How to decrease Bounce Rate

So how do you reduce Bounce Rate, or the rate at which a person only visits one page of a web site and then jumps away to another site. Seems like I've been looking at this awhile now, with some page updates devoted to reducing Bounce Rate. I even see a blog posting from last May concerning Bounce Rate.

Yet I don't see any improvement. However there has been a slight improvement in a few sections of the web site;
Dictionary of Resistor Terms.
Component Derating.
But no changes in other sections.

Of course it's a bit hard to track individual pages, because there are so many. I could improve one page yet the bounce rate on another page decreases for some random reason.
Bounce Rate is important, but only because I see a steady decrease in page views, but no decrease in site visits. Which means the same amount of people visits the site, but over the last few years click to less pages per visit.

New Pages and the result [Title, Date added, Bounce Rate]
Diode Terms, 11/3/07, Bounce rate GOOD [80% last year to 60%]
Minimum Annular Ring Definition, 11/4/07, Poor, [75% steady]
Dictionary of Capacitor Terms, 11/17/07, Average, [60% steady]
Multi-Colored LEDs, 11/20/07, Average [60% falling]
CEC Interface, 11/29/07, Poor [80% rising]
MicroUSB, 11/28/07, Poor, [80% rising]

More Page Views equals a low Bounce Rate. Web Site; Personal Computer Signal Assignments.

Number of Daily Visits

Looks like there is an increase in Daily Visits to the site. January had on average about 8,000 total visits per day, with a few days having around 8,300 visits [as reported by AWSTATS].

This week the visits have grown to around 8,500 visits a-day, or up about 500 per day.

There is no way to tell why the site sees an increase, but it could be due to what ever work was added to the site back in December. Of course there is also no way to tell how long the increase will last.

The important points in the graphic are Number of Visits and Pages. Web Site; PC Bus Pinouts.

Hits for the 7th are no yet in the report. On average from Jan the highest hits come in on Tuesday and Wednesday. Monday is a close third after Tuesday. Visits start to fall off on Thursday, Friday and Saturday. Slowing rebounding on Sunday, I assume from over-seas.

Total Unique Visits

I haven't posted this chart in a while. The chart to the left contains the same data for Unique Visitors as the last posting. The difference here is that the data goes back to 2001, and the chart does not contain the other traffic information. Click the chart for a larger view.

It may appear that there are large fluctuations in the traffic over the last few years, but the graph is misleading. The 'Interface Bus' site receives around 8,000 unique visits a day so a short month or a holiday will reduce those peaks by several thousand hits. A holiday could reduce the number of hits from 8,000 to 2,000, while a short month could lower the graph by 16,000 visits. Most of fluctuations are down 10 to 15,000.

In 2006 the server went down for a few days and the site lost more than 20,000 visitors. In 2005 the counter was changed to one that appeared more strict in counting. But a few months later the graph still showed more visitors regardless.

So as far as I'm concerned unique visitors to the web site have been flat from 2006 on. Stable is fine, but I would rather see some increase in site visits. I still work the site, but it seems the best I can do is keep things stable.

One common reason for a lack of any increase is interface bus obsolescence. Pages written years ago see less and less visitors as the topic moves to obsolescence. For Example the page on the PCI interface bus has fallen from 1,000 hits a week in mid 2006 to about 500 hits last week. The page on the AGP bus has seen the same reduction over the same time period. Because of PCIe both those computer interfaces were already on their way out by 2006.

The chart covers January 2001 to January 2009. I do have data back to 1999, but it would not help the graph and just make the graphic larger than needed.

Internet visits for the month of Jan

Web site visits for January came in at about what I expected. I still can't get the site's visits to increase, it may not look like it but these number are really flat.

The site receives around 8,000 visits a day, so on months with a holiday or with only 28 days the site hits could reduce by over 20,000 just because it's a short month.

Server Bandwidth:
The lowest curve is server bandwidth and does not relate to the other numbers on the chart. The bandwidth is hovering around 100,000 [on the chart] but really equates to 10GB as the numbers were changed to fit the graph.

Unique Visits:
Are visits from a computer within a month, but any one computer is only counted one time. If any one computer returns for a second visit it's counted by the Visits curve.

Visits:
A site visit is registered each time a person visits the site within a month and each time the person returns to the site. Site Visits should always be equal to or greater than Unique Visits.

Page Views:
Are the number of pages a person views per month, regardless of how many times the visitor returns to the web site. Page Views should always be equal to or greater than Site Visits. Page views are really the only data point that is falling. Page Views is related to Bounce Rate, which is the percentage a person visits one page and then leaves the site.

Bounce Rate:
The Bounce Rate for interfacebus is 71% for January. Or 71% of the people that visited the site viewed only one page during the month. Some individual pages have Bounce Rates of 20% while some may have rates as high as 98%. One page I just looked at [30-Pin SIMM] had a 77% Bounce Rate. When I looked at the data the Bounce Rate would cycle from 100% to 50% or 0%. So I went ahead and added links to simular memory modules; 72-Pin SIMM. Not really sure why I cared those 30 pin SIMMs have been obsolete for a decade now. But I know why the links were missing, key-word contamination, confussing the search engine as to what page the 72-pin SIMM was really on.

Google Knol Awards

Seems two of the Knols I wrote have won some awards. The two Knols both won a top viewed award, and a Top Pick Knol award. I assume somebody at Google made the notation for the "award", because they don't indicate any thing other than a small medal. Or this could just be like a smiley face kind of award. If it makes the knols show up sooner in a Knol search, that would be good.

I really would like to have the knols bring in more traffic to my web site. I think I posted the first Knol back in August 2008, now I have nine Knols. The two Knols that have the traffic banner have seen thousands of page views, but have only resulted in a few hundred visits to interfacebus.com.

Incoming traffic from Google Knol is sitting at 0.03% tracked over the last several months. But the search engine brings in the most when tracking hits due to Google.

Of course writing a web page and hoping for page views or traffic is more luck than another thing else. I added a few pages on FET derating over the last month, but they are only generating a few page views a day. In fact the page views are so low for that section they could just be from me checking the page.

Page hits and visits are up for January over December and may end up higher than either October and November.

Internet Visits from Africa

Here is a graphic showing web visits from the continent of Africa. The five sub-continent regions are shown indicating the amount of incoming traffic.

Northern Africa = 10,546 visits
Eastern Africa = 4,599 visits
Middle Africa = 269 visits
Western Africa = 3,241 visits
Southern Africa = 11,901 visits

In this map Southern Africa is made up of 3 countries; Namibia, Botswana, and South Africa. As you might guess South Africa produced all the visits with 11,379 hits. What's interesting is that when I zoom into South Africa to see the visits from the cities places like Cape Town don't place high. Johannesburg produced the largest visits followed by Pretoria and Auckland Park. I had to look up Auckland Park...

Eastern Africa is made up of 18 different countries, with Kenya producing the most visits; 1,460

Middle Africa only provided 269 visits from 8 different countries for an entire year. Cameroon had the most visits at 138.

Western Africa is divided into 16 different countries. Nigeria brought in the most traffic with 1,908 visits.

By far Egypt provided the most traffic from Northern Africa at 6,471.

Internet Visits from Asia

Normally I'll post graphics showing internet visits from countries or states within the US, this time it's the continent of Asia.
Western Asia visits = 54,370 hits
Southern Asia visits = 155,262 hits
Central Asia visits = 568 hits
Eastern Asia visits = 127,655 hits
South-Eastern Asia = 98,395 hits

Because of all the traffic from India, Southern Asia shows the largest amount of traffic.

While at the same time frame Central Asia has only a few incoming visits. Central Asia is made up of five break-away provinces from the USSR [in the 90's ?]. Guess that's why they were permitted to break-away, they had no internet or technology.

Israel provided the most traffic from Western Asia, just slightly above the traffic from Turkey.

China which was closely followed by South Korea sent in the most traffic from Eastern Asia.

The Philippines in South-Eastern Asia sent in the most traffic from that location.

Generating traffic from any of these countries is really out of my control. As I could always just want more traffic from the U.S. and spend time trying to generate more interest in this country. It's still interesting to see the location of incoming traffic. Plus, judging from the locations or time-zones, this is all traffic during the middle of the night ~ which is great.

Resistor Manufacturers

As part of interfacebus, I've had a page covering Manufacturers of Resistors on the web for about ten years now. The page covering Resistors is well established after 10 years on the internet

The Resistor vendors page grows in size or reduces as links are added or removed. The other day four 'redundant' manufacturers were removed from the listing as they all pointed to the company that acquired them. Normally I'll leave a reference to a company that was purchased so people can still find the new company, but after awhile the old link gets removed.

From the "What's New blog" I see a new page dealing with Current Sense Resistors was added in April 2008. Adding a related page to resistors means removing all references to Current Sense Resistors from the main Resistor listing, so that they only show up on the new page, other wise both pages would appear to cover the same topic. Of course removing any information relating to current sense resistors reduces the amount of text from on the main resistor page.

April 2008 also saw a number of new pages that related to Resistors but did not require information being drained off the main resistor page. The new pages provided more incoming links [external pages linking to] the Resistor Manufacturers page. Incoming links should help a page gain visits by providing additional ways to find a page topic. How ever at the same time similar topics on the same web site could end up dividing the traffic between all related topics.

The point is I see a reduction in visits to the Resistor Manufacturers page over the last few years. Page views in 2006 seemed flat, but appeared to start decreasing from 2007 on. Page visits look stable from July 2008 on but are down 60 percent.

Googlebot Crawl Stats

Here is the latest graph on Googlebot crawl stats, or how often the Googlebot spider checks interfacebus.com. For reference, the previous Googlebot crawl rate.

The previous 90 days ending in October showed an average of 452 pages spidered each day, while this chart shows 523 pages a day.

The amount of data down-loaded has increased to over 6,405k bytes [6Meg], but the time spent downloading the information has decreased by 100mS. So the spider is reading more data faster, which is good.
July crawl rate.
Sep 07 crawl rate.
Jan to July o7 crawl rate.

2008 Visits from Countries

Attached is the Google Analytics map for the world showing hits from each country. Google uses a color code, I added the text.

United States; 907,029 visits
United Kingdom; 126,399 visits
India; 121,684
Canada; 98,013
Germany; 64,810

Highest hits during last year

Took a few data points from last years visits.

Language; 57,297 visits; German [de]
Browser; 1,352,562 visits; Internet Explorer [30.01%]
OS; 1,680,333 visits; Windows XP [84.16%]
Java Support; 2,052,254 visits [96.13%]
Traffic Sources, Referring Site; 19,938 visits [en.wikipedia.org]
Traffic Sources, Search Engine; 1,681,379 visits [google]

India had the highest non-US visits [U.S.A. had the most hits].

Total from all sites in 2008 include:
1,982,780 Visits
3,717,226 Pageviews

Remember non-Java supported Browsers are not counted.

External Web Linking

Here is a listing of web page addresses with only one external page link from another web site. In most cases the external link is really coming from the web site's sitemap, which is located on another server.

The list only indicates a total of 1,032 pages, while the main sitemaps page indicates 1,414 URLs.

How to Derate a Connectors.

Limiting the maximum current used by connector pins.

How to Derate Power Supplies.

How to Derate Switches.

Manufacturers of DC-to-DC Converter ICs.

A list of Voltage Converter IC vendors.

Joystick for the Apple IIc computer.

An out dated interface on the Apple Computer.

PC104 Board Dimensions.

Graphic of PC-104 card.

Manufacturers of Embedded Backplanes.

Industrial or non-commercial Back-Planes.

Listing of Single Board Computer Form Factors.

Sizes of SBC formats.

Manufacturers of Cable Clamps.

Companies that produce cable clamps.

Strain Relief Manufacturers.

Definition of Gray Code.

Hex Symbol Table.

Some of these pages should have showed up in the 'what's new blog' in addition to the sitemap, but must be missing. Having external web links help increase the page ranking of the page being linked to. But the page linking in needs to have a page rank of its own.

Web Statistics

Web analytics indicate an average year The server side counter AWSTATs indicates 2,326,184 visits while Google Analytics shows 2,131,879 visits with 3,987,364 Pageviews.

2007 had 2,354,985 web visits and 4,063,901 page impressions

2006 had 1,550,092 web visits and 3,688,317 page impressions

It appears that 2007 had a bit more traffic than 2008

Sitemap Generator

I up-loaded a new XML sitemap to Google [Google, Webmaster Tools, Sitemaps]. The current XML sitemap contains 1,859 submitted URLs, as reported by Google. The previous XML sitemap was last up-loaded on August 16th and contained 1,727 URLs. So 132 new pages were added to the web site over the last 4 months. The last site map caused 1,318 URLs to be indexed by Google.

During this same time frame 968 html files were changed or up-dated in some fashion. However the up-dates could be just html fixes in some cases.

The program GSiteCrawler is used to spider the site and generate an XML Sitemap. Because of the bandwidth required I only generate a sitemap once every few months. The program has to check every page on the web site. But once the site map is generated and up-loaded to the server the bandwidth usage remains the same. Google reads the sitemap once every few days regardless of how old it is. So if Google is going to download the sitemap from the server it may as well be up to date.

Other search engines find the site map via a comment in the robots.txt file on the server. That comment happens to point to another web site which also holds the sitemap.xml file. So only Google downloads the sitmap from my server, all other search engines get the same file from another server [saving me bandwidth].

I also started to update the HTML version of the sitemap, the human readable version. I added 18 new page addresses to the sitemap that had been added to the site from 8/16 to 10/16. I up-loaded what I had and will add the previous two months as time permits.

A few days ago I spidered the site using Xenu to check for broken web links. Xenu indicates 2% of the 6,105 URLs on the site were bad. However giving them a few days to come back proved that only about a dozen URL links were really gone. I did end up deleting or removing a few links.

Adding Pages to Increase Visits

I keep track of new page additions to the web site via another blog. So far this year about 190 new pages have been added to interfacebus.com, a few other new additions to the other web sites. However many of the new pages are small in size which makes it hard to compete with similar pages with more content on the internet. Of course even if the new pages do not show well in the search engine rankings [SERPS] they still serve to support the web site.

What Page Rank a page receives is determined by the search engines, as is when a page receives a Page Rank. Page Rank and Search Engine position are both determined by the search engines, with Google bringing in the vast majority of incoming visitors.

Of course some pages just might not do well because they have little or no content;
Manufacturers of Solderless Terminals,
Manufacturers of Water Alarm Units,
Synchro definition,
Mazda Engine Control System, [New page]
Residual Current Device definition,
Manufacturers of Thermal Adhesive Compounds, [No page linking in]
Industrial Mezzanine Board Formats, [No page linking in]
Industrial Tool Manufacturers,

So those are a few of the pages that have had the fewest visits this year. In two cases it appears like there were no other pages that pointed to them, which would explain why the page received no hits.

It's a bit hard to find the low performing pages because the list is full of misspelled page addresses, which return a 404 page not found code. The list of 'Top' Content' is generated by Google Analytics and contains 18,718 line items for the year. However I only checked the top 2,500 as all the rest of the list is just misspelled page addresses. The Mazda page is 1,670 line items from the top of the list [surrounded by misspelled web addresses].

Adding new pages does not guaranty more traffic

Web History Stats

Went ahead and started to compare web stats for this year and last. So far they look about the same. Of course I don't have any data for December yet, but I can estamate that. Last year. December saw 126,000 unique visitors. So I should see around 1,650,000 visitors this year, or about the same amount as last year.

Google PageRank Reduced

I noticed my Google PageRank reduced from a 5 to a 4 on interfacebus.com. The site has had a PageRank of 5 for years now, although I have seen a reduction in the past. Using Google Webmaster Tools I can see which page on the site has the highest PageRank. Back in August a page other the the index page had the highest PageRank, so it does go up and down.

It could be that the Home page was just a 5.0 instead of say a 5.8 so maybe it's common for the index to slide between a four and 5, but I don't see a PageRank of 4 that often.

Why does the PageRank change; well it's recalculated each month. Some page that link to you may be off-line when the GoogleBot checks them making appear that fewer external pages are being linked in at any given time.

MS Web Site Issues

OUTDATED TEXT REMOVED, 6-28-2010

Same thing happened with the browser that Google came out with; Chrome. I could not access the site data using that browser either ~ There is a previous post about that issue, I think. So now, Firefox now longer works. What, it's an MS site, so only Internet Explorer will work on a Microsoft Soft web site.........

Any way I posted a graphic which shows hits to serialphy.com, which is another site I operate. The two main items are the Orange line which shows unique visits, and the yellow line which depicts number of visits.

I see an increase in visitors, so I'm fine with that. Notice I do not provide any hard numbers ~ there to low...... Oh, November is low because it's still mid-month, so that data will not be available for another 15 days.

Spider Crawl Stats

I just checked the crawl stats for interfacebus.com using Google's Webmaster Tools. The attached graphic shows Googlebot's [Google's spider] activity over the last 90 days.

The first two graphs appear normal, but the last graph does not. Check out the reduction in time to spider a page even as the amount of data downloaded has increased.

Does this mean that my server has gotten faster over the last few months? Or maybe Google is coming out at 3am when there are less people using the server?

What's new blog

Looks like the companion blog used to indicate new pages to interfacebus.com is a year old now. The blog just provides a link to new pages as they are added to the site. There were 35 entries last year and another 73 additions this year for a total of 108 separate blog pages. Many of these pages contain several 'new' links to pages that were just added to the site, so a few hundred pages have been added over the last year.