The Internet Marketing Driver: Glenn Gabe's goal is to help marketers build powerful and measurable web marketing strategies.

Monday, April 28, 2008

Performing Keyword Research and SEO, Don’t Assume You Know the Right Words to Target!


How to perform keyword research.When it comes to Natural Search and SEO, performing extensive keyword research for your given business is critical. In my experience, most people are too close to their businesses to understand what people are really searching for. You may have seen this too, like using terminology and acronyms that only industry folks use. Or, if you have been in an industry for 20 years, then you surely must know how people search the web for your products or services, right? Don’t make this mistake! You might get a few by chance, but I’ll guarantee you are missing huge opportunities if you ignore keyword research. So don’t do it. :-)

Skepticism is Good
Right now, some of you are probably skeptical. That’s good, and I’ll give you some examples to curb your skepticism. Let’s say you are in the summer rental business at the Jersey shore. If you performed keyword research for your business, you would find that beach rentals is searched 4X more than summer rentals, which in turn is searched for 10X more than nj shore rental and beach house for rent. Without keyword research, it’s all based on opinion… I’ll take real data over opinion 99% of the time. That’s one thing about keyword research that I love… it takes guesswork out of the equation. Armed with data, you can make the right decisions from the beginning of your seo project before wasting time, money, and effort.

Here are some more quick examples:
Do you sell jewelry? Did you know that the keyword jewelry showed up 12X more than the keyword jeweler in Keyword Discovery? Let’s shift our focus to a buggy business? Pest control is searched 10X more than exterminator. Sell infant bedding? Did you know that the keyword baby bedding showed up 16X more than the keyword infant bedding? That's 16X more! I think you get my point… Do your keyword research and move opinions to the side…focus on real data, real searches, and don’t waste your time and effort trying to rank for keywords that won’t pay off.

Keyword Research Tools:
The two most popular options for keyword research are WordTracker (WT) and Keyword Discovery (KD). I have used WordTracker much longer than Keyword Discovery, but I can tell you that I’m really digging KD. Both are great tools and will give you excellent data. WordTracker’s database holds approximately 330 million metacrawler searches where Keyword Discovery holds over 36 Billion from over 200 search engines. I often find myself using both tools to find the right keywords, and if you focus on SEO, I would probably keep accounts with both services. Their prices won’t break the bank… WordTracker is $59/month and you can get a fairly large discount for an annual purchase ($329 for the year). Keyword Discovery is $70/month and I believe both are a small price to pay for finding the right keywords via the multitude of tools they provide. Your return on investment should be huge, to say the least.

A Closer Look at Keyword Discovery:
Let’s say you sell women’s jewelry and wanted to do some keyword research. You would log into KD and enter jewelry in research mode (see screenshot below). You will see the top searched terms with the keyword jewelry in them. The one column provided at this stage is “Searches”, or the number of times that the keyword was searched for over the past 12 months.

Screenshot from Keyword Discovery (Research Screen):
Click the image below to view a larger version.
Researching a keyword in Keyword Discovery

Now, if you click the icon for “Analyze”, then you will see those keywords with some additional columns like “Occurrences”, “KEI”, and “Predicted Daily”. Occurrences shows the estimated number of webpages the keyword shows up on. KEI is a formula for showing you how competitive the keyword is. I can dedicate an entire post to KEI and you can read more about it on the web, but not all keywords are equal from a competitive standpoint. KEI helps you determine which keywords are worth going after and which ones might be too tough to rank for. Predicted Daily is just that, the predicted amount of times that the keyword is searched for each day.

Screenshot from Keyword Discovery (Analyze Screen):
Click the image below to view a larger version.
Analyzing a keyword in Keyword Discovery

Drill in further to find targeted, long tail keywords…
At this point, you can click on any keyword to see a list of longer tail keywords containing the original word you clicked on. For example, click diamond jewelry to see all the keywords in the database that have the words diamond and jewelry as part of the keyword. This will include diamond jewelry watches, black diamond jewelry, diamond jewelry stores, etc. Then click “Analyze” again to view the additional columns I mentioned above.

I have my keywords, now what?
Let’s say you performed keyword research, found your target keywords, and have the spreadsheet sitting in front of you. Now what? Well, you would want to include these keywords on your website within the right HTML elements. For example, you would want to use these keywords in the title tag, the meta description tag, in the page copy, within your page headings (H1, H2, etc.), in your navigation and anchor links, and in image alt text. You would want to take a hard look at the pages on your site and optimize each one for the specific content they hold. Yes, it’s a lot of work, but well worth it. If you have a large site, definitely work with your developers on how to optimize the site dynamically. I can also write an entire post on optimizing the elements I just listed, but you’ll unfortunately have to wait for that one! I want to keep this post from being 25 pages long. ;-)

In SEO, your work is never done.
Once you optimize your website, you can’t just sit back. Like everything in web marketing, you need to track your results and refine your strategy as needed. Maybe some of your optimization isn’t paying off like you want it to, so you may need to go back and research more terms and optimize more pages. Or, you might want to tweak some of your pages, based on changes in your industry, your products, or seasonality. If you are using a robust web analytics package (Omniture, Coremetrics, Google Analytics, etc.), then you should have some great data to analyze. Then learn from the data and make changes to improve your rankings. I have written several posts about web analytics and you should definitely check them out.

OK, I’m sure you are chomping at the bit to get started (at least I hope you are!) Definitely stop back and let me know how keyword research works for you and your business. Go ahead, real data awaits!

BTW, did you know that SEO is searched for 3X as much as Search Engine Optimization? We are lazy typists, aren’t we? Quick tangent...do acronyms affect your business? ;-)

GG

Labels: , , , ,

Monday, April 21, 2008

E-Commerce Customer Reviews, Common Pitfalls That Can Impact Sales


Customer Reviews and e-Commerce ImpactIf you run an e-commerce website, then chances are you’re fully aware of how ratings and reviews can impact sales. I think everyone agrees that enabling customers to review products is a powerful way to leverage user-generated content to improve the overall buying experience for visitors. That said, what constitutes a quality review, what are prospective customers looking for in a review, and how can the various types of reviews impact conversion? Not all reviews are created equal, so if you are thinking about implementing reviews on your website, I've listed four pitfalls below to watch out for. You might be able to plan your implementation with these in mind!

When Are Reviews Necessary?
Before I list the pitfalls, I also wanted to quickly explain when reviews are beneficial to prospective customers and when they aren’t necessary. I don’t really need to read a review for GAP jeans or a Banana Republic belt or a Canon calculator. We all know they are high quality and they are fairly simple items. In my opinion, it comes down to price, safety, and how a product impacts your life. Lesser known brands from smaller companies might require reviews from consumers versus bigger, well known brands. In addition, how the product will impact your life is an important factor for reviews. For example, you would probably want to read reviews for infant car seats, a high end camcorder, or an expensive piece of furniture. Price, safety, and how that product impacts your life will dictate if reviews are necessary. Again, just my opinion.

Without further ado, some pitfalls of e-commerce reviews include:

1. When there are no reviews!
If visitors are expecting to find reviews and they can’t find any for the product they are looking for, then there’s a chance they will lose confidence during the purchase process and move on. That's especially true if you boast about your website reviews! Even the slightest second thought can be a conversion killer. If you are having trouble gaining reviews from customers, then I highly recommend launching a campaign to drive more reviews. Leverage your in-house email list and get people back to your site to review the products they purchased. Heck, give them an incentive…maybe 15-20% off their next purchase in exchange for a review. Having no reviews can impact more than sales for the product at hand, it could be interpreted by visitors that you don’t have enough volume or customers to generate reviews. Again, not a good thing when someone is ready to buy from you…

2. Who is actually writing the review?
Let’s face it, customers aren’t stupid. They want to read reviews from similar people who are in similar situations. For example, anonymous reviews are close to worthless in my opinion. Depending on what I’m buying, I want to hear from someone in a similar situation (who has a name). I don’t need to know them obviously, but I want to know that it’s from a somewhat quality source. For example, whenever I buy a tech book (like a programming book), I want to hear from other developers. I have an entire bookcase full of programming books and not all were written perfectly, to say the least… So, something like, “I’ve been programming for 13 years and this book was outstanding. The chapters started with a solid foundation, then moved to basic coding examples, and then real-world coding projects. The samples always worked (don’t laugh, many programming books come with code examples that don’t work) and the book is a great reference for when I get stuck.” A review like that would get my attention. Last year, I wanted to read reviews when I was looking for a new golf driver. I definitely wanted to hear from golfers in my skill range. Hearing from a scratch golfer wouldn’t be helpful, nor would hearing from a beginner. Solid reviews helped build confidence and got me closer to the sale… Now, I still needed to blast a few golf balls at the range before buying the club! You get my point.

3. Mixed Reviews (Great reviews mixed with poor reviews).
Products that have mixed reviews will tend to give customers less confidence in moving forward with a purchase. It makes sense, right? How can there be 5 fantastic reviews and 5 horrible reviews? That’s a definite red flag for me (and others too). I wouldn’t take a risk on buying something online that half the reviewers thought was a waste of money. Would you? I’m not referring to reviews that rate a product as mediocre. I’m referring to those weird set of reviews where some people loved the item and others hated the item. Seems fishy, doesn’t it? I always move on… It also leads me to think that there’s a fake review or two in the mix… Read on. ;-)

4. Fake Reviews
I think we’ve all come across these, right? (probably written as anonymous)? It’s funny, most people can’t write a positive review if they are forced to. Don’t believe me? Go ahead…try and write a positive review for something that you don’t really like or that’s your own product. I will guarantee you that it sounds obvious. ;-) If I come across a fake review, there had better be a real one for that product…or there’s a good chance I’m moving on. My hope is that you can pick out the fake reviews, and if you can’t, then hopefully there are plenty of other reviews for the item at hand.

So there you have it, four pitfalls when offering customer reviews on your website. I definitely believe reviews are a great feature to add on your e-commerce site, but I think you should implement them with a good understanding of the types of reviews out there and how they can impact conversion, user experience, and return buyers. My hope is that you will keep these pitfalls in mind and try to form strategies to overcome them. For example, launching campaigns to increase reviews, dealing with anonymous reviews, tracking the impact of reviews, etc. When you break it down, reviews can help your customers make informed decisions, and also help you determine the right products to sell on your website. Now if you’ll excuse me, I need to go review the Tonka Truck I just bought my son. ;-)

GG

Labels: , , , , ,

Friday, April 11, 2008

LiveHttpHeaders and SEO, How to Check Your HTTP Response Headers for Red Flags


Using LiveHttpHeaders to Check Your HTTP Response Codes for SEODoes your website throw a 302 when it should throw a 301? Does it throw a 200 when it should be a 404? Are there 500’s thrown on your site that look like 404’s? Do you think I’m insane yet? Hear me out…

Whether you understand the introduction above or don’t know what I’m talking about, there’s still something extremely important for you in this post. Every time you load a webpage, your browser REQUESTS a file and then the server provides a RESPONE to that request (also called a Response Header). Response headers can help you identify critical issues on your site (especially from an SEO standpoint). Now, you probably have a few key questions.

1) How do I check my response headers?
2) What should I be looking for?

Although I can’t cover everything about response headers in this post, I will answer the two questions listed above and provide some examples along the way.

Let me start by answering the first question since it’s the easiest… I highly recommend using LiveHttpHeaders, an add-on for Firefox that displays http headers in real time (as you browse webpages). This tool can save you a lot of time and possibly help you diagnose some serious SEO-related issues. I will answer the second question later in the post.

Install LiveHttpHeaders Now:
First, visit the LiveHttpHeaders project website and install the add-on. You will need to restart Firefox after installing LiveHttpHeaders. Once restarted, you can trigger LiveHttpHeaders in two ways. You can click Tools, LiveHttpHeaders, which will trigger a new window where you can view header responses in real time as your browse the web. You can also click View, Sidebar, LiveHttpHeaders to view response headers in a sidebar within Firefox. I prefer the new window, since I have dual monitors and it doesn’t take up any browser space. :-) Either way works fine.

To quickly test it out before we go any further, go and visit Google with LiveHttpHeaders running. When you hit the homepage of Google, you will see a bunch of information scroll by in LiveHttpHeaders. For our purposes, let’s look at the top of the window (the first piece of information sent back to you). I have stripped out some of the information you don’t need to focus on for this example.

http://www.google.com/

GET / HTTP/1.1
Host: www.google.com
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.1.11) Gecko/20071127 Firefox/2.0.0.11
Keep-Alive: 300
Connection: keep-alive

HTTP/1.x 200 OK
Content-Type: text/html; charset=UTF-8
Content-Length: 2547


I have bolded the response code, which shows 200 (or ok). 200 is a good thing... and now I can explain more about response headers and codes. By the way, where you see 200 in the window is where you would also see other response codes like 301, 302, 404, etc.

Back to our two questions for a minute. The second question was “What should I be looking for?” In a nutshell, you should be looking for the file requested and the response code sent by the server. Let’s start with a definition of a response header and identify some core http response codes.

As I mentioned earlier, when you load a webpage, two things happen. There is an http request (by your browser) and then an http response is sent (by the server). There is a response code returned as part of the http response. Some of the following codes might be familiar to you, and others aren’t. If you focus on SEO, then get to know them…they can really help you diagnose problems on your website.

Some of the most common http response codes are:

200 – ok (the webpage was returned ok)
301 – Permanent redirect (seo friendly) J
302 – Temporary redirect (don’t use this unless you absolutely have to!)
400 – Bad request (uh, not good)
401 – Unauthorized (you need to be authenticated)
403 – Forbidden (doesn’t matter who you are, it’s forbidden)
404 – Not found (not necessarily a bad thing…I’ll explain more later.)
500 – Internal server error (something went very wrong processing the webpage…)
502 – Bad gateway (also not a good thing)

Why are http response codes important? One of your goals as an SEO is to enable the bots to easily index your site. You don’t want them to get caught up in any way, shape or form. For example, 302 redirects are not the SEO-friendly way to tell Google where a page you removed now resides (you should use a 301 redirect instead). So constantly providing 302’s would be a very bad thing to do. Or how about throwing a 200 (ok) when you really should be throwing a 404. For example, the page isn’t there, but you just told Google that it is. Again, not a good thing to do. Therefore, finding 302’s, 404’s, 403’s, 500’s, etc. is critical to creating a clean path for the bots, which means you can have more of your content indexed and at a solid frequency. Let’s take a look at how LiveHttpHeaders can help you out.

Checking Your Response Headers:
Let’s take a look at a hypothetical situation. One morning you wake up and decide that you want to increase your natural search rankings. You launch an SEO initiative and get moving quickly. The first thing you want to do is to audit your current site structure (since you know that without a sound and clean structure, you’re dead in the water). So as part of your audit, you want to ensure your response headers and response codes look ok.

You fire up LiveHttpHeaders and visit your website:

* You hit the Homepage, 200 returned,
* You visit a Top Level Product Category, 200 returned,
* Then you try and visit a product detail page and you hit a 302 redirect. Hold on… You find that all links to your product detail page go through a 302 redirect. This was implemented as part of a recent code change. This is something you would want to change ASAP. The content on your product detail page is obviously important so you wouldn’t want to be throwing 302’s prior to the bots hitting those pages…
* But it doesn’t stop there. You know that you changed dozens of older product pages recently and created new URL’s. You check out the old URL’s and find 302 redirects to the new product pages. You would want to change that too… and provide a 301 redirect from the old page to the new page, safely passing link power from the old page to the new one.
* Then you check out some product categories on your site that have been removed completely (you won’t be selling those products anymore), but you find 200’s instead of 404’s. A 404 (page not found) is the proper response code to throw in this situation, as it will tell the engines that the page has been removed and that it should be de-indexed. You don’t want the page to be indexed if it’s not actually there, right?
* Last you check some newly added pages and find they are not displaying correctly. It looks like they aren’t on the site, which means you should see a 404. But…the server returns a 500 (or internal server error). Again, not a good thing as the bots traverse your website content and this is something invisible to the naked eye as you test your website. You would need to be checking response codes to find this issue…

OK, I think you get the picture! Keep in mind that a full SEO assessment covers much more than just checking response codes, but it’s an important part to revealing SEO-related issues. And you know what? Sometimes it’s darn easy to find a serious issue that can be resolved fairly quickly. For example, providing a 302 redirect right on your homepage! Or throwing 200’s for any page that’s been removed from your site.

Think Like a Bot:
I’ll end this post with an analogy. Imagine you needed to check every room in a 10 story hotel to document the type of TV that’s in each room. But the elevators don’t work properly, some of the staircase entrances are locked, and every now and then the room numbers change on you. Would you have an easy time completing your task? Would you keep trying to come back to “index” each room? Or would you stop a few rooms in and say, “Hey look, Lost is on.” And then sit back and watch the show….and forget about the TV’s (or your content). ;-)

GG

Labels: , , , ,

Tuesday, April 01, 2008

The DVR and Its Effect on TV Advertising Recall, Do Your Commercials Stand Out?


DVRs and TV Advertising RecallOr does that really matter? More on that later. First, I’m a big DVR user and have been one for a number of years. I bought my first TIVO about 5 years ago, quickly added a second, and then moved to the Comcast HD DVR last year. So, I read an article in late February about a study conducted by General Electric’s NBC Universal to document the recall power of TV commercials when DVR users were whipping through them at 6x speed. In other words, do you recall an advertisement as you are fast forwarding through commercials on your DVR? By the way, that’s exactly how I watch the shows I record. In addition, if I choose to watch a live show, I just start watching the show 15 minutes in (for a 1 hour show) and then I can still fly through the commercials. Now back to the testing. The following quote is directly from the Wall Street Journal article:

Tracking biometric measurements such as eye movements, heart rate and sweat, the study found that the ads people concentrated on the most and recalled the most shared several traits. The most successful ads concentrated the action and the brand's logo in the middle of the screen, didn't rely on multiple scene changes, audio or text to tell the story, and often used familiar characters. People were also more likely to remember an ad in fast-forward mode if they had seen it once before live.

Glenn’s translation: A big static logo in the center of the screen. ;-)

Uh, tracking biometric measurement?? OK... The most successful ads didn’t rely on audio to tell the story?? The more I think about this topic, the more I think that this entire study wasn’t necessary. I guess it was for networks trying to hold on to TV advertisers at all costs, right?

Paging Dr. Gabe… Dr. Gabe Please Come to the Living Room:
So, I decided to conduct my own study. That’s right…and without biometric measurements, my heart rate wasn’t monitored, my eye movements weren’t scanned, and my sweat glands weren’t checked! I flipped on my TV, drilled into my recordings, armed with a high tech toolkit for a high tech doctor of technology (a pad and pen) and started my first show. My goal was to fast forward through each set of commercials at the highest speed and see which ads I could recall. That is, if I didn’t go into convulsions first! :) Since every study needs a name, I am calling it “The Strobe Logo Study” conducted by Glenn Gabe, Technology Scientist at Large. ;-)

Without further ado, here are the results! I will also give my quick analysis of the results following the data.

LOST on ABC (Probably my favorite show right now...)
1 Hour in Length
I started zipping through the commercials at the highest fast forward speed. Believe me, the highest speed is darn fast…each segment of commercials was over in a few seconds. I remember seeing a flash of the Wendy’s logo (centered on screen), an Applebee’s logo and a Dunkin Donuts logo (both also centered on screen). More on ad positioning later. Each was up for a flash…probably a quarter (1/4) of a second. Also keep in mind that the three logos I remembered are big brands that have been advertising for years. This obviously helped with my recall of their ads.

The Sports Reporters on ESPN
30 minutes in length
The only advertisement I remember seeing was CDW. I saw this logo twice during the show and for a little longer than the logos in LOST. This intrigued me, so I rewound the show a little to watch the CDW ad. It ends up they sponsor the show (see, I didn’t know this because I never watched it in real time!) So, as a lead into the show, they have a voiceover say, “The Sports Reporters is sponsored by…CDW”. This takes a few seconds, which leaves their logo on-screen longer. I saw this twice during the show when whipping through it on my DVR. This is an interesting note for TV advertisers. Also, the logo was centered on the screen, an important factor during my scientific study. :)

American Idol on FOX (like I had to tell you…)
60 Minutes in length
I remember seeing a Citizens Bank logo (I think) and an Infiniti logo. And, maybe…just possibly… a Lowes guy. Then again, it could have been a 7 Eleven guy or Wawa guy. I told you…it’s darn fast! Any quick movement or elaborate camera angles looks like a blur in fast forward. It makes sense, though. The common thread for TV commercial recall was becoming apparent. Any commercial that ended with a large logo centered on screen had a chance of recall (unless I blinked during the 1/4 of a second!)

Dancing With The Stars on ABC
60 Minutes in Length
I’ll cut to the chase…I recalled 3 ads, Advil, Petsmart, and a Nivea product shot. Again, each logo was centered on screen and fairly large (and the Nivea product was also large and centered). Are you seeing a trend here?

Here’s an interesting side note:
When you watch a recorded show, it will abruptly stop at a random frame at the end of the show…which is rarely when it fades to black. Dancing with the Stars ended on a BBC Worldwide America logo. So, since my DVR stopped on that frame, I saw this logo for about 10 full seconds before it returned to my DVR menu. If you are still looking for ways to appear in a DVR world, this could be one of them. BBC Worldwide America was not an advertiser, but that spot could be opened up for advertising...

Conan O’Brien on NBC (Just a brilliant comedian!)
30 Minutes
I literally didn’t recall any specific advertisement… Not one. So, I went through it a second time to make sure my eyes weren’t just fried out of my head from the previous shows… Nope, not one ad, logo, brand, etc. I guess none of the commercials used a logo centered on-screen. Read on for my scientific analysis of my DVR study.

Is this what it’s come down to for TV advertisers?
Triggering convulsions in people to see if they remember a flash of a logo? Really? I know not everyone has a DVR (yet), but if this is the type of study we are conducting, then there’s something very, very wrong. I can hear the scientists who conducted the NBC Universal study now. {In my best European scientist voice} “Yes, it seems that even 1/8 of a second can impact your brand’s recognition in the mind of lowly consumers. TV advertising is still hugely powerful and we may turn to 60 strobe flashes of logos in the future versus 30 second spots. Sure, some people may be hospitalized from the flashes of light, but it’s all worth it if the Advil logo shows up, is recalled, and then helps the hospitalized person's family overcome its collective headaches by using Advil.” OK, I’m a bit sarcastic, but it’s hard not to be!

How To Really Enhance Your TV Advertising, ONLINE
When I think of the cost to produce high end TV commercials, then the cost for airing those commercials, and then combine that with the growing number of DVR users, it doesn’t take a mathematical genius to understand why traditional TV advertising is not in good shape. I’m not saying that TV commercials should go away…but I believe that you need to supplement your TV campaigns with other campaign elements to maximize exposure and engagement. For example, I always recommend adding an online viral component to your TV commercials. Have a micro-site or landing page to engage your brand and advertisement, to help with lead generation, to work in a contest, to spark user generated content, to add a blog, etc. and then utilize online marketing channels to drive visitors there. So, combining your high end TV commercials with a robust micro-site, and then utilize paid search, organic search, email marketing, social media, blogging, display advertising, etc. to drive people there is a smart way to go. Then you’ve got yourself a serious campaign, covering all avenues, and using innovative methods. Versus…trying to justify your TV commercials with “flashes of a logo” or what I’m calling “The Strobe Logo”. There is so much you can do online to enhance your TV advertisements, and at a reasonable cost. To me, it’s a necessary addition that can unleash the true power of viral marketing.

Now let me go put an ice bag on my eyes and give my poor DVR and TV a break! Until my next scientific experiment, this is Dr. Gabe signing off. Does anyone know a good eye doctor? Maybe I should get in touch with the people who went through the original study to see who they recommend!

GG

Disclaimer: I am not a medical doctor, scientist, or PHD. Please don’t ask me for medical assistance or to conduct clinical studies. If you need assistance from a medical doctor, please consult your healthcare company for a referral. I am not authorized to prescribe medication, recommend time off from work, or advocate medical procedures. :)

Labels: , , , , , , , , , ,