Web Site SEO Analysis Request Your Free Analysis
|
Home Business OnlineLocation: Case Studies 4 A Case For Regressiom--> Home Business OnlineCase Study 4: A Case For Regression Pauline has been running Internet businesses since 1995 and during those years she has observed many changes. As new technology emerged she had to make changes to her e-businesses and for many years the technology changes came faster than any webmaster could keep up with. In 1995 most marketing techniques were largely offline promotions and campaigns. The only online advertising was through bulletin boards. By 1998 the marketing had completely changed and rechanged a half dozen or more times. Linking to other sites and those sites linking back lead to link lists and webrings. Link lists and webrings lead to toplist exchanges which then made way for banner exchanges. Yahoo gave directory a whole new meaning from the phone book directory to a worldwide database of sites and businesses. Directories like Yahoo made room for search engines with searchable databases for the directories to use. The directories and search engines grew in number and the competition began to heat up. Meanwhile webmasters became aware of unscrupulous techniques that were at play. Link exchanges, banner exchanges, and toplists were being cheated and in some cases the list owner was cheating the list. People were writing programs to emulate traffic to the lists and thereby making their banner or link rise to the top of the list and take advantage of the majority of real traffic. Soon programs known as bots were being used to crash websites, steal mailing lists, and other malicious activities. By 2001 the Internet was a piracy and scam artists were driving the consumers away and back to the malls. Creditcard fraud was the leader in causing the markets to shrink and online businesses to fail. It wasn’t uncommon for websites to collect customer information over unsecured web pages and a bot watching and copying the customer data. A merchant would never be aware that they were setting the customer up for a major creditcard swindle. Many search engines and directories were failing too. The markets were all but dead and those expensive bandwidth-devouring databases were very expensive to maintain. Pay per click advertising came to the scene as a means for search engines and directories to make money. Merchants would pay for a listing and then pay for each time the search engine or directory sent a customer to the merchant’s site. Many webmasters knew it was another scam and just ignored the pay per click resources. Yahoo began to charge for businesses to be listed in the directory. Though they grandfathered existing sites, the writing was on the wall that free advertising was about to be a thing of the past. Merchant accounts began to take charge of Internet purchases and the providers developed secure ordering forms. By the end of 2001 creditcard fraud was all but gone. Secure websites continued to improve and having a merchant account for Internet businesses became more and more expensive. Purchasing a shopping cart gateway was also increasing in costs and gave way to another burst of Internet businesses and service providers. Search engines began to follow the Google example and Yahoo began to feel its first real threat. AOL, the most popular Internet Service Provider was also being copied as more chat rooms, instant messages, and webcams became the standards. All the while the merchant was swimming in a whirlpool of marketing changes in all directions. Adjusting web pages and styles to keep up with all of the changes. Images in bitmap form gave way to jpeg and gif images added motion to an image. Slow webpages were the result of too many or too large of an image. Design had to be clasic and a bit boring. As the year 2004 approached the business owners saw the first year of few and far between changes to the markets. During 2003 the purpose of web designs became directly related to how search engines measured a site for listing in the database/rank. Search engine optimization (SEO) became more of a mathematical equation than a chance and guessing game. Marketing by means of directory and search engine listing became the primary focus for most every Internet business. So much so that weblog-analyzing programs became dedicated to revealing traffic from search engines and directories and the better analyzers even show what search words or phrase the customer used to find your site. By the last quarter of 2003 the Internet gave birth to another boom of web businesses: The SEO specialists. Now with few changes and a stablization period it would seem that all was ready for the Internet business owner. Especially for those who had survived the many years of constant change and the booms and busts. However, there were still a few glitches to the utopian scene on the distant horizon. The glitches were centered on three key elements and two obscure variables. Those pesky pay per clicks were number one, paid advertising was a developing and bizarre twist to directories and search engines and that’s number 2, and SEO matrixes round out the three glitches. The obscure variables are the business owners and/or webmaster’s preoccupation with SEO and resulting keyword placement in major search engines and directories, and the lack of attention to marketing planning principles. Web design went through the various stages of providing information, services, products, and portals to junk sites designed to fit a search engine equation. The quest for link popularity finds many webmasters and business owners drafting more junk sites in effort to drive up their count of cross links. The focus becomes directed to a popularity contest where webmasters with the highest PR and the top position within Google for desired keywords have bragging rights and the admiration of all others. And, somewhere in all of this is the ability to buy your way to the top of the list. Pay per click engines are making money from the new startups and those who just haven’t been able to understand the math of designing junk sites to the equation. The business owner and webmaster watch the web logs with hungry eyes for the telltale sign that the Google search engine crawled their site deeply and completely. Carefully they look through the logs to catch the entry pages and keywords that surfers (once-upon-a-time these were called customers) are using to enter the site. Victory is declared when the Adwords right up was approved and the cost of advertising will bring the Google hits up on the log records. The count of surfers from the Pay Per Click Engines are counted and compared to yesterdays or last weeks totals. Traffic becomes the ultimate goal, Google ranking is evidence of superior web design, PR is scorned if it is lower than 5 and glorified if above 7. Just when the definitions were all coming together Google changed everything. The matrix was redesigned and the myriad of equations was improved, but no one knew what the new math was. Google introduced a virus of sorts that unseated the best PR ranks and robbed the best-placed keywords. Suddenly virtual unknowns had the bragging rights and the coveted PR standings. To mix it up even more Google drops the paid sponsored spots (more to the point the companies who paid $120,000 a year and more for those spots discovered that the return on investment was negative). Pay per click engines begin to force guidelines on site design and content which pushes many top bidders out. The scramble was on again and webmasters and business owners tried new web designs and watched the telltale signs of the web logs for proof they were on the right track or not. Through all of this madness Pauline stayed the course outlined in her marketing plan and business objectives. She never became engrossed in the quest for PR and keyword placement. Instead she based her site design and business changes on bottom line performances. Were sales increasing, was income increasing, is the business growing? These were the questions she had when she observed her web logs. The traffic from Google or the use of Pay Per click and Adwords advertising were each compared to ROI results. The return on investment was always her measure of success for paid advertising the same as it is for buying new technology and expanding by way of new websites. Her business grew steady over the years. Even during the Internet bust years and the crisis of creditcard fraud she managed her business to guide customers to different means of shopping and safer means of payment. Link exchanges were handled as business decisions and the potential of the sites complimenting each other in terms of customer wants and needs. She found link lists, webrings, toplist banner exchanges that were market specific for her products and services. She followed the strict discipline of annual strategic planning, setting objectives and tasks to meet the planning targets. She measured only those variables which directly affected the four targets. All changes to business direction, marketing, advertising (including web site design) were based on measured fact. As her business continued from one year into the next she was realizing a steady gain to gross sales of 25% on average. Her profits were slightly better than gross sales coming in at 27% per year on average. With all the success and concurring the struggles and changes the first 8 years of business on the Internet brought her she still had one haunting question left unanswered. Where were the sales coming from? When I explained the correlation analysis I had done for Debbi and the results, Pauline was intrigued to put it mildly. The more I would explain correlation the more confused Pauline became. It wasn’t that she didn’t understand the concept or the purpose it was that she couldn’t see how each variable by itself would tell the whole story of total sales. When she explained to me that what was really important to her was to understand how each source of Internet customers affected sales as a whole. In a nutshell, she wanted to know what sources of Internet traffic were valuable and which were simply wastes of time. Not in terms of PR or total traffic from the most popular search engines and directories, but in terms of bottom line performance. How could the correlation analysis, which proves valuable for identifying key variables be used to identify how the entire marketing of the business is contributing to sales? “A regression analysis” was my answer. The regression develops a formula that shows how each contributing factor ads to or subtracts from the target measure. The formula would look something like, 2.003xGoogle + 3.07xYahoo – 1.06 x something else, etc. It produces a total picture of all contributing factors. Pauline was very interested and wanted to know the answers. Where were her sales coming from and what are the most valuable sources of customers? We began to set up the experiment. The first step we took was to identify all of the variables. The variables would be sources of traffic and where in the site the traffic was looking i.e., entry pages. With a good website log program you can see where the traffic to your site is coming from and where in the site it is going. We looked at the logs for the previous two months and choose the top 5 sources in each category as a place to start. Traffic Categories: Search Engines, Link Exchanges, Webrings, Top Lists, Bookmarks, Direct Access. With a list of 37 variables to track we then decided on gathering 100 consecutive days of data for each variable. A sample size of 100 data points is generally considered adequate for statistical inference. I also found that grouping the data into subgroups of five days each provided a normalized data set. Three of Pauline’s sites were chosen to be analyzed and cross compared. Her best producing site, her worst producing site and one in the middle were selected. We established some assumptions as we began to collect data. A) If a search engine (SE) traffic was significant then a keyword (KW) should also be significant: A positive correlation should exist. When the analysis was complete I prepared the explanation and graphics for Pauline. The significant result of a regression analysis is the “Prediction Equation” which shows how each variable contributes to the whole. For the middle of the road website the formula looks like this. Daily Sales = - 0.0259102*keyword1 + 0.0184514*keyword2 - 0.0945501*keyword3 - 0.0440144*keyword4 + 0.0314224*keyword5 + 0.00123985*Google + 0.110487*Yahoo + 0.0441073*AOL + 0.152602*AltaVista + 0.00616965*bannerlist1 - 0.00274115*bannerlist2 - 0.00293838*bannerlist 3 - To simplify this equation you can state it as sales per source as follows: Correlations did exist between keyword 5 and AltaVista. That was the only 2 way or three-way correlation. The “Other Sources” category, which produces 48% of identified sales, is spread between word of mouth advertising and other forms of BUZZ and offline advertising. The significance of nearly half of sales coming from other than Internet marketing and advertising held up in all three website analysis results. AltaVista was generally the least provider of traffic numbers and always in the top three for revenue production. AOL and Yahoo were also significant contributors to sales in all three results were as Google consistently proved a poor provider of sales. Once the results were translated and understood Pauline put together action plans based on the analysis. Case Study 1 Case Study 2 Case Study 3 Case Study 4 Case Study 5
Call me with any questions and for help making your business more profitable: 1-888-464-3646, extension 0405115 Your personal small business consultant is ready to work for you right now. |
Any time any place.
|
