eBusinessBlog.org Leveraging marketing & technology to solve business problems.

4Dec/060

Title tag optimization

Posted by Eric Long

This article got me thinking on how I handle title tags on various sites. It's a good, quick read and great points are made by both strategies (put brand/site name first in the title tag, or put it last). Title tag optimization should also take into consideration any SEM intiatives you have in place. SEM can help drive home your brand on a search engine results page (serp) (sign up for this hitwise whitepaper on brand marketing in search engines, or read an excerpt here). For example, for Levolor.com, we have the following (typically 1st placement in SEM results and always the 1st placement in natural results when searching for "levolor"):

levolorserp.jpg

As you can see, we employ a combination. Because this industry is dominated with independent online retailers, we attempt to improve natural search results using the above title tag on our homepage, thus the reason for "window blinds, treatments, shades, & drapery hardware." These keywords hit home on most popular search phrases that drive traffic to our site in combination with products we offer. Natural search results are very important to us right now as we came into the online world late in the game and have a lot of traction to make up. Our home page title tag is a bit wordy right now and by itself, doesn't lend to obvious clicks if you're really looking for the Levolor site. Additionally, we are a well-known brand name to many, but the blind and shade industry is home to many new consumers who are brand agnostic -- it's very possible for consumers to initially discover us via search engine results without having heard of us before.

Using the meta tag and a combination of SEM, we can very easily craft our message to drive home the point that this is the official manufacturer site, and not an online retailer selling our product (of which there are several). The fact that we are the first natural search result also gives us flexibility to be a little more "wordy" with the homepage title tag. Levolor is a very well-known brand name and as a result, we also place Levolor at the end of every title tag on the site. There is slight overlap in the "official manufacturer line" which could be tweaked in either our SEM or homepage meta tag. Since it's already established in SEM, we may modify the homepage meta tag in the future so we have less redundancy. This will take some A/B testing to prove out which method is more effective at driving clicks.

On the flip-side of this coin, let's take a look at SuperMotors.net search results:

supermotors-serp.jpg

As you can see, we've got some work to do. First, we do not do an effective job of indicating on the hompeage in the title tag (or meta tag, for that matter), that this is SuperMotors.net. Actually, the data is there, but the title tag is so long on the homepage, that is gets truncated on the serp. Furthermore, the title tag and meta tag are very redundant. The title tag should almost be converted to "SuperMotors.net | Photo, audio, and video hosting" and just leave it at that. In anyone's bookmarks list, the current homepage title is really too long and isn't easily identified as SuperMotors.net.

In the second result on this page, you can see how SuperMotors.net is included in the title tag. This is done dynamically on every page of the site. Having your brand name or domain name in the title tag of each page is recommended if users will see it in their bookmarks or on serps. Placing it before the title of the page or after the title of the page will be a decision you make on a case-by-case basis depending on your brand recognition, competition for natural search, and your current SEM initiatives that pick up where your natural search leaves off.

Tags: , ,

2Dec/061

apache mod_deflate reduces bandwidth usage by 27% on SuperMotors.net

Posted by Eric Long

In the ongoing battle of optimizing, tweaking, and testing, in November we enabled apache's mod_deflate module on SuperMotors.net. Literally adding about 10 lines of code to httpd.conf (the apache config file), all text-based content on our pages now use gzip compression when delivered to web browsers that support gzip compression. By using this tool, you can see if your site currently utilizes gzip compression. When testing SuperMotors.net with this tool, our homepage size was originally 49K, but with gzip compression thanks to mod_deflate, it is reduced to 9K. This is an 81% savings in bandwidth usage.

So, when implemented over the course of the entire month, we saw an overall 27% reduction in bandwidth. Surprisingly, our absolute unique visitor count was down 14% for November (holidays do this to our site), but page views actually increased by 1%. Page views are ultimately the driver in bandwidth utilization. I interpret this to mean that fewer users were able to do more on our site because they were downloading content in less amount of time.

So, why does the test tool from above show an 81% reduction in bandwidth yet we only saw an overall reduction of 27%? This is because images (and other files) are not compressed when delivered to web clients. Since we primarily serve images and videos, the rest of our bandwidth still remains largely untouched by mod_deflate -- which is by design.

This bandwidth reduction is very good news for us. As previously posted, we are leaving the world of colocation and moving to managed hosting -- with a new billing model. The "total bandwidth used" billing model favors us even more now that we've reduced bandwidth usage by 27%. This translates into squeezing more data out of the pipe than previously anticipated. We'll be able to maintain our fixed costs and increase revenue, thanks to this little module. The more data we push out, the more page views we serve, and the more revenue we make from our CPM-based advertising model with Tribal Fusion and our in-house ad inventory (read the challenges we face when we introduce Ajax functionality and how it'll affect our CPM-based ad model).

The mod_deflate module does increase CPU usage due to the need to compress each page sent out. However, this had very little impact on us as we had plenty of processing power to spare. Your results may vary, so keep an eye on CPU usage when you implement this module. The slight increase in CPU usage was worth the risk, because the 27% reduction in bandwidth was a much bigger gain for us.

12/4/06 Edit: Ajaxian has an article on gzip compression with some user responses. Interesting insights (in the comments section) on the law of diminishing returns with gzip compression used on ajax-enabled pages. 

Tags: , , , ,

1Dec/060

Page views out. How to monetize Ajax-enabled sites on a CPM ad model?

Posted by Eric Long

In this post, the author predicts 4 more years until the "page view" metric is no more and advertisers/publishers are scrambling to find other ways to measure traffic on a site. This is something I have had in the back of my mind as well. While I'm not interested in attaching a timeline-to-extinction to the page view metric, it is of legitimate concern to us on SuperMotors. We generate revenue via user subscriptions (for advanced site features) and banner advertising. We sell our own banner space but also rely on the likes of Tribal Fusion and Google AdSense to fill our unsold inventory (and there's a lot of it!). What's interesting is that with Ajax, you essentially eliminate the need for several page views to navigate a site quickly and easily. Granted, pages will still need to remain, but certain functionality within pages definitely can take advantage of Ajax.

For example, we allow users to post media files to their accounts on SuperMotors. Here is the hierarchy of clicks (from the home page) to post 1 picture, sound, or video:

  1. (From Homepage) click "My Garage"
  2. Choose "Edit Album" from your list of registry entries (vehicles, atvs, boats, etc.)
  3. Choose the album section to add to, or create a new one
  4. Select the file you wish to post, click submit.
  5. Site refreshes. Repeat for each file. (it should be noted that subscribers can post more files at a time and use drag & drop functionality we have implemented, thus requireing fewer page views at this step)

Including the visit to the homepage and the 5 page views above, that's 6 total page views. Each page view displays 2 advertisements for our site. That's 10 total ads being displayed in the course of posting one photo. Add 2 page views for each additional photo/file posted. Our members have posted nearly 300,000 photos, sounds, and videos -- over time, this adds up to significant revenue from CPM-based advertising campaigns like Tribal Fusion and our in-house ads.

Now, enter Ajax (which we will be implementing in the next 2-3 months). Ajax could potentially eliminate page refreshes on steps 3-5, thus bringing total page views down from 10 (for 1 file posting) to 3 (for 1 file posting), with no additional page views generated for each file posted. Makes for a superb user experience but at the cost of sacrificing advertising revenue.

So, how do we measure this traffic and use of the website? Or do we? From an analytics standpoint, yes, I want to measure the use of our site even through Ajax functionality so we can determine if users like it, are stumbling with it, or if it doesn't work at all. From an ad revenue standpoint, Tribal Fusion and Google AdSense policies state that you can't automatically refresh ad units on a page. So, do we include ad units within the Ajax functionality? There are technically many screens passing by and that's a lot of potential advertisements to expose a user to, just like we do today with our page view/non-ajax model. Where's the happy medium?

So, for any site selling ads based on CPM, Ajax is a double-edged sword. Increase usability, decrease revenue.

So, what will the new metric be on Ajax-enabled websites (or Ajax-enabled features within a site)? Time spent on site? Will ads run in 15, 30, or 60-second allotments, similar to television? How will you prevent someone from just leaving a page open forever and allowing ads to rotate through? I'm excited to introduce Ajax to our user base to see how they react to it and also to see how we can best create the best of both worlds: improved usability & sustained revenue.

Tags: , , ,

1Dec/060

Applying web usability to real life

Posted by Eric Long

My wife and I built a new home this summer and are still in the midst of getting it organized. One of our biggest challenges is dealing with the fact that we no longer have a basement -- in North Carolina, apparently this is quite common; we come from Minnesota, where it's mandatory. Anyway, I've been on a recent "usability in the home" kick as we've optimized the locations of our belongings based on our own internal usability study.

Usability for the morning routine
Subconciously, I had been doing this all along, but never really stopped to think about how usability could be applied to the home. For example, our master bedroom has a master bedroom closet, but you must walk through the master bathroom to access the closet. In our old house, this was not the case, so we had hanging clothes in the closet and folded clothes in the dresser in the master bedroom. Getting ready in the mornings was a fairly quick task. With our new place, there's a lot of walking back and forth, and all 3 (closet, bathroom, and bedroom) are considerably larger than in our previous home. I was finding myself zig-zagging back and forth from the closet, through the bathroom, and to the bedroom every morning when getting ready for the day. Normally, I am not particularly astute in the mornings, so it has taken me a few months to finally realize what I was doing. Because we had previously had a different layout, we gave no additional thought to where our clothes would be positioned in our bedroom and closet. I'm now optimizing the layout by attempting to put all of my clothes in our master bedroom closet to eliminate the back-and-forth trips that waste time.

Usability in the kitchen
Another example is our kitchen. We've been working on the best way to organize pots, pans, dishes, glasses, utensils, etc. based on workflow in the kitchen and commonly used items. Glasses are above the dishwasher and fortunately, our dishwasher and glass-holding cabinets are at the end of the counter in the kitchen -- makes for very good usability. It's quick and easy to access, and since more of these typically fit in the dishwasher than anything else, unloading the dishwasher directly into the cabinet above is also quick and easy.

Usability in the garage
Yet another example is our garage...a man's haven. I enjoy working on our cars, so I have many, many tools and liquids (oil, transmission fluid, cleaning products, etc.) to keep organized. I have yet built a work bench yet because I'm still evaluating if I can buy a modular workbench system (like this, or this), or if I am in need of a custom-built work bench. I have however gotten most of the clutter off the floor by using Rubbermaid FastTrack organization kits. Much like a business, I am evaluating my three options based on time, money, and the value it adds. My rollaway toolbox has been organized for several years now, and it's interesting to see how my interpretation of organizing my toolbox is compared to friends who also have their own vast array of tools. A good friend of mine who spent a considerable amount of his time helping me work on my vehicles would always joke about how my toolbox was setup poorly compared to his. I guess a usability study is in order. ;) I'm really waiting to pull the trigger on a work bench after I get some more of my own usability studying in with the current garage setup/layout.

Usability for...lawn mowing? :D
Lastly...and out of fear that I'll get laughed at if I mention how else usability is being applied in our household, is that lovely task of mowing the lawn. At our previous house, our lawn was big enough to require a riding mower. The riding mower alone transformed the chore from a 2-hour pushmower nightmare to a 45-minute dream. I regret to say that I did not spend enough time figuring out the best mowing routes for my lawn in the shortest amount of time -- I had thought the rider solved all of my problems, after all, it provided a 62.5% reduction in time. Relating this back to web usability, I think this is quite a common mistake with web marketers. A significant improvement is great, but that doesn't necessarily mean your job is done. You have to constantly monitor progress and determine through testing and analysis if more optimization can be done. At our new place, we have a much smaller yard and only require a push (self-propelled) mower. So far, I have concluded that mowing the lawn perpendicular to the street is the quicker method (as opposed to mowing parallel or diagonal to the street). I need to figure out alternate paths so the grass doesn't grow in one direction from me mowing in the same direction all the time.

Optimize your site for your customer's primary tasks in doing business with your company
Again, mowing can be related to website usability. Mowing is a chore, it is something that I enjoy doing because I take pride in the appearance of our yard, but I would definitely rather be doing about 1,000,000 other things than mowing. For this reason, I need to minimize the time spent on this task to maximize my time elsewhere as well as maintaining my own sanity -- nobody really wants to mow all the time. On websites, consumers often feel the same way. They may use your site as a tool to get certain tasks done, are not interested in Web 2.0 features, social media, bells and whistles, unless these features help them get their task done more efficiently. They may not be there for an emotional experience, they are there to perform tasks as quickly and efficiently as possible. This is a balancing act many companies play when they sell products online and also provide support and account management of the products online as well (think online banks, cell phone companies, etc.).

Build, launch, test, optimize, re-launch, test. Repeat.
Furthermore, I think web marketers get caught up in the hype at times and are so focused on trying to architect information and cram as much data online as possible that they lose site of the task at hand -- making it easy. A lot of this is trial and error, A/B testing, etc. And most importantly, your job is never complete! Building, launching, and forgetting about it is a bad habit to get into. Just because you have a website, doesn't necessarily mean you're doing anyone any favors. You may be doing a disservice to your customers if you do not analyze how they use your site -- they could be wasting valuable time working with your site and can very easily switch to a competitor that has their kitchen properly organized, knows where their clothes need to be in the morning, and knows the best lawn-mowing path for their yard. You never know until you do the research and get in the mentality of always improving what you offer.

Tags:

Tagged as: No Comments
1Dec/060

Artistic Design vs. Usability Design

Posted by Eric Long

Today, while working with my favorite UI/design guy, we were nit-picking over subtle navigation elements (in a step-by-step process, is tabbed navigation better? floating buttons labeled, 1, 2, 3, 4, etc.?) and your standard "add to cart" buttons. In my ifinite Web 2.0 wisdom (ha!), I said I wanted the global navigation to be larger tabs and the footer navigation (add to cart, save to wishlist, etc.) to be larger and more obvious as well. At one point he said "that's too Fisher Price." I laughed. The WordPress software has me a little biased towards big fonts, buttons, and text. However, there's power in this because it took me all of 2 minutes to setup the blog and figure out the software. Simple and straight forward.

His comment did get me thinking about how much the web is a constant struggle between usability and aesthetics. The theories on both are a bit blurred. Our problem right now is trying to apply brand guidelines for what was previously a 100% print-focused brand on the web. Not good (small serif fonts, limited color palette that doesn't work well with text, colors that don't translate well on computer monitors...you know the drill). Fortunately, we'll be taking closer look into this for '07.

Anyway, I don't have a definitive right way of approaching it, other than I think the struggle between artistic design and usability design is a healthy one to have, regardless of the website's content. In our line of business (blinds and shades), it's as much ease-of-use as it is fashion, and the emotional connection. Consumer marketing online is challenging, and I really look forward to the first usability study after we push some of our latest changes to the live site. Heatmaps are what I'm really interested in doing this time to determine the visibility of our navigation and to see if we're any worse off with "Fisher-Price" buttons vs. smaller, "on brand" buttons. At the end of the day, it's in your best interest to have both graphic designers who are non-web-savvy and web-savvy UI/designers eye-balling the site. This, combined with information architecture based on natural talent (and some data from web analytics and usability studies, too :) ), should always produce a superior website.

Tags: ,