out how to:

a.  Give great service

b.  Ask for local reviews on site

c.  Get those reviews to the appropriate page

 

I know the developer and his work with the business operator.  Much of the beauty of this relies on running a great ship that generates great and high volume reviews on site.  Then there is the code that shows it.  Finally there are the high volumes of calls that come off the organic results that have highlighted the reviews.   Its been beautiful.   (Iamlost would love it  ;) )

 

After all is said and done, though google is changing the game.  The 3 or 4 ads at the top of the screen have sucked away organic visits and organic results.  PSDNJ is now advertising when they didn't need to before.  Its not because the organic results aren't working, its because google is sucking the organic traffic out of the market with its increasing monetization of search.

 

Lastly and I have to look at this more closely:  While the precision door schema shows in google it does not show in bing.  As I currently reviewed Bing is currently accepting one form of schema markup and not the other.   Google is accepting both.

 

Again:  Any comments to add, Grumpus:   I've got UP to (I know I won't hit this) 35 pages/cities to tackle.  I'm currently debating which code to use.





RSS Feed


0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users