Redesigning Without Losing Traffic Options

[hitechsol]

Hello there,

We are expecting a client to assign us an SEO project for their eight year old website. The existing site is reasonably ok getting tons of traffic every single day. However, there is a lot of room for SEO work to be done.

The issue we are facing is that they are already selling 500 used cars through their website per month. How to do the SEO work while not losing the healthy amount of traffic? Any thoughts on maintaining and increasing the amount of traffic on the site?

Thanks in advance.

Regards,

Maqsood Ahmed

[Black_Knight]

Start with a thorough and detailed analysis of where their existing traffic and sales are coming from. Understand the demographics of that traffic as best you can.

Then... don't do anything that will screw it up.

Make sure all your changes will enhance and improve the experience of those existing sources of custom, and those pre-existing customers. This is especially true of those already on the site, in not adversely affecting the quality of the after-landing experience.

So don't add keyword-stuffed headings that get a high ranking but cause higher abandonment of the page or cause customers to question the integrity of the site and business. Don't mess too much with things that are working well - the old adage being "if it ain't broke, don't fix it".

There are many things you can do to help a site rank well that are entirely imperceptable to the average customer. There are many things that can help rankings that require no on-site changes at all. There are probably hundreds of small changes that will not only help gain better search positions, but more importantly may increase the existing custom conversion rates.

When a site already has some good sales, approach SEO with another adage firmly in mind: "A bird in the hand is worth two in the bush". It means don't do things to attract more of the birds you don't have in your hand if those actions will cause the one you have to be lost.

[Jozian]

Nice summation, Ammon.

There are many things you can do to help a site rank well that are entirely imperceptable to the average customer. There are many things that can help rankings that require no on-site changes at all. There are probably hundreds of small changes that will not only help gain better search positions, but more importantly may increase the existing custom conversion rates.

I know we have great lists and guides here for general SEO activity. But it might be nice to have a list or guide to: How to Improve existing SEO without blowing up the current traffic and conversions.
Ammon hits the generalities, anyone care to add specific tactics?

[hitechsol]

Excellent Ammons, Thank you so much for the wonderful insight on the issue.

There are many things you can do to help a site rank well that are entirely imperceptable to the average customer. There are many things that can help rankings that require no on-site changes at all. There are probably hundreds of small changes that will not only help gain better search positions, but more importantly may increase the existing custom conversion rates.

Could you please elaborate on the small changes you mentioned above if you have few moments?

Thanks again for the big help.

[Black_Knight]

Well, I'm no SEE, but simply use of good tags and elelemnts can go a long way.

Just as with that other thread, you've leapt right into changing the tags, the on-page factors, and the copy (headings and links) without fully determining whether or not the existing tags and copy are what is making the site sell 500 cars a month. If your changes mean that the next month they sell only 400 cars, are you going to pay the difference in income?

The opening post is entirely correct in placing immense concern on not harming what is already a successful site. Selling 500 used cars per month, over the inernet, where they cannot be seen, touched, sat in, or test-driven is no mean feat. The site is doing something right, and you must know what that is before you risk messing with it.

Put it another way, if you were building this site from scratch, would you offer a cast iron guarantee to sell 500 cars per month? If not, then accept that this site has things to teach you about selling cars, that take precedence over what you can teach the site about rankings.

Until you have run thorough web analytics throughout the site, and made very certain you know what not to do, so that you don't make even one of those 500 customers each month not buy, the only changes you can make must be either effectively invisible to the user, or must be off-site factors, such as link building activities.

Also, recalling the "bird in the hand" issue, be sure you understand who is buying and identify the differences demographically between those visitors who buy and those visitors who do not. What if the analytics show that not one of the search referred visitors has ever converted? There are people who are buying, and people who are not, and before you worry about getting the rankings, you must know which rankings the buyers are looking at compared to which the vast majority of visitors, the non-buyers are looking at. You may well find that the word "car" or "cars" are never once used in the search of an actual buyer, for just one example why this is so vital.

The average person does not sit at there computer saying, "Hmm I'm bored, oh I know, let's buy a car! I'll google 'cars' and buy one". You need to know what are the trigger words. Are successful sales based on people searching for a particular make and model or just make? Do the successful sales pay attention to the year as a search term? Was price a deciding factor, or merely one of several, along with colour and condition? Is Google image search actually driving more end sales than the regular search?

Whatever you do, do not make any visible change to the website until you understand these things.

We spoke of professionalism a lot recently. Remember the hypocratic oath: First, do no harm!

[hitechsol]

Whatever you do, do not make any visible change to the website until you understand these things.

You are right Ammons, We have got to be extremely careful of what we are going to do with the site. To my understanding, the site has been successful mainly due to off-site factors including the age of the domain. off-line marketing activities in various parts of the world and the like. Anyways, thank you sooooooooo much for sharing your valuable knowledge with a newbie SEO like me.

[joedolson]

2) There is no conceivable way that "upping" a tag to a "heavier" weight could cause any negative results.

Wow, is that a dangerous statement.

No matter WHAT you're aiming to do, any time you jump into an absolute statement like that you're asking for trouble. (Speaking of absolute statements...)

No, I can't prove to you that this activity could cause trouble. But that's not what I'm interested in: what is most interesting and problematic about the approach you're suggesting is that it has nothing to do with study, thought, or analytics.

Personally, the first step in nearly any SEO project for me starts with statistics. Where are customers coming from; what are they doing; and, most importantly, where aren't they going.

Yes, in many cases some basic re-coding can be beneficial --- but I don't take that for granted.

The problem isn't actually with what you're suggesting - it's in the way you're taking it for granted. The first step isn't to change the page - it's to understand the page and why it's functioning the way it is. Then, as a second step, you can begin making trial changes in pursuit of improvement.

I can, in fact, easily conceive of ways that "upping" the "weight" of a tag could have negative impact. No, it's not a WISE use of the method - but what you've said is that you don't find it conceivable. Try this:

CODE

<h1 class="h1">Recent News:</h1>
<h1 class="h2">Autocrat Can Coffee Available Throughout RI & SE MA</h1>
<h1 class="p">
Look for Autocrat Premium, Autocrat Decaffeinated, and Autocrat 100% Colombian at fine retail establishments such as Stop & Shop, Wal-Mart, Dave's Fresh Marketplace, A & J Seabra Supermarkets, Brigido's Fresh Marketplace, Trucchi's Supermarkets, Belmont Martketplace, McQuade's Marketplace, Eastside Marketplace, Ruggieri's Market, Shore's Fresh Marketplace, Carcieri's Supermarket, Roch's Market, and Phred's Drug, as well as several other local independent retailers.
</h1>

I'm not a search engine, but if I saw code like this, I'd immediately be inclined to be suspicious. All they've done is increased the weight of their tags, right? But in my mind, there's absolutely no question that they've done this with no thought except to try and influence search engines. I don't really believe that I'm the only one who thinks this.

[DianeV]

I have to go with Ammon on this one.

Thing is, sure, you can just start slapping tags and keywords on the pages. However, without studying the site, you don't know what pages do what. What if there's a golden page that isn't optimized but converts like crazy with content stated just as it is ... and what if in optimizing it, you destroy or lessen its effect? This isn't as outlandish an idea as it may seem. And plenty of direct marketers use language that isn't of the sort that you'd think their pages need to rank for — and they may not. But that golden page is worth gold. If you just go ahead and do the usual, and the sales drop, then that will be a problem you've created. Ouch.

Optimizing an existing successful site isn't like building a new one from scratch. There are issues to deal with, as in this case, where the site is doing very well at the point where you take over the SEO. And the bottom line isn't necessarily just to get higher rankings; it's to increase the sales. That means, necessarily, ensuring that the current successfulness of the site isn't hindered. So, unless you have a sure-fire way to increase sales without having to preserve whatever is currently causing conversions, I'd suggest treading very lightly indeed.

Now, I know, Autocrat, that you're saying that you didn't want to reiterate what was said before — understood, and no reprimand intended here. Thing is, though, that since we're conversing in text, that means body language doesn't play a part, so voicing whole thoughts can be helpful. But, that's how it goes with text-based communications, sometimes. Understood.

[DianeV]

Aw, this is really not that intense, Autocrat, or at least it shouldn't be. I'm certainly not taking it that seriously, and I don't ask or expect everyone to agree with me; I don't wish anyone to feel s/he has to agree with anything at all. It's just a discussion, after all — an exchange of ideas.

Now, I have no argument against using H tags, etc. Generally that's good advice.

However, for the sake of argument (and because I've run into this before), I can well imagine a scenario in which a "golden page" (or series of pages) is made up mostly of words that you wouldn't necessarily want to have the page rank for but which are direct contributors towards the sales-worthiness of the page. So, before launching into doing anything (and assuming that we know *how* to optimize), it can still behoove one greatly to find out what's going on with the site — and to find out what information was left off the site. Because *then* you're better equipped to know what to do with the site, and how to approach it.

[Wiseguy]

Totally agree with Ammon's advice.

I would also do the following:

Create a PPC campaign on Google AdWords in order to test keyword efficiency.
Many people don't realize that Google AdWords is an amazing testing platform that helps you discover within days stuff that could take you months to complete in the offline world.

Use it to learn what are people actually looking for, and how well these keywords actually convert to traffic and sales on the website you are trying to promote. No guessing, just pure data.

Also, try Google Website Optimizer - a relatively new tool for A/B split testing.
This way you can test new page versions and compare them to the existing pages in order to measure the effect on the conversion rate. It is a safe and efficient method if you do it persistently over time.
Make sure to test 1-2 page elements at a time, not more - otherwise you will not be able to know what actually worked.

As Google put it in their own words:
'Rather than sitting in a room and arguing over what will work better, you can save time and eliminate guesswork by simply letting your visitors tell you what works best.'

[egain]

Think you hit the nail on the head Ammon, until you have a good idea of the site, its performance and where/when site traffic comes from, making wholescale changes to the site is probably inadvisable.

As Ammon mentioned, there are a number of things you can do outside the remit of on-page changes including link development, removal of duplicate content issues (the number of sites I have taken over where incorrectly implemenmted test areas exist that have been spidered) etc can all be done without having to put a single finger on any site code.

Really can't go against anything Ammon says, really its case of knowing your client, knowing your audience, and implementing the right strategy accordingly.