Omniture SiteCatalyst vs. Google Analytics

Posted on October 14th, 2006 by Bart Gibby.
Categories: SEO.

At InfoBase Media Corperation we use Omniture SiteCatalyst which is very complex and is meant as a enterprise solution. We’re not that big trust me, the actual license is actually like $30,000 a year, thankfully we just buy from a reseller called 10XMarketing.

The funny thing is that both companies are located in the Provo-Orem Utah area just like us. I guess the apple doesn’t fall too far from the tree, unless of course your Google. What I mean to say is that SiteCatalyst frankly is too much for us and is just too expensive for what we use it for. None of us are trained to use it to its fullest capacity and its even been giving us bad data the last week or so.

I think it is mainly because the SiteCatalyst code may not be on every page. It was the past but since we have been adding the Google analytics, which is free by the way, to our site some f the SiteCatalyst code may have been deleted by accident or something. I am not sure the date error could be caused by not having the code on every page due to the nature of the data, which is visits vs. unique visitors.

First I feel it is important define each, both definitions are taken from the SiteCatalyst Glossary of Terms.

Visit:
The number of times a visitor browses to your site. A visit begins when a person first views a page on your company’s Web site, and lasts until that person stops all activity on the site for a period of 30 minutes (Web analytics general industry standard). The industry assumes that, given the length of time, the person likely closed their browser and reinitiated a separate “visit.” A unique visitor could be responsible for multiple visits (or sessions) during a day, month, and year.

Unique Visitor:
Registered when a person visits a site for the first time within a specified period of time. In other words, while one person may visit your site and view the front page three times, a contact page twice, and several other pages once each, this person would be recorded as one “unique visitor” so you can tell exactly how many individual people are coming to your site.

There are four different time frames SiteCatalyst uses to define unique visitors: hourly, daily, monthly and yearly. An hourly unique visitor may visit your site at 2:01, 2:45, and then again at 3:24. This visitor will register as two hourly unique visitors, because only the 2:01 and 3:24 visits would count as unique visits for their respective hours (2:00-3:00 and 3:00-4:00). This same standard is used to determine daily, monthly and yearly unique visitors for their respective time frames. It can be helpful to ask the following question to see how SiteCatalyst determines who is a unique visitor: “How many different people saw my site during this time period?”

Note: Daily Unique Visitor reports are different from the others in that it is not only cookie-based, but is also IP address based to help catch everyone who visits your site.

Note: The sum of all daily unique visitors is not equal to the total monthly unique visitors for that month. This is because a visitor who returns twice in a given month will count as two daily unique visitors – one for each day they visited, but only as a single monthly unique visitor.

Because a unique visitor can return and make many visits in the course of a day or a month that there should always be fewer Unique Visitors than Visits or they should be equal, Unique Visitors < or = Visits. Another way of putting it is that Visits should never be less than Unique Visitors. But this is exactly the bad data I have been getting from SiteCatalyst for over a month now.
Here is some data examples:

Date Unique Visitors Visits Variance
Sep 1, 2006 807 812 5
Sep 2, 2006 997 1003 6
Sep 3, 2006 1383 1348 -35
Sep 4, 2006 1053 1030 -23
Sep 5, 2006 1006 993 -13
Sep 6, 2006 926 903 -23
Sep 7, 2006 964 949 -15
Sep 8, 2006 1299 1249 -50
Sep 9, 2006 1184 1134 -50
Sep 10, 2006 1635 1569 -66
Sep 11, 2006 1327 1320 -7
Sep 12, 2006 1394 1364 -30
Sep 13, 2006 1153 1127 -26
Sep 14, 2006 1082 1074 -8
Sep 15, 2006 988 987 -1
Sep 16, 2006 1237 1217 -20
Sep 17, 2006 1681 1637 -44
Sep 18, 2006 1092 1093 1
Sep 19, 2006 1110 1136 26
Sep 20, 2006 1144 1121 -23
Sep 21, 2006 1824 1755 -69
Sep 22, 2006 1179 1177 -2
Sep 23, 2006 1238 1219 -19
Sep 24, 2006 1651 1575 -76
Sep 25, 2006 1214 1226 12
Sep 26, 2006 1215 1208 -7
Sep 27, 2006 1132 1128 -4
Sep 28, 2006 1090 1095 5
Sep 29, 2006 999 1009 10
Sep 30, 2006 997 970 -27
Total 36001 35428 -573
Oct 1, 2006 1720 1690 -30
Oct 2, 2006 1285 1253 -32
Oct 3, 2006 1166 1166 0
Oct 4, 2006 1075 1095 20
Oct 5, 2006 979 979 0
Oct 6, 2006 987 1001 14
Oct 7, 2006 961 910 -51
Oct 8, 2006 1511 1476 -35
Oct 9, 2006 1366 1398 32
Oct 10, 2006 1659 1618 -41
Oct 11, 2006 1580 1537 -43
Oct 12, 2006 1003 1007 4
Oct 13, 2006 948 941 -7
Oct 14, 2006 623 595 -28

Google on the other hand has not been giving me bad data like this. Now keep in mind we just got all our pages added with the Google analytics code, or at least we think we have. Speculation still exists because Google consistently gives us lower numbers than SiteCatalyst for everything, but at least the numbers have been consistent. Here is our Google data:

Date Range Unique Visitors Visits Variance
Fri 9/1 743 921 178
Sat 9/2 831 1000 169
Sun 9/3 1272 1476 204
Mon 9/4 904 1045 141
Tue 9/5 921 1083 162
Wed 9/6 888 1054 166
Thu 9/7 886 1026 140
Fri 9/8 1192 1440 248
Sat 9/9 1063 1218 155
Sun 9/10 1489 1713 224
Mon 9/11 1230 1458 228
Tue 9/12 1093 1291 198
Wed 9/13 1072 1257 185
Thu 9/14 987 1181 194
Fri 9/15 899 1105 206
Sat 9/16 934 1099 165
Sun 9/17 1489 1718 229
Mon 9/18 1015 1188 173
Tue 9/19 1010 1213 203
Wed 9/20 1009 1226 217
Thu 9/21 1356 1677 321
Fri 9/22 585 700 115
Sat 9/23 591 682 91
Sun 9/24 846 991 145
Mon 9/25 644 763 119
Tue 9/26 580 708 128
Wed 9/27 575 688 113
Thu 9/28 547 664 117
Fri 9/29 459 579 120
Sat 9/30 480 607 127
Sun 10/1 826 1014 188
Mon 10/2 694 814 120
Tue 10/3 585 697 112
Wed 10/4 552 682 130
Thu 10/5 523 627 104
Fri 10/6 473 584 111
Sat 10/7 457 533 76
Sun 10/8 735 873 138
Mon 10/9 955 1175 220
Tue 10/10 1482 1726 244
Wed 10/11 1286 1514 228
Thu 10/12 944 1116 172
Fri 10/13 865 1034 169
Sat 10/14 467 518 51

Conclusion: Google Analytics can actually give you more reliable data than Omniture’s SiteCatalyst and it’s FREE!

41 comments.

Michael M. Choi

Comment on December 10th, 2006.

Very interesting article Bart. It was very informing.

Robert

Comment on February 22nd, 2007.

I have come upon the exact same conclusion. I have had Omniture training, and have seen it in action on 3 major sites I have helped manage and analyze. For my personal sites I go for the free Google Analytics.

Omniture is full of bugs and the interface makes it difficult to visualize correlations properly. GA has an extremely simple drilldown process in which your information is presented so nicely that you don’t have to split your head open to analyze.

Granted, there are less custom variables, and less possibilities with certain tracking elements (campaigns, trkids, etc), but let’s face it, when you’re analysing every little thing to such detail, you’re not spending time adding promotions, content and ideas to your site that could make a nig difference in attracting qualified traffic…

Google Analytics is more than enough for any small to med-large site…. omniture is for muli-million dollar companies… that have the time and resources to waste money on a whole “tracking” team.

Jeff Lawrence

Comment on March 21st, 2007.

Thank you for the information Bart. Have you noticed a substantial unique visitor count beyond 10-15% when using Google Analytics compared to a log based solution?

BartGibby

Comment on March 22nd, 2007.

Umm… I have not done the math on that. I do have a hunch that Google doesn’t give me all my keywords like my log based program(s) seem to do, or at least that tracking of keywords will be different between the two.

But it would make sense that there would be an increase in the accuracy of the statistics gathered by Omniture Site Catalyst, Google Analytics, and other solutions that use more than just logs such as cookies.

However, this may even decrease the unique visitor count. I would really like to get more into conversion and tracking. In fact I’ll take a look at this in one of my future blog posts.

Thanks Jeff, I hadn’t even thought about how big the difference between the two would be.

Cheers,

-Bart

Kelly

Comment on March 27th, 2007.

We installed Omniture late 2006 and I have been noticing odd behavior since installation on the reporting end. Such as; 15-20% inflation compared to the engine reports (visits, views); pages reporting traffic that weren’t live yet; pages appearing within the clickmap report that aren’t part of a page. When I ask about these items, I am told the reporting is fine; like I am suppose to accept that answer and forget about the mystery occurences. Has any one else experienced these behaviors?

BartGibby

Comment on March 27th, 2007.

Pages not being \”live\” can still get traffic, search engines are amazing today they can drill, prod, and poke in places they have never been able to go before.

This could be one reason for your traffic. Another is if you have a large company or even many team members just working on the pages could count as traffic.I am betting its the last, I would go into the Omniture settings and have your office specific IPs or cookies not tracked. I am not sure to what extend Omniture can filter out developers traffic, but I would assume it would be adequate.

If you have a changing page due to development and have added the Omniture code to that page before all the changes then that would also most likely explain your click map issues.

its probably a good practice to add the Omniture code to your pages as your final step before making the pages live.

I hope this helps, Cheers!

-Bart

Kelly

Comment on March 27th, 2007.

No, that page had not even been developed yet; the page was built weeks close to a month later. Omniture was the last step when we did go live with this page, though.

In regards to the click map, the page has not been touched (this is a different page btw).

Thanks for your feedback :)

Kelly

pria

Comment on April 25th, 2007.

i noticed differences in the numbers reported via click map and the numbers from the data warehousing report??

Bart Gibby

Comment on April 25th, 2007.

pria,

I am not sure whether you are asking a question or making a statement?

-Bart

Matt Newman

Comment on May 23rd, 2007.

Bart, what is up? This is Matt Newman, from high school. I stumbled onto your site when I googled:

SiteCatalyst vs. Google Analytics

I am the webmaster at Wheelchair Foundation, and I have come to the same conclusions you have about Google Analytics and Omniture SiteCatalyst. With Google’s recent upgrades to Google Analytics, I think Omniture is going to lose market share fast unless they do something different. You can’t beat free, especially when it provides more accurate, more user-friendly, and better-managed data. Like all things Google, Analytics is an amazing product that only a big company with nothing better to do with hords of cash and a bunch of geeks with phds would produce for free.

Peter Andersson

Comment on July 13th, 2007.

Do you know if omniture by default use 1st party cookie or 3rd party cookie counting Unique visitors/browsers?

Bart Gibby

Comment on July 13th, 2007.

Peter,

From what I know they use their own cookie for tracking. I do believe there is settings in the sitecatalyst admin to set the cookies for certain time frames.

Useful if you want to expire a special sale or something.

Does that answer your question? Or are you asking if they can read other websites cookies placed on the users machine?

A coworker of mine also mentioned that:

“All tracking systems use third party cookies, cookies are domain based, and you can only drop a first party cookie from the domain or root domain that the browser is accessing. For example: if I am on toys.ebay.com, I cannot drop a cookie for http://www.ebay.com, but I can drop one for toy.ebay.com or ebay.com.” - Jared Turner

Britton

Comment on August 11th, 2007.

Don’t want to burst your bubble here. Omniture data isn’t bad folks. They simply have a different setting that doesn’t count a visit if the visitor rejects third party cookies. Google obviously chooses to count visits even if the user rejects a third party cookie. So this boils down to a choice of when to count a visit or not based on user browser settings.

I do agree however that if your company is small, free is hard to beat. If you’re enterprise, Omniture is a no-brainer over Google. There’s a reason why more and more enterprise folks choose Omniture. It simply gives these companies more of what they need: the ability to customize their tracking solutions like GA can’t. If your small and don’t need all that customization, Google isn’t necessarily a bad idea.

Bart Gibby

Comment on August 11th, 2007.

Britton, I really appreciate you defending Omniture’s Site Catalyst. They are a local company here and I have even had them contact me for to be a “Best Practices Consultant”. But most of all I would love to see them maintain their dominance in enterprise level analytics soultions.

Too many Utah (Utah Valley mainly) based companies like WordPerfect, Novell, and others have been destroyed by giants. The two mentioned of course were by Microsoft.

Google has yet to devastate anyone else that I know of as Microsoft has when they release a product. Google just doesn’t have the distribution channel that Microsoft has had… so they have had to create a new one.

Back on track here, I still do not see why SiteCatalysts numbers would have been so much higher than Google’s if is counting less visits than Omniture. But the most important thing is that there is a negative variance with Omniture and not with Google. Thus I deem Google’s data to be more reliably consistent when it comes to Visits and Unique Visits.

Please, pleas elet me know if they have fixed this… eventually I think my new work OrangeSoda would love to make an agreement with a top tier analytics company.

Ryan Roberts

Comment on August 31st, 2007.

Well how valid is your comparision if you think your Omniture code is implemented incorrectly anyway. Granted their shouldn’t be more uniques than visits. My experience with Omniture has been really good. When we have run into weird reporting anomalies they’ve been able to correct them after working with their support.

Ilkka Jormalainen

Comment on December 5th, 2007.

Hi! Im from finland and there is a really intresting web tracking program here called “Snoobi”. I haven`t seen omnitures reporting but compared to google analytics, snoobi gives me much more valuable information about my website. There`s a really good search option in snoobi. For example you last list every visitor with snoobi who has came to my website in 5 days and used word “holiday” in google. Then i can see what pages every separate visitor has clicked in my website and in what order. Then i see for how long they have been in every page. I can also see in what firm they work for and how many times have they visited in my site. I can also make a search that how many of them has visited in some exact page :) This is really good for my addwords campaign.

Bart Gibby

Comment on December 5th, 2007.

Ilkka,

do you have a URL for the “Snoobi” analytics? I am currently testing three or four different analytics on my website. I would love to add one from Europe.

I am always looking to highlight the best tools for webmasters.

Cheers, - Bart

Snoobi Oy

Comment on December 11th, 2007.

The URL is http://www.snoobi.com. For further details you can contact Jukka Ropponen 0400423910

Victor Edinian

Comment on February 1st, 2008.

Hey folks, interesting thread we have here. I agree that this is a very difficult comparison because the page tags in the above examples were not implemented correctly. And although I have used Google Analytics and WebTrends, I am anxious to use Omniture and Unica at an enterprise level.

With this said, I think the core reason to use any analytics solution is to bring more value and less waste to your site. If you have millions of people visiting your site, a 10% improvement is 100,000+ unique improvements, which hopefully should increase conversions by an X amount.

Now the brass tax, which is the holy grail out of all enterprise analytics softwares that you think will make it simple to optimize your efforts that 10% more?

Let’s keep it simple…and discuss…

Google in my opinion does the job, if you have tagged your pages correctly, and you have 100% implementation. The problem is that it does not do a good job with offline efforts.

WebTrends on the other hand, is decent. It does have many features that will allow you to work with databases, email programs and more. The problem I found is that they lack in many add on’s like multi-variant testing and the sort.

Therefore, any true success from Omniture with these other issues? I’m curious to find out…

Thanks,
Victor

Melissa

Comment on February 26th, 2008.

Hi,

I come from an interactive agency, and we have implemented SiteCatalyst 13.5 on about 20 different client sites. We started with Omniture about 9 months ago. It seems to be an uphill battle trying to decifer good data from bad. On some sites, our campaign ids are getting dropped, so we can’t tell what email, banners, prints ads are working and which are not. On one of our more complex sites we have A LOT of custom settings, but the reports contradict each other. SearchCenter data conflicts with SiteCatalyst data. On other sites, we see orders that we know were purchased, have been lost by the analytics system.

Is it possible that there is no one product out there with 99% accurate data? Is it possible that we are supposed to use these tools to help us with trending, and give an overall idea of what’s going on, but have to accept the fact that it’s not accurate?

Bart Gibby

Comment on February 26th, 2008.

Melissa,

This is a great comment. I appreciate you taking the time to share this perception with us. It has been over a year since I posted this blog post. Since then I have also come to the same conclusion.

That no one tool can actually give you 100% accuracy. I have three free analytics solutions installed on my blog at the moment. They all differ with the data I receive from them.

Someday I hope to post a real case study about them. I have not because there are several out there already.

Again thank you for taking the time to write to all of us.

Cheers,

-Bart Gibby

jack casey

Comment on March 14th, 2008.

We TRIED to implement site catalyst for over 6 months and found so many inconsistencies that we junked the product. It is so complicated and un userfriendly that you need 2 dedicated people just to error check the product. The overhead of running the product was way too high. We ended up going back to Google for some fresh air. However we quickly grew out of the top line numbers we got from Google as you could not drill down to any degree and it was funny how Google always appears to account for over 60% of site traffic when our logfile analyser showed categorically they accounted for 40%. This made us suspicous that Google overinflates their own figures to support their PPC and since they are not audited i think it is easy to believe that Google are looking after their own. There is a huge conflict of interest to sell PPC and then report on it impartially.
We have since trialied Webtrends, Intellitracker and unica. We found all systems to be ok but unica and Intellitracker stood out a mile as the product for the serious business. Check them out both out as they make omniture and Google look like clowns when it comes down to power, userability and ease of use. Amazing products guys!

Bart Gibby

Comment on March 14th, 2008.

I have never heard of Unica. And I didn’t know WebTrends sold analytics. It seems everyone and their dog is doing analytics now a days.

I know we have very basic analytics program at OrangeSoda, Inc. but we don’t sell it separately from our PPC and SEO services. Its more of a value added service we provide.
Funny thing is I have not even used it on my blog yet. Simply because I have not received authorization for a an account yet.

I would like to check out these others you have mentioned Jack. I appreciate the time you took to tell us about your experience with SiteCatalyst and Google Analytics.

Ken

Comment on April 1st, 2008.

I’d like to make a bold statement - there will never be a package that offers 100% data accuracy (even statistics don’t have 100% confidence anyways). The reason is simply because what is involved in the process:
- all of what you have already mentioned;
- tracking implementation (in flash vs. html);
- scripts operation calling the right tags;
- client server sending data;
- vendor server receiving data and depositing data accordingly;
- reporting platform pulling data accordingly.

If an error occurs in any of those steps, you will not see accurate data. Can they be fully operational 24/7/356? …. didn’t think so! Being said, I think we need to focus more on how we should utilize these tools.

I wouldn’t want to repeat what’s already said - large firm with good staffing should go for Omniture, etc. What I want to bring up is the lack of collaboration between Marketing and Analytics. I’ve yet to see any company doing a good job at using these tools to support their marketing objectives - I don’t just mean offline analytics, I mean the integration of all types of analytics. Web is only one of many marketing channels. If we have the Integrated Marketing Paradigm, why not an Integrated Analytics Paradigm?

Bart Gibby

Comment on April 3rd, 2008.

Ken,

I really like where you are going with your comment. “Integrated Analytics Paradigm” is a very thought provoking topic. Especially for companies who use their website as one of their many sources of advertising and revenue.

Maybe that the next step with Omniture is to fully integrate a enterprise level full analytics system.

Crunching numbers from all sources.

I also like your bold statement. That the perfect data gather analytics solution is not out there and probably never will be. Which as you mention is due to the complexities involved and all the different types of technology working together.

I completely agree with this.

I would still like to point out that, just because it is expensive doesn’t mean it is more accurate. On a side note just because it is free doesn’t mean that it is less accurate. As my data I collected above suggests.

I hope Omnitire has fixed this issue. It is sad day for a company when a competitor can give a product away for free that gives a higher rate of data confidence.

Again Ken thank you for the awesome foresight into what a true analytics solution can become.

Patrick

Comment on April 23rd, 2008.

Has anyone run into a scenario where the clicks being reported by the search engines and omniture are close, but the difference between clicks and unique visitors is huge? I’ve been able to accept a 10-15% difference, but when it gets up to 40% a red flag goes up. Say a keyword drove 9,000 clicks, but that only translated into 6,000 visitors. I find it hard to believe the difference is from the same visitor clicking on an ad multiple times, unless there’s fraud involved.

curtis

Comment on May 14th, 2008.

Yes, a very good info on analytics software and the competition. After reading a little more on snoobi I see some faults in Google analytics . Intellitracker’s site did not offer much info at all. What I found most was their client list. And they measure site visitors..kind of funny there.

Matt

Comment on May 24th, 2008.

Re Integrated Analytics Paradigm - wouldn’t this just be database marketing?

Bart Gibby

Comment on May 27th, 2008.

Matt,

Do you mean “database marketing” as the database is the product or service that is being sold? Or using databases to do marketing?

We are talking about internal data used to sell more product or services, add more value to our current and future customers with those products or services, and do it with the least amount of waste by optimizing and integrating marketing campaigns.

See the largest problem with databases is they have to be used by everyone involved inside the company. Even with proper relationships between the databases people still need to understand the data and use it.

It also needs to be resource effective. Which right now a system that does all this would either have to be patched together from various 3rd party and internal databases or all custom built internally or by a 3rd party.

Do you know of any solutions?

-Bart

Lisa

Comment on June 5th, 2008.

We tried to use http://www.omniture.com, but we encountered highly questionable JavaScript coding practices. Our own appraisal of the Omniture code has shown that their developers have attempted to generally follow the variable naming requirements that are described above, but have only done so partially, rendering the effort useless. Thus we are faced with not only a lack of cooperation with the vendor, but a general incompetence. Please see the Omniture representative’s response below.

/Omniture representative wrote: I’m sorry, but it is not possible for me to have our code rewritten to stop using “f” as a variable, or to have all of our functions and variables renamed. In addition, I cannot provide you with a list of all variables in the s_code.js file because of intellectual property protection issues.”/

The claim that a simple variable name could have intellectual property implications is patently ridiculous.

Anthony

Comment on July 4th, 2008.

Hi we are web interactive company consist of less than 20 workers with less than 5 person handling technical/database and analytics issues.

we had been using Google Analytics all this while until yesterday our directors bring front Omniture calling it the future of analytics.

they are not a technical person. they know nothing much about GA as well as Omniture.

Even though we are a small company, we run web campaigns and solutions to multi million corporation like Samsung, Pepsi etc.

the question here is, can 2-3 technical person handle the problem that might arise using Omniture? should we stick back to GA?

thank you for reading

Bart Gibby

Comment on July 4th, 2008.

Anthony,

If I had a team of 2 to 3 people using Omniture’s SiteCatalyst I would be the happiest man alive.

What I mean is, even though Omniture has its flaws, its main advantage in the future is that it is not Google.

To answer your question if 2 to 3 people is enough or not. That depends. I would want at least one person focusing on website conversion using the SiteCatalyst product.

If your 2 to 3 people know and are trained on the program it can be the most powerful tool you will ever use.

Again, you have to be trained. It is so in depth and covers so many areas of website analytics that you need to be patient and learn it. It is completely customizable!!

I only used it for 6 months. I taught my self what I could. But I regret to inform you that my main function was to report on SEO campaign metrics using it not increase website conversion.

So be sure that your company knows using the product will be a waste, a huge waste of money if they do not pay to have you trained properly and in depth on how to use it.

If they don’t you might as well stick with GA.

P.S. Love your website. Some great design work. Never actually seen a website offer to go full screen and actually have it make my entire browser disappear. very different. Hope you guys get some targeted website traffic from the search engines with it.

ben

Comment on July 26th, 2008.

The reason that Visits are lower than Visitors is because Visits are only counted for people that come to your site who are accepting persistent cookies. Visitors are counted even if the persistent cookie is not able to be set.

You actually quoted this little tidbit straight out of the documentation, though it’s not obvious from the wording. I would imagine any of the support people at Omniture would be able to tell you that as well.

Anyway, Google counts Visits for both persistent & non-persistent cookied visitors, hence the difference in numbers. For the most part, comparing two different systems like that is dangerous. Some numbers will match, some won’t, and it’s difficult to tell if they’re supposed to or not. Anyway, hope this helps.

Bart Gibby

Comment on July 28th, 2008.

Ben,

I appreciate your feedback. Very excellent point about the cookies.

However, don’t you think it contradicts the whole idea of tracking both visits and unique visitors? Why not just track one and avoid confusing the user(s) of the SiteCatalyst system.

I prefer a system that forces me to use consistent measurements. That way I know whether the data is truly related to each in valid methodologies.

Just set me up a system that works, is consistent and doesn’t contradict itself. That is all I ask for.

Ben

Comment on August 1st, 2008.

I agree completely. I’m not saying I agree with the way they set it up. I actually hate it, but knowing is half the battle right?

BrownBoy

Comment on August 1st, 2008.

Bart,

All pathing reports in SiteCatalyst are, quite naturally, built around the concept of a VISIT. If you include the the visits of non-cookied users with the visits of cookeid users, you potentially compromise the validity of your pathing data. Non-cookied users are identified using an IP+browser method. It’s the next best thing after no cookie, but still inherently inaccurate given shared IP addresses via proxies, IP-pooling as practised by most ISPs, etc.. In essence, including the visit data of non-cookied visitors in your pathing data pollutes good data (from cookied users who have good visit integrity) with bad (lots of users looking like the same user due to IP sharing and single users looking like multiple users courtesy of dynamic IP allocation).

I take your point that this can be confusing for those new to SiteCatalyst and the lack of consistency between visit and visitor could be a problem to some, but to flip it around, does it make sense to compromise the validity of pathing data just to achieve this consistency?

Note also, the pages seen and events triggered by non-cookied users are of course recorded by SiteCatalyst; it’s just that the visit metric itself does not include the visits of non-cookied users.

For the rest of the comments, it’s difficult to know where to start. Nobody’s going to have a quarrel with “I prefer X” or “I think Y only works for big companies”. These are opinions and everyone is entitled to them. But there are several statements of “fact” above that are, I’m afraid to say, borne of sheer ignorance. My favourite concerned the “highly questionable JavaScript coding practices”. Quick, someone call AOL, Microsoft, Yahoo, eBay, Disney and the rest of the Fortune500 companies out there whose due dilligence failed to uncover this priceless gem.

Apologies for the sarcasm, but really….

Bart Gibby

Comment on August 1st, 2008.

BrownBoy,

So if I am calculating conversion rate and I have polluted data, then that gives me a polluted conversion rate. Great, I now have data that is either saying my rate is too high or too low.

If Site Catalyst has a way of turning IP address or Cookie tracking on/off. Then I say Good for them and all who use it. If not then I say bad for them and you are welcome to your contradicting data.

There is no point in tracking a relationship between two sets of data if you can never get the proper data to leverage that relationship. So just pick one set of data, forget about the relationship. Gather just that set of data as accurately as possible.

Google seems to get it right… I am fine with lower numbers so long as they are consistent and the relativity of the relationships between the data collected are sound.

Stop overloading me with data I can’t use. I need data to make business decisions now. Don’t make me have to do more calculations. I don’t need corrupted data that leaves me wondering if it is accurate enough to get the job done.

Bart Gibby

Comment on August 1st, 2008.

Ben,

I agree, knowing how they get their data can be very useful. Especially when the data starts to make you wonder, “How did they get this data?”

When data starts to get you thinking like that, it is either very good or its very bad. its usually very bad in my experience.

Ya so the other half of the battle would be compensating for any data that is not valid. Or just starting over.

BrownBoy

Comment on August 1st, 2008.

Bart,

You can run reports in SC that allow you to see both cookied and non-cookied visitor numbers. You can get visits per visitor using like for like (i.e. for cookied user only) at the flick of a switch. The key to all analysis is knowing where you’re figures come from. If you know that your visits metric does not include non-cookied users - who you can quantify using other reports - then your covnersion data is not “polluted”. It’s only “polluted” if you either:

a) don’t know that visits excludes non-cookied users, or
b) have no way of assessing/quantifying the number of non-cookied users.

The reality is that when trending data, the number of users rejecting cookeis is pretty much immaterial anyway. This is not quite so with pathing data, as a relatively small segment of users who share a same proxy IP, or who are getting different IPs in mid-visits due to dynamic IP allocation by their ISPs, can disproportionately impact the integrity of pathing data in quick order.

Look, I accept the above is my opinion on what’s more important, but I’m willing to concede that the point is arguable. I’ve no doubt it is *possible* to get decent pathing data out of Google Analytics and, even if the data is compromised by non-cookied visits, this ceases to be quite so importnat if analysts know that their data includes non-cookied visitors; their fallout trends over time will still yield quality intelligence that can be acted upon. Likewise, I’d aruge that the absence of non-cookied visits from SC’s visits metrics is not the big deal you might think it is so long as analysts know that this is the case. Knowledge is power, and all that.

I have no problem with users who prefer one approach over the other. That’s the type of stuff that makes the world go around. But your insistence that one approach yields *bad* data as oppsoed to *good* data using the other solution, needs to be challenged.

Stop overloading me with data I can’t use. I need data to make business decisions now. Don’t make me have to do more calculations. I don’t need corrupted data that leaves me wondering if it is accurate enough to get the job done.

The point is that you wouldn’t have been “wondering” anything if you actually knew how Omniture calculated visits, something every user discovers on first day of basic user training. This isn’t an attack on your intelligence - I’m sure there’s things you know that I don’t. But now it has been explained to you why you saw what you thought was a data anomaly but is in fact perfectly valid data, why not just accept that this is just a different approach that has it’s own benefits, just as Google’s approach has benefits, too? Why the insistence that your analysis efforts in SiteCatalyst are rendered by pointless without an exponential increase in effort? It’s just not the reality, my friend.

Respetfully,

BrownBoy.

Bart Gibby

Comment on August 4th, 2008.

BrownBoy,

I am so glad you have actually been able to help me finally. Its been almost 2 years since I posted this. I never received training, so I was left to my own understanding of the program.

So for the analytics laymen, which is almost everyone who has posted a comment here. Omniture’s SiteCatalyst really needs either to be more usable for the laymen or Google analytics will always win.

I love SiteCatlyst, its just not meant for people to understand without training. If you read my summary at the bottom of my post or even my followup post you’ll see that.

And yes BrownBoy, you win, when it comes to how much effort I put into knowing how to use SiteCatalyst. But that’s my point BrownBoy.

SiteCatalyst needs someone trained on it and working it every day to maximize conversion. Where on Google’s anayltics its just nice to look at every once in a while and check out certain stats when it comes to conversion optimization.

But it does all I need it to do, I sue it mainly for traffic sources, I do SEO.

I am actually looking forward someday to getting a certificate or two using an website analytics program like SiteCatalyst.

Cheers,

-Bart

abc

Comment on December 11th, 2008.

hi iam layman when it comes to sitecatalyst and i am using .net application our company has signed up with omniture so i have no choice than implementing it can you help me with some implementation tips and other stuff

Leave a comment

Comments can contain some xhtml. Names and emails are required (emails aren't displayed), url's are optional.