Case Study: Six Sigma Methodology for SEO

In the previous article in this three-part series, you learned the details involved in forming a Six Sigma team and were introduced to the various tools and guidelines involved in going through the DMAIC Six Sigma methods. In this last part, you will see everything in action through a complete case study.

We’ll start with the case study’s background. I won’t be showing you the real keywords and website name, to protect owners’ privacy and confidentiality.

A music download website originally aimed to target consumers looking for music that can be included in various commercial and non-commercial projects. They asked for a fee to use the song (at full length), using industry standard set prices; the cost to include the song in any projects depends on the type of project . The website server side script handles the form processing and computation automatically.

Below is the original flow of a transaction before the Six Sigma project was implemented:

Dissatisfied with the website’s performance over the course of one year, the website business manager, following the advice of their president and chief executive officer, called together a team of experts to create a six sigma team project to improve performance. The SEO in charge is the team leader, and the six sigma improvement project was initiated. When they convened for the first time, they talked about commitment and said that all members are ready and know the process. They formulate the schedule below, which they find to be realistic:

The team assumed that the problem was pretty obvious, but it wasn’t. Aside from the fact that the website business owner is reporting no sales, there is also a report that there is no downloading. Defining the problem correctly takes you halfway to a solution. So the technical members of the team hold discussions with the marketing head and website business manager to slowly define the problem.

According to the website business manager and marketing head, their competitors are those royalty-free download websites. They offer a single fee to download the whole package (for the entire artist catalog). While this seems less costly to customers, it has a disadvantage: the artist/writers themselves who are affiliated with performance societies (e.g. BMI), earn no money at all when their work will be publicly performed.

The team finally admits that they can offer all downloads at no cost to consumers, as long as authors get paid for their song’s public performance when songs get cut and released.

The download itself, even though the website’s customers are not paying for it, can make the authors happy because somebody just downloaded their song. The customers are also happy because they avoid paying a lot and get the song they like at no charge. The website business manager is also happy, since the authors/writers that are registered with their site are happy and will likely continue their membership. In the long run, the download can translate to money when the song is performed in public.

Finally the team executes the deliverables in this phase:

Problem statement:

“In the span of three months, all traffic sources sent a total of 1,539 unique visitors. This translates to 513 average unique visitors per month. The conversion rate is 0%. So there is no single download in that three month period. (This was even true for an entire year of operation)”

Project objective:

“The objective is to at least increase the conversion rate to above 1%.” (This is an initial safe target, which the team recommends because it seems to be the lower end of a reasonable website conversion rate).

The data collection plan is straightforward, since all information can be gathered using Google Analytics. The performance baseline data before the implementation are as follows:

a. %Conversion rate: 0%

b. Average monthly unique visitors (search engines, direct traffic and others): 513

Analysis Phase

The team conducts a brainstorming session, writes down all possible causes on the white board and produces the “cause and effect” diagram below:

After the possible root causes are identified by the team, they seek out numerical evidence to confirm the root cause. With the help of the statistician and the SEO expert, they dig out analytics data, interpret and find relationships between the historical information presented.

They discovered that certain causes of the problem can be grouped, since they have similar origins and can be confirmed at once.

The first group of possible causes include:

1. Download fee forces customers to seek free alternative.

2. “Buy now” buttons imply it is not free and fans/customers are looking for free products.

3. Artist fans will not download songs with fee.

4. Too many steps before successful downloading (8-stage process; see first screenshot in this article)

Confirmation result: The above causes are confirmed to be true because the %bounce rate of the home page (where those "buy now" buttons first appear) as well as the check out pages (where a form asks for a user’s name, address, etc as well as the downloading fee) averages around 70%. This is too high; it means visitors tend to navigate away quickly and are not interested in digging further (especially if it requires them to go through an 8-stage process before successful downloading). Knowing it is not free immediately convinces them to leave the website and visit competitor websites.

5. Music consumers needs the full length to properly decide on a song.

Confirmation result: The separate check out page for sample downloading (less than one minute) has a conversion rate of 0% while the %bounce rate is very low, unlike that for the payment pages(<10%). This means potential customers are attracted because it is free and convinces them to click further – however, on the form’s page, it is stated that it is just a sample and not a complete version, so they navigate away from that page, failing to be converted.

5. No search engine optimization has been done.

Confirmation result:

The website is NOT ranking in the Top 10 for the main targeted key phrases as well as the secondary key phrases. Traffic from search engines is not significant as compared to other traffic sources:

Traffic from search engines = 47.56%

Traffic from other sources/referring sites/direct = 52.44%

Total = 100%

Now that the root causes are confirmed and found to be true, the team formulates the list of corrective actions to solve the problem. The corrective actions, naturally, are based on the root cause analysis.

1. Offer full length song downloading for free. Remove any sample downloading and its associated forms.

2. Remove the “buy now” button and replace with “Click here for free downloads.”

3. Remove all pages or portions of website content indicating “downloading fees.”

4. Shorten the check out to a 3-stage funnel process (Shop -> Standard agreement -> Download). No filling out of forms required for fans and customers.

5. Do some SEO onsite work strengthening the word “Free.” This includes revising title tags, internal links, affected content, etc. An offsite campaign was also done to get more references from trusted, related websites to the client website in relation to their updated content. Featured content to get quick attention/buzz is also considered by the marketing head, by putting content in front of thousands of viewers daily in other related sites to get immediate results, such as “link baiting” or traffic from Google-related services such as News.

Control Phase

A month after implementation, the website reaps a lot of benefits from the improvement actions:

a. There were 560 unique visitors and 16 successful downloads. This makes the conversion rate increases from a 0% 3-month average before implementation to 2.86% in just one month of observation. This is above the objective set by the six sigma team, which was just 1% as stated in the problem definition phase.

b. Search engines contribute 60% of the traffic, while other sources contribute 40%. This means the search engine optimization improvement is effective in bringing in more traffic from search engines.

The above are the most important benefits produced by the Six Sigma project. It is expected to pick up even more as continuous improvements are periodically reviewed by the team on a monthly basis.

The above data is based on Google Analytics reports. Permanent control measures are being implemented in the site by the web developer to ensure that the corrective actions implemented are not changed over the course of time.

Google+ Comments

Google+ Comments