Sem-Faq

Search Engine Marketing 4 Less

Search Engine Optimization Benefits

The benefits of Organic Search Engine Optimization (A Primer for Online Retailers):

So, you are a retailer trying to extend your bricks and mortar “Cashmere shawls and Silk Scarf” business on the web. Or perhaps you have sourced a great product and want to sell it online. Or you are a seasoned e-tailer in online marketplaces such as Yahoo Shops or Amazon Universe, and the transaction costs make you want to high tail out of there. Or you are tired of the bottom feeders at E-Bay who are unable to truly appreciate your unique products. And do you truly want to have your shop look like it's a part of some virtual mall? To avoid casting pearls before swine, your experiences lead you to the logical conclusion: Set up shop by yourself.

How do you to make sure that your website is not in the middle of the Sahara desert? After all, your wares will not have exposure to the teaming millions of E-Bay junkies anymore. Or if you were a brick and mortar store, you will have to develop a new set of “regulars”. Or would you?

The situation is  not quite so bad, since you can plan ahead, and put a plug in for your e-commerce website while still  selling  at Amazon. A “multi-channel selling” strategy using something simple like a 5% off printed coupon, every time you ship the item to your amazon customer. Great idea, but limited to your existing customers. Should you not aim higher and gain new customers through search engine visibility? The good news is that brick and mortar marketing has many analogs to the online world. The common sense decisions that a small retailer makes to promote her offline business are useful metaphors for online business promotion.

Let us make a few assumptions to get us off the ground, and to focus on the marketing of your business. You have this great source of cashmere shawls and silk scarves and you want to sell it online through your website. You have already developed it using a professional firm, or yourself. You are also using a website statistics package to measure your performance. Without actionable measurement its guess works. It is impossible to have consistent improvement based on “hunch”. You would think that this was self evident! You also need to know about this software called a Robot(affectionately called “bot”) of which there is no equivalent in the bricks and mortar world. Sometimes, I wish there were!

What is a bot and what does it do?

We can imagine the bot to be a busy body, checking out every shop in the city. It starts with (say)  the chamber of commerce directory and starts visiting all the retailers. Every retailer it visits, it makes a note of their prominently displayed items (indexes them), and asks the business owners if they know any other business that they might wish to recommend “refer” to him. If the bot already has visited the business, it just awards the business some bonus points. The number of bonus points it awards depends on the “importance” of the referring business. If the bot has not visited that business before, it adds it to it's “to-visit” list.  If the bot never gets to your shop, well it won't know about it.

The search engines use this agent, “bot”, to rank sites and keep the “indexed” list. When you use a keyword in a search engine, you  are effectively asking for information out of that indexed list. And you definitely want to be ranked high for the keywords of your interest. Research shows that 2 out of 3 people don't look beyond the 5th site when they are trying to use a search engine results. 9 out of 10 don't look beyond the 20th site.

What is “split testing”?

Assume your shop in the mall had two different doors and everyone entering from the door “A” got to see your most profitable shawl displayed right by the door at eye level. While the door “B” people got to see that shawl along with other shawls, but somehow contrasting from them. Which group of people would end up buying the shawl? Sure, you can guess.  But the real answer is that you don't know until you try it out. Remember, professionals don't guess. This technique of testing your hunches before going full scale on them, is vital to your success. Do you sometimes see mall sales like “half off sale”, and a few days later “50% off sale”. Confusing? Most likely split testing. Your hypothesis about what's going on is soon confirmed when the retailer puts the same thing “2 for the price of 1”. The marketing mavens of that store are busy measuring what is effective. The final arbiter of the measurement is the cha-ching sound in the ca$h register. All these experiments are done on a smaller scale and the results applied to a full scale campaign.  This should be easy to do with a good web analytics package. Don't make a move without it.

One of the first concerns of any retail business should be generating “targeted traffic” to their shop. Let's go “offline” to illustrate this point. Imagine that you just started this store. One of the first “marketing” activities you will do is put up a sign on  your store describing the items you sell (in this case, shawls and scarves), instead of (say) “deli” or “produce”.

The bot need to be told the same. Except that for the bot to be convinced that you really are selling shawls (and not “deli” or “produce”), you have to write a crafted copy with appropriate density of keywords. In practice you would need to make a separate page for all the keywords you  wish to target, in this case “cashmere shawls”, “fine wool shawls” etc., as well as another page for “silk scarf”, “silk scarf gifts”. These pages are called “landing pages” and this process is called “On-Page Optimization”.

The choice of keywords is very crucial. If it's too broad, your traffic will not be focused enough. If you were to optimize on the word “gift”, the person who is searching for “gift” may very well be thinking of a baseball bat for his son(instead of shawls and scarves). You don't want to spin your wheels on something which will provide a very low “conversion rate”(a fancy term for people who end up buying from you, as a retailer). Such people may end up buying but the odds are against it. On the other hand, too narrow scope may also be bad. If you optimize for “angora shawls Kashmir”, many people may not find you since they may not search for something so obscure. Attracting “targeted” or “qualified” traffic is a science, and should be treated as such. With hypothesis testing, skepticism and proof.

There are some tools such as “word tracker” which provide you statistics on how many times a certain keyword is used. For instance in September 2005, SEO and its variants were used (top 10 shown).

67867

 seo

 13856

 seo company

 10611

 seo services

 9399

 seo firm

 7494

 affordable seo

 6066

 indianapolis seo

 5720

 seo optimization

 5092

 seo expert

 4654

 seo consultant

 4623

 affordable seo services

We have to mention the importance of testing your ideas against reality. If you think a keyword is useful, you can  optimize. But even with the help of word tools, can you really tell that your best customers are being reached? You have to test different keywords, measure the effect of your test with a web analytics package, and see if the new keyword is getting you more “targeted” or more desirable traffic. This continuous improvement process must be used to maximize what we call “visitor efficiency”, which is the ratio of people who enter your website, to the people who end up “buying” (or agreeing to give email and phone number, if your goal is “lead generation”).

In real estate business the cliché is “location, location, location”. Same applies for retailing. Would you choose to set up shop by the “auto parts store”, the “army navy discount store” or by the flower shop? I would say the flower shop is your best bet, given that people who buy flowers may also want to get your silk scarf as a gift. Ideally the flower shop may refer it's customers to you (while you reciprocate the kind gesture). If you are a really nice guy, and have some interesting product to sell, a lot of other neighbors, such as the chocolate shop and the cigar shop in the same strip mall may also refer you. They are even more likely to do so, if you also send traffic their way. In the online world, the analog to this activity is “link building”. Your service is so  interesting that others feel that a link to your site would enhance the experience of their customers. So they link to you. A link is just like a “referral”, and they may also reciprocate links for mutual benefits. The link text has to be crafted so that it's targeted, and therefore the appropriate link text must be used. Again, if you are selling shawls, don't  advertise “fresh produce” or even “clothes”.

Now do you want a referral from the “auto parts shop”? Maybe the guy working on maintaining his muscle car wants to buy that “cashmere shawl”. Maybe he doesn't. Sometimes, it's a judgment call. The army-navy discount store shopper may well think it's chic to wear fancy silk scarves. If you had to prioritize, you would want to go for a relevant (ideally complimentary) partner, who is not a direct competitor.

Our friend, the search engine “bot” is very interested in how many links point to your website. From the bot's point of view, each website linking to your site is a “vote” that your site is worth linking to, and hence is “credible”. It's really quite clever, because the bot can't really tell which website is for real (or as the old New Yorker cartoon goes “On the Internet, Nobody Knows you're a Dog”). Therefore, the bots assume that the people reciprocating the links do some due diligence, and link to credible sites. One example of this link voting mechanism was proposed by two Stanford Graduate students. Since their paper achieved a modest level of acclaim, I will link to it . You can read all about the gory details of Page Rank calculations in it (here's our summarized information on the general principles of Page Rank ). In practice, getting targeted links is very beneficial, and this activity is called “Offsite Optimization”.

Now let's assume that your visitor is enticed into coming into your “offline” shawl and scarf store. Are you going to show him your wares and convince him that the cashmere scarf is just what he needs for that anniversary present? Well, you need to make your pitch to the online customer as well, and convince him to go to the relevant section of the site (instead of just standing in the doorway, or the “home page”). You therefore need to link your site appropriately. Perhaps, by having them browse the items all the way to “check out”. The bot also looks at your links and calculates the importance of each page by counting how many other “important” pages from within the site point to it. This process is called “On-Site Optimization”.

Here's a survey for you. If your visitor enters your store, will you dazzle her with neon lights and disco music? A shocking number of websites waste bandwidth, confuse the customer with pointless flash and music (although the use of such multimedia aids is justified when the point is to give the customer a demo). Again, you need to do something like A|B or split testing where half your traffic sees the site without the flash, and the other half sees it with the flash. You have web metrics running, so you can tell if the group of visitors who saw the flash converted better (or did they get alienated). The bot, by the way, ignores images and sound, one way or another.

A smart use of website measurement tools would allow you to attract paying customers, guide them properly through the site, so they don't walk out on you as well as to retain them. This whole process of optimizing your site is called “Organic Search Engine Optimization” or SEO. So does this SEO thing really work? Given our background in website analytics, we have found that Organic Search Engine Optimization is often very cost effective. PPC is popular because it gratifies and it's simple to understand (kind of why sushi is popular). PPC has also made the Google folks, billionaire geek rock stars all dressed in black. It is enormously profitable(for _them_), and it is _your_ money. With Organic SEO, there is no instant gratification of PPC, and there is hard work involved. But your efforts are rewarded over the long term.

Here are some actionable items, a check list if you will that you can use to take care of some obvious improvements.

1>   Careful about the “flash intro” on your homepage:

The bot doesn't like it. Not one bit. It came to your site looking to get an inventory of items and instead it saw this flashy commercial before he could enter your store. There's nothing much to index on your home page. What a waste. See you later. Not! Also, do your eventual human visitors really need that introductory flash, or is it an ego trip for your website designer. You decide. But, we recommend that you split test your idea to avoid guesswork.

2>   Do you need music on your site?

Don't try to hypnotize the user with soothing music to get them into the mood for buying. Most likely it ain't gonna work. Your customer is probably this harried, overworked lady trying to sneak some office time to buy the shawl for her mother in law. Does she suddenly want the speakers on her office PC to start blaring  away? The bot is deaf, by the way and will ignore it.

3>   Describe your items prominently:

If you want the bot to see your stuff, you should make it accessible. This is achieved with good, detailed description. Ultimately what's good for the bot is also good for the human visitor buying your wares.

4>   Careful with Dynamic Pages:

Bot's often have issues with dynamic pages (you know the one's ending with ?&prod=1321&cat=12&col=17). Bot will likely refuse to index it. Next. Try to create static pages to make bot's life simple. It's not practical every time (please write to us if you need further pointers on this).

5>   Reduce ClUTtEr:

Don't make your site Aunt Edna's attic. It's nice to be creative for your human, cash carrying visitors, but (almost) everyone appreciates clean, consistent décor. Use of CSS (cascading style sheets) along with <H> tags is a great idea for instance. Works well for humans as well as bots!

6>   Keep the important content at the top:

The bot usually has a limit of 100 k or so that it reads and interprets. The rest is ignored.

7>   Use Metatags Judiciously:

These (title, description, keyword tags) are the big product signs on the aisles of your stores. It's an opportunity to tell your bot (and human visitors) about your store. It would be a pity to waste it on trivia. Try to convey the most important and relevant information.

8>   Don't hang out with the bums:

As far as the bot is concerned, the “Free For All”(FFA) pages are pond scum. If you  link to them, the bot will lose some respect for your site. However, a FFA page linking to your site won't hurt you. It's like a bum living under the bridge telling the bot that he shops at Nordstrom's.   Really!

9>   Don't pull a fast one:

I recently was explaining what SEO's do to someone, and she told me: “Oh, it's easy…you just use the word that you want to rank on a thousand times on your page”. I wish! There was a time when the bot's were simpletons, but now they recognize it and in their R2D2 like voice complain of “keyword spamming”. Five point deduction, take that Mr. Spammer. Similarly, domain name spamming and sub domain spamming are caught. Another one is “cloaking” where people try to show bot different items than the human visitors. The bot's are often times recognizable, and it's technically possible. However, the bot's sometimes use anticloaking audits, just to check up on you. And they really throw a hissy fit if they catch you. These “black hat” tactics are a very bad idea.

10> Measure, measure, measure:

It's very important to keep track of which keywords perform for you. Sometimes you can get ideas from PPC search words which result in sales. If there is a pattern, and your site is not optimized for those words, time to optimize a landing page for those words. Why pay for PPC, when you  can often optimize some niche words.