Monitoring your Web Stats - How do you use them?

[greeneagle]

Those of you who minitor your and client web stats know what a valuable tool they are.

Hosting Companies that do not offer Statistical packages are a dieing breed.

NO ONE can effectively manage a Site without stats.

Some of the Sites we manage have Webalizer. Webalizer is better than nothing, but pales in comparison to AWS (Advanced Web Statistics).

Obviously there are many great and important features that can be utilized to show a webmaster or Site owner:

The effectiveness of any given marketing campaign.
Overall visitor preformance.
What needs improvement and what doesn't.

Web stats are the most effective tool we have, in that it is the nearest thing we have to "real time" performance data and visitor satisfaction.

One of the AWS stats I pay particular attention to is the "Save as a Favorite" reporting.

Our site is running a phenomenal 19% this month.

I believe that the higher this stat is, it reflects better overall visitor satisfaction, not to mention the fact there is at least residual indication of future traffic indicated.

Stats are important and are a lifeline into Site management. Without them Site management is nothing but..."Shooting in the dark".

I would like to discuss Stat packages and the ways we use them to help promote ourselves and clients in this thread.

I don't think there is a better optimization tool than Statistical reporting and the the developing analysis and action expresses thereby.

Everyone seems to develop their own technique.

Ken

[dawilliams]

I couldn't agree more with greeneagle.

Web stats have proven invaluable for both the client and developer providing quantifiable data that clearly illustrates what 'works' and 'what doesn't work' while marketing on the internet.

Those of our clients whom have chosen to plan future web site marketing actions based on this information have been truly rewarded for their attention to detail.

How better to determine the 'stickiness' of one's content than to measure it's success with unimpeachable data.

Affordable Web Design | Professional Web Development
www.DavidWilliams.Biz

[brian.mark]

We sell things, so our #1 metric is conversion rate. We want more visitors, more bookmarks, etc. - but our true measure is how many of the visitors we got (which we want to be more) acually bought something. That tells us if we're targeting the right terms or not.

Brian.

[wilderness]

I agree with you Greeneagle, regarding stats. Sites that are with my host have good information but I have actually set some of them up with Statcounter so that my clients can go look at their stats and have encouraged them to do so since there is no way that I can determine whether visitors to their sites have actually booked with them. Unfortunately, most of them don't. I have two questions for you. My host has excellent and thorough stats, but they don't match with Statcounter's. Do you know why this would be so?
Second question is related and will probably show my ignorance about SEO even though I try to read every newsletter I can get my hands on. One such newsletter suggested checking traffic ranking with Alexa. I did so today and the numbers just don't make sense. If I could direct you to http://www.alexa.com/data/details/re....resortsbc.com
you will see that resortsbc shows Traffic Rank for resortsbc.com: 4,887,136
The two sites listed below as related are far older and more established than mine, which is a fairly new site, yet their Traffic Rank numbers are lower. Do you know why? Does this mean that none of these numbers can really be trusted? Thanks ahead for your answer.
J Baker

[brian.mark]

First off, you're comparing apples to oranges by comparing statcounter and a logfile analyzer. One is JavaScript - run by browsers with JS enabled. Spiders, visitors who decide to turn it off, and some old browsers won't ever show up in a JavaScript tracker. On the other hand, logfile analysis only shows requests, so people hitting the back button show as another view in a JS solution and don't show the new view on a logfile stat package. JS and logfile won't ever be the same, no matter how much they work to get similar results.

Second question was about Alexa. Well, that only shows visits by people with the Alexa toolbar installed. I personally don't have it installed. Nobody that works here does (here or at home). That means that our visits aren't counted by their system. However, when we see visitors on our site from Amazon, they always have it installed. The fact that their employees all have it says that their numbers are naturally going to be higher.

It's a measure, but not the best. Most certainly something you can look at, but realize that it doesn't show you any facts - just some sort of a number as a guess. You could also look at the numbers from trustgauge, but that has the same problems.

Brian.

[wilderness]

Thank you Brian,
It never occurred to me about the javascript being the difference on statcounter, but it makes sense.
I don't have the Alexa toolbar installed either and won't but it doesn't quite answer my question as to why the numbers on a new site such as mine would be so much higher than on the two sites listed as being related. I don't understand how their numbers can be so far out of line and this would indicate to me that their information in other areas regarding SEO may not be accurate, however, I could be wrong. But thank you again for the information. I much appreciate it. I've learned a tremendous amount over the past few months reading all you guys' posts and it's been a great help to me.

J

[ronparrs]

We've been using SuperStats from Network solutions for about 2 years. It's been great even at the basic level especially when providing me information such as keywords and referrals. At the higher levels it's probably worthwhile if I didn't want a home life!

SuperStats is probably the single best tool in growing the web side of our business.

[Steve W]

Hi, how are you detecting the 'add to favourites' count please? Is that through a link on the site or is there some magic java way to do it?

Thanks :)

[Songwriters7]

On the subject of webstats, can someone please explain how it's possible for pages to have more visitors Exit than Enter? Seems illogical, but that's what our site stats say about some of our pages.

Thanks!

Songwriters7

[intelliot]

J Baker (wilderness),

"Rank" means that lower numbers are better, higher numbers are worse. According to your post, you're rank 4,887,136. The most popular site would be rank 1, then 2, then 3, etc. So it makes the most sense for more established sites to have a lower Traffic Rank number.

As your site gets more traffic, its Rank will approach 1st place (decrease).

The number cannot be trusted completely because there are limitations. However, it is a good estimate.

[brian.mark]

Originally Posted by Steve W

Hi, how are you detecting the 'add to favourites' count please? Is that through a link on the site or is there some magic java way to do it?

Thanks :)

That's a feature in awstats. It just counts how many times favicon.ico is accessed. This is really skewed now with FireFox using those icons for the tabs and in the address bar.

Brian.

[brian.mark]

Originally Posted by Songwriters7

On the subject of webstats, can someone please explain how it's possible for pages to have more visitors Exit than Enter? Seems illogical, but that's what our site stats say about some of our pages.

Thanks!

Songwriters7

Usually, what is reported as Entry is when someone enters the site through that page. What is reported as exit is when that's the last page viewed by that visitor.

It's very easy for a page to not be a landing page too often but be the last page that someone views.

Brian.

[Steve W]

Thanks for enlightening me Brian. I guess you have to consider % browser id=Gecko when tallying up hits and make that your error margin?

[Tubby]

Quote
'One of the AWS stats I pay particular attention to is the "Save as a Favorite" reporting.'

YES! This one is the King. .

I have not yet found a stats program that I like. But I do run stats programs on sections I am working on, outward bound exit links I like. Keyword information I like.

These latest 50 are from as new section today
they seem to say - 'Who are you aiming at ?'
You have to love stats, Its prime feedback..

- classic australian cars
- 1958 Oldsmobile Super 88 Holiday Coupes
- michigan 1962 pontiac windshield
- "1957 cadillac" parts
- LOUISIANA CLASSICS
- (562) 762-8244
- 1971 Chevy. Impala parts front grill front lights
- cowl hood for 1979 malibu
- @us.cibc.com
- "1966 oldsmobile Tornado"
- chevelle for sale ga 1965 chevy
- 4 door willys wagon
- 1957 chevy parts in iowa
- impala 1963 to sell in louisiana
- custom mirrows for pickups
- 1948 ford f 100 truck parts
- 1973 volkswagon super beetle ragtop
- torana for sale
- PARTS FOR A 1979 OLDSMOBILE CUTLAS SUPREME
- 1950 pontiac chieftain convertible interior pictures
- dodge rampage
- 1981 mrecedes 500sl buy
- door 1978 trans am patch panel
- 1967 Datsun 1300 Pickup Truck
- 1958 pontiac starchief sale
- VW BUGS FOR SALE IN OREGON
- 1963 CHRYSLER IMPERIAL FOR SALE ALABAMA
- John Early Whitehall, PA 18052
- 69 cutlass sheet metal
- Triumph Herald sedan
- classic steering wheels australia
- 1967 chevrolet eye brows
- "1955 ford f-100"
- kannapolis nc used cars
- 64 ford fairlane bumbers
- 1963 ford nos parts wanted
- 94 oldsmobile cutlass cedar s
- 1957 F-100
- 1967 olds delta88 355
- 1979 cutlas bumper
- usa ebay nova console

Stats like this say to me oops, no prominant single words double word phrases yet..

[wilderness]

To Intelliot,
Thanks for the information on Alexis Ranking. I had no idea that is how it works but it definitely answered my question. Unfortunately, finding out I was that far down Alexis' list also ruined my supper so lots of work to do on the site this winter.
Thanks again guys!

J

[KenGalo]

This is all some great insight. Being in the marketing end only of the my biz, I use the stat counters to differentiate between surfers and customers. Since we target market based on geographical locations we can tell within a week or two if a particular marketing scheme is working or not by the traffic, which pages are visited and return visits.
Great info, thanks.

[mike]

I like getting stats from multiple sources and comparing. We use awstats like some of the other folks here and I like it quite a lot. I keep an eye on keywords, referring sites and unique visitors mostly.

[Tubby]

I love looking at other website stats, they can turn what you think is normal upside down, unfortunately you don't get the chance often, especialy cometitors stats, maybe somebody should start a new site 'Stat-swapping.com'.

But who would trust another webmasters stats, most of lie through our teeth when quoting visitor numbers etc.

Has anyone done any research on longer key-phrases in comparison to shorter key-phrases. are long phrase users more likely to buy because they are better targeted?. .

gonna have a look at awstats now,

[brian.mark]

Originally Posted by Tubby

I love looking at other website stats, they can turn what you think is normal upside down, unfortunately you don't get the chance often, especialy cometitors stats, maybe somebody should start a new site 'Stat-swapping.com'.

But who would trust another webmasters stats, most of lie through our teeth when quoting visitor numbers etc.

Has anyone done any research on longer key-phrases in comparison to shorter key-phrases. are long phrase users more likely to buy because they are better targeted?. .

gonna have a look at awstats now,

I've seen those stats at SES a few times. 1 - 2 words are considered "Browsing" terms. 3 - 4 is the "Consideration" phase. 4+ is considered the "Buying" phase. I think it's comscore that puts those out, but I haven't seen that presentation since Chicago last year, which was December, so I don't remember who published that for sure.

Brian.

[Tubby]

Quote
'1 - 2 words are considered "Browsing" terms. 3 - 4 is the "Consideration" phase. 4+ is considered the "Buying" phase.'

Yes that probably what I read. But I have done my fair share of writing and reading, often writers having wondered about something then create some 'facts'. as good reason to write an article. I would far sooner read a post from someone I knew that can state his own observations and where he observed them. There are far to many webmasters creating facts to post in return for traffic. I call it 'Concept spam' I do like to see or know the source. ( I used to write articles part time for womens magazines - does your man realy love you? type of thing where answer (a) was worth three points etc - just for the fun and some spare cash) I do know what crap some writers can come up with,, , just bunkum, but edited more carefully. . .

I would like to see more configurable stats, where a webmaster could punch in the combination he is curious about, something like - Exit links/ original search terms/ length of stay/ sort of thing. and only get stats for these items. Now somebody tell me I can get this.(then tell me it's only $*** per month)

[MarcieZoob]

To provide effective SEO services, you need decent stats available. We combine Webalizer and Awstats for our monthly analysis of client sites. We pay particular attention to Referring SEs, Unique Visitors and the Keyword Phrases so we can make sure we're hitting the target market (or, if not, we can make adjustments).

[Patrick Goff]

I have been analysing my stats every month for the last four years. I have reached the following understanding in relation to them, and would be grateful for confirmation from others that I am understanding them right:

1. People visiting are more important than hits (in the last six months I had 207,000+ visitors and about 600,000 hits, which seems to confirm my stats average of 3 pages read per person)
2. Repeat visitors are more important still, but stats packages only measure a return visitor twice (first visit and then first return, and not after that) - from my 207,000 I had about 60,000 repeat visitors.
3. If a site stat shows that there is no code against indicating where entry to the site was made, then the most likely explanation is that they are coming in response to a hyperlink. I send out on Sunday night a weekly digest of new articles from my site which is all hyperlinks, and which has led to nearly 60,000 hits on a Monday compared to a daliy average of only 25,000
4. The reading of 'whois' data no longer leads to individuals but only to the generic ISP address. For example my site was visited in one month by over 150 individuals from a major hotel chain (my site at www.hoteldesigns.net is a major information resource for those involved in the design and construction of hotels)but 'whois' search led only to the chain not any department or individual - I would like to have known whether it was the purchasing side or the sales side that was taking an interst). I think this is a change over 18 months ago when individuals were more easily identified. Is this interpretation correct?
5. Do you agree it is the general trends that are most important e.g. are the figures steadily climbing? Figures showing what happens if I do this? Showing how the key search words relate to google rankings (mine do)?

I always take the raw logfiles and put them in another package to crosscheck Webalyzer, and the trend attendance figures check out. However my own package shows more pages read than webalyzer - nearly 40% different with a monthly figure from wablyzer of about 69,000 compared with a figure of over 102,000 from my own package. Why should this be from the same raw log files?

Or does this last simply show we should use common sense in looking at our figures? For they are like using the I Ching and will respond to our own spin and that we all have to be well grounded before we start this process.

Happy interpreting, and thanks for all the informative reads the site continues to provide

[khurramali]

Originally Posted by wilderness

Second question is related and will probably show my ignorance about SEO even though I try to read every newsletter I can get my hands on. One such newsletter suggested checking traffic ranking with Alexa. I did so today and the numbers just don't make sense. If I could direct you to http://www.alexa.com/data/details/re....resortsbc.com
you will see that resortsbc shows Traffic Rank for resortsbc.com: 4,887,136
The two sites listed below as related are far older and more established than mine, which is a fairly new site, yet their Traffic Rank numbers are lower. Do you know why? Does this mean that none of these numbers can really be trusted? Thanks ahead for your answer.
J Baker

in short, the way alexa tracks and compiles its statistics does not reflect a true picture of the current traffic or ranking of your website.

it only tracks the users who has alexa bar installed and computes by averages.

so I would suggest to look closely at the site statistics while only using alexa rankings for Amusement.

[khurramali]

I think Aw stats is a better product then Webalizer.

I have even taken Urchin for a spin but still i feel that Aw Stats does a better job than anyone else in the market.

My Hats off to Awstats.

[timmathews.com]

We use webtrends and have been happy with it for many years

[incrediblehelp]

Great thread on web stat packages:

http://www.webmasterworld.com/forum39/3633.htm

[brian.mark]

Originally Posted by Tubby

Quote
'1 - 2 words are considered "Browsing" terms. 3 - 4 is the "Consideration" phase. 4+ is considered the "Buying" phase.'

Yes that probably what I read. But I have done my fair share of writing and reading, often writers having wondered about something then create some 'facts'. as good reason to write an article. I would far sooner read a post from someone I knew that can state his own observations and where he observed them. There are far to many webmasters creating facts to post in return for traffic. I call it 'Concept spam' I do like to see or know the source. ( I used to write articles part time for womens magazines - does your man realy love you? type of thing where answer (a) was worth three points etc - just for the fun and some spare cash) I do know what crap some writers can come up with,, , just bunkum, but edited more carefully. . .

I would like to see more configurable stats, where a webmaster could punch in the combination he is curious about, something like - Exit links/ original search terms/ length of stay/ sort of thing. and only get stats for these items. Now somebody tell me I can get this.(then tell me it's only $*** per month)

Instead of reciting someone else's research, I'll tell you now that I spent some time digging through our stats. We get conversions from many 2 word phrases on our parts site, but very few 2 word phrases on our tools site. 3 word, 4 word, and 5 word nearly always convert on the parts site (20 - 60%), but most of our traffic comes from 1 - 2 word phrases. On the tools site, conversion rates don't get high until 6 terms, and that's only at about 45%. 1 word phrases haven't converted in the past 12 months. 2 word phrases are again a lot of traffic, but the conversion rate is pretty low. 3 word phrases are more common on that site (2 words for the product with a brand name added.) Most conversions on that site are at least 3 words, staying consistent through 5 and really picking up at 6.

Most of your requests can be done by preconfigured reports on our stats package, with the length of request to conversion rate being a fairly simple custom report. We're paying $900/month for our stats right now, though. I haven't seen things like that in low cost solutions yet. That's considered "Enterprise level reporting". Scenario analysis like this is very useful, but most webmasters wouldn't ever have a chance to look at that. Most have enough trouble understanding the reports from webalizer and awstats.

Brian.

[Steve W]

Very interesting, thanks for posting!

How do the SEs break down for you regarding referrals, if that's something you're willing to share also?

Cheers

[brian.mark]

Originally Posted by Steve W

Very interesting, thanks for posting!

How do the SEs break down for you regarding referrals, if that's something you're willing to share also?

Cheers

On the parts site, the referrals from search engines break down as .59% Google, .13% Yahoo, and .07% MSN. That's right, we have enough other sources of traffic for this particular site that all 3 of the major search engines combined amount to less than 1% of our traffic.

For the tool site, .29% Google, .14% Yahoo, and .06% MSN. Again, less than 1% of referrals from the 3 major engines due to the number of traffic sources we have. Looking at just search engines, Google provides 55% of our search traffic, Yahoo provides almost 19%, and MSN provides 16%. AOL and Ask Jeeves are both about 1.5% of search engine referrals for us.

Did you have any other specific questions that won't give anything of any real value to our competitors?

Brian.

[Steve W]

Hi Brian, thanks for answering.

I expected the numbers to be low, but not that low!

I've read a post elsewhere at WPW (maybe Greeneagle or DMC?) that said something along the lines of if you rely on the SE's for trafiic, you won't do very well... your post would appear to back that up 100%.

So, if it's within your criteria to answer, what are the major sources of traffic for your sites?

And/Or what would you recommend for Open Choice Mortgages as an alternative way to boost traffic.

[Tubby]

brian.mark Brilliant! Thanks brian . .

I have found that longer search terms have more value. My stats are nowhere as specific or sophisticated as yours. I work more on an individual user 'spot check' and comparison between different sections. I am more like a traffic cop, not selling anything. just directing people. (I get the scavenger rights for anything collected on the journey)

I must admit I do like the variety of sites represented at WPW

[brian.mark]

That's an easyt one. This is pretty much internet wide, and there are tons of other sources of visitors.

1) Links from customers. We include HTML code in every delivery confirmation email we send out for a link back to our site and encourage people to relate their experiences on their own websites and/or blogs.

2) Forums. I'm not just talking about WPW. We have a few people that post at various diy type forums that answer questions about some pretty technical stuff in the construction industry (fasteners suitable for certain purposes, what tool has enough power to do a job, what battery does a specific tool need, etc.) and those signature links get tons of traffic.

3) Email to your customers. Simple newsletters can drive tons of qualified traffic to your site.

4) Shopping Portals. (Not for every industry, but certainly any e-commerce site). BizRate, NexTag, Pricegrabber, MSN Shopping, Become, Shopping, Epinions, Shopzilla, CNet, and anything else that matches your product offerings.

5) Vertical creep. Vertical searches (product searches, news searches, image searches, etc.) all are starting to creep into the organic results. Those can be great sources of traffic, and they tend to convert well for our products.

6) Press Releases. We haven't utilized these as much as we should, but they can drive a ton of traffic. Even one that we did up in 2001 has been still supplying us with traffic via CNN.

7) Business partners. We sell products, so the manufacturers of those products link to us as a factory authorized seller of those items. Many are linking from the products themselves to the products on their distributors' sites now. USPS has sent us traffic since we switched a lot of our shipping to them. Every company relies on someone else in business, and some links between those sites can make quite a difference.

8) Sponsorships. We're sponsoring a truck that has been putting up some videos using a lot of bandwidth due to the number of downloads. This simple site doesn't do much, but the forums he runs send us some traffic, as do the forums he posts to about his new achievements. He's also having a simple commercial put in all of the videos now that is generating traffic (now that he's broken 1000 horsepower without the nitro, he's getting a lot more attention.)

9) Newspapers / Magazines. We have had some articles in magazines over the past few years as well as some ads. Articles drive the most traffic (we sold thousands of pairs of ear plugs in a matter of weeks when a pair we carried were mentioned in an article), but ads do increase brand awareness as well as getting some people directly buying because of the ads.

10) Catalogs / Flyers. Yes, you have a website, but that's tough to sit down and thumb through when you get home from work. You've got to get in front of them many times before they'll remember your name. We don't do nearly as much with this as we could, but we're working on it. How many companies send you a catalog? I know I usually thumb through and have a wish-list when I'm done for the musical equipment suppliers. Flyers can work well to. Postcards have worked well for us to increase awareness and gain new customers. Lots of options here.

11) Billboards. They're not that expensive. Just pick out a target area and get a billboard set up. Usually for a few hundred, you can get one to test out your message pretty easy. Being very geographically targeted (physical signs), you can track results pretty easily by comparing leads from that area historically to what you get when the billboard is up.

12) Type in traffic. We know that a lot of people call Spotnails brand products that name without the "S" at the end. We bought that domain (spotnail.com) and do a 301 to our site. Same with rolair.com, justpowertools.com, justcordlesstools.com, justhandtools.com, and the one we get inquries about all the time because there are so many companies with this name - wholesaletool.com. You'd be amazed how many people type some of those names in to see what's there or they think that's the site they wanted, and we see some conversions from that traffic.

13) Use your imagination. There are a ton of possibilities if you think about what an offline business would do. Radio and TV have a lot of opportunities if you target your audience well. Local newscasts (you can select local newscasts in other markets, doesn't have to be local to you) usually have a larger audience than a lot of other programs. Those times work well.

Just because you're online doesn't mean you have to do all your marketing online. You're primarily a business - online is just your main method of contact.

Brian.

[Tubby]

brian.mark

Brian that last post should be a 'Sticky' in the marketing thread. .

[Steve W]

Wow, speechless. Thank you Brian. I shall be re-reading that post many times.
Ditto the sticky comment.

[brian.mark]

Speechless? Wow. I must have out-done myself on that one.

:-)

Brian.

[wilderness]

Wow Brian,
I am truly impressed. I just learned more from your last few posts about marketing a business on and off the Internet than I have in countless hours and days of reading everything around. It's all helpful and if you don't mind, I just copied everything and printed it out for future reference for both my online sites and off line businesses. Thanks and Hats off!!!
JB

[send2paul]

What a great thread!

At some point in the future I will have to get "more savvy" about using stats. (At the moment I just use this guy URL Stats for all my stats needs).

So, tell me anyone, apart from the diasadvantage of perhaps URL Stats not being available one day due to server issues etc, or the fact that the stats produced there are not as flexible as the other stats produced in Awstats etc - what is the big adavantage of all these "script based"/"load files on your server" type stats programmes?

The reason that I'm asking is that I haven't got a clue where to start with something like Awstats etc, as my techhie knowledge is sadly lacking in that area :)

[Tubby]

Impressive isn't it. I am sitting here wondering what brian.mark missed - I think he covered letter box drops in flyers ect - Posting message on notice free shopping centre noticeboards, maybe bumper stickers . . But it is the post of the month in my view
In fact the post does look like a chapter summary that could be presented to a publisher. If I were a publisher I would forward him a check and the contract to sign . . .Before the opposition publishers got hold of it.

[DrTandem1]

Originally Posted by Steve W

Hi Brian, thanks for answering.

I expected the numbers to be low, but not that low!

I've read a post elsewhere at WPW (maybe Greeneagle or DMC?) that said something along the lines of if you rely on the SE's for trafiic, you won't do very well... your post would appear to back that up 100%.

So, if it's within your criteria to answer, what are the major sources of traffic for your sites?

And/Or what would you recommend for Open Choice Mortgages as an alternative way to boost traffic.

I may be wrong, but I think Brian forgot to multiply by 100 to get his percentages. I doubt he was tracking figures of less than 1 percent as that would be statistically zero. If he actually is receiving "less than 1%" of his traffic from the search engines, there is definitely something else going on with his site.

Brian went on to list a lot of very good traditional methods of advertising. If that is where the site is receiving its traffic, then indeed his statistics could be correct and very detailed at measuring values <1. This also would be an anomaly of a post, as most here preach SEO as if it were the Gospel.

[brian.mark]

Originally Posted by DrTandem1

I may be wrong, but I think Brian forgot to multiply by 100 to get his percentages. I doubt he was tracking figures of less than 1 percent as that would be statistically zero. If he actually is receiving "less than 1%" of his traffic from the search engines, there is definitely something else going on with his site.

Brian went on to list a lot of very good traditional methods of advertising. If that is where the site is receiving its traffic, then indeed his statistics could be correct and very detailed at measuring values <1. This also would be an anomaly of a post, as most here preach SEO as if it were the Gospel.

Yes, Doc, you are wrong. While we do quite a bit of SEO, we don't get most of our visitors from Search Engines. Sending out postcards to a targeted list has resulted in some huge numbers of visitors. Shopping Engines (I suppose those could be also considered Search Engines, but they're more predictable than an algo) provide a lot of good quality, high converting traffic. Links from other sites and forums as well as our email campaigns actually result in more traffic than search engines send.

I didn't list everything we do, but those are the ones that didn't require as much budget. Sure, you can get a Superbowl ad or a national ad during a Nascar race or some primetime show, but that's not in most small business budgets. We have really gone for stuff we can track the ROI on by targeting geographically in small quantities first, then growing it as we see fit.

Froogle to me is a different campaign than Google, so the "Product Results" on the organic pages don't count as Google for me. That's providing nearly as much as Google is at the moment for us. Sure, it took some real creativity to get traffic to our parts site via Froogle, but I got it working and the conversion rate is great.

The main thing that most online businesses miss is the fact that they are a business first and online being secondary. I hope some of you get some ideas out of this thread that work for you. I know I'm always surprised when I look at our stats and see how small of a percentage comes from the big 3 search engines, but I know that means that the rest of our marketing program is working and we're algo-change proof.

Brian.

[Tubby]

I have mentioned on a previous post somewhere, that on sections of my site I get more traffic from users favourites folder than google, So I am not suprised.

brian.mark is extremely thorough and a shining example of just how thorough you need to be to keep on top of a competetive field . . Lets face it he is competing with the whole world and doing it extremely well.

When I was a kid in the city I used to see shopkeepers stood at the front door of their shops in the afternoon, waiting for customers. Those days are gone. The New shopkeepers now do not have customers walking past. . They have to go out and get them.

I saw our local hairdresser the other day stood at his shop front door. I do need a haircut, but when your hair has best part fallen off the top - the sides seem far less important.

This thread is about statitics, I for one am very pleased it turned out not to be
another Googlegrovelgrumble

[DrTandem1]

Originally Posted by brian.mark

The main thing that most online businesses miss is the fact that they are a business first and online being secondary. I hope some of you get some ideas out of this thread that work for you. I know I'm always surprised when I look at our stats and see how small of a percentage comes from the big 3 search engines, but I know that means that the rest of our marketing program is working and we're algo-change proof.

Brian.

Exactly. It also shows the importance of a good domain name that is easy to remember and spell. No one needs to search for a name they know.

[swordfish-1]

If not for my webstats, I would be lost in cyberspace. Not only do they let my know how to best target my audience, but they also let me know when to stay put with my SEO when the Google dance begins.

I have gone up and down in rank at times when Google
updates, and have suffered through a few traffic swings. But my 30% bookmark stat helps keep my feet on the ground. These people keep comming back for a reason and buying our products is one of them. So any tweaks that I do for the search engines are usually minor.

My other favorite stat to look at is the duration of visit. I watch this one the most carefuly because the longer people stick around, the more likely they are to be a repeat visitor and customer. This stat guides me to make my site more appealing to the eye
and more simple to navigate even at the risk of my pages taking more time to load. The short time visitors are usually just surfers or redirects.

I check all my stats weekly. Keywords,origins, and page visits are all important, but I like to look at load times, systems,servers, as well as the stats mentioned above to see if I am doing my best for the people who visit us.

[Tubby]

Stats are Terrific for reducing bandwidth.
I know it can sound silly to boast about reducing page views , but in some cases this can be a fantastic thing. If you can get a couple of hundred users to avoid loading a single page while navigating it has big benefits.
Why use up 10 page views to sell a chicken, If you can sell it in one or two?

[brian.mark]

Originally Posted by swordfish-1

My other favorite stat to look at is the duration of visit.

I used to like that one, until I noticed that the long visitors we see are from competitors. The longest visits are from Amazon employees. I suppose that's cool, but doesn't tell me squat about my actual site visitors. It just means that they're searching to see if we have some of the products that they carry and what our prices are.

On the plus side, that means we made their radar. ;-)

Brian.

[Serj]

I use AWStats also,and my favorite thing to watch when starting your next site is to watch how the GoogleBot, The Slurp, and the MSN bot, and who your first referers are, etc. The beginings of a site are always fasinating to watch. I also use the error pages hits in AWStats to make sure that there are no goof ups anywhere...also interesting to see if, I have a password-secure section for webmaster purposes, and you can see how some people who figure out that that section exists, try figure out the user name and password...

[jackson992]

The thing I would most like to see is something that will tell me specific engine traffic by day. This lets me know exactly when I go up or down in the engines and on what engine. Presently my stats don't tell me this and I use awstats.

[brian.mark]

Originally Posted by jackson992

The thing I would most like to see is something that will tell me specific engine traffic by day. This lets me know exactly when I go up or down in the engines and on what engine. Presently my stats don't tell me this and I use awstats.

Most higher end packages will do this. Webtrends will, ClickTracks will, Urchin (Google) will, HBX will, Index Tools will. Anything that you can do a real-time report by querying a database (instead of just viewing a webpage and having it pre-made) should be capable of this. There are a lot of other names, I just don't remember them all.

Brian.

[jackson992]

Thx Brian:

This was very useful as I do have Urchin but I've never taken a look at it. I'll be sure to check it out now.

[greeneagle]

Brian.mark:

“While we do quite a bit of SEO, we don't get most of our visitors from Search Engines.”

This month, we are running:

Direct address / Bookmarks - 68.5 %
Links from an external page (other web sites except search engines) - 19 %
Links from an Internet Search Engine - 12.2 %
Unknown Origin – 0.1%

“..our email campaigns actually result in more traffic than search engines send.”

From what I have seen, this seems to be one of the best ways to generate repeat business for many on-line wholesalers and retailers.

Ken