Jump to content

  • Log in with Facebook Log in with Twitter Log In with Google      Sign In   
  • Create Account

Subscribe to HRA Now!

 



Are you a Google Analytics enthusiast?

Share and download Custom Google Analytics Reports, dashboards and advanced segments--for FREE! 

 



 

 www.CustomReportSharing.com 

From the folks who brought you High Rankings!



Photo
- - - - -

Ecommerce Site Search Tip


  • Please log in to reply
42 replies to this topic

#16 adibranch

adibranch

    HR 5

  • Active Members
  • PipPipPipPipPip
  • 332 posts

Posted 20 May 2009 - 10:13 AM

QUOTE(Randy @ May 20 2009, 09:04 AM) View Post
You may also want to consider usability concerns adibranch.

I had something similar happen with one of my sites when I first added a search tool to it.

yeah true i think it makes a difference, but in both my circumstances stated above the search is a standard quick search box, on every page, in the header, above the main categories, against a contrasting background similar to what you suggest.. still only 1-4%. I'm not saying this is the same for everyone of course.

#17 BBCoach

BBCoach

    HR 5

  • Moderator
  • 402 posts

Posted 20 May 2009 - 01:08 PM

QUOTE
NASA - However, do you not wonder how come you get so many idiots to your site who are incapable of searching or finding the product they want without this tool?
They're not idiots, well mostly they're not. There are several factors that contribute, but the number one factor is a person is new to our products. Number two is that our industry is very complicated to match our products to use/accessorize with their products. However, another significant factor is the average reading ability of Americans is at the 8th grade level or lower (which means they can't spell very well), not to mention English may be their second language. Bottom-line is I still want to SELL to them by removing any barriers and I'm doing that by making it easier to quickly find our products. Even if I were selling products to a group of PhDs, I still want third grade drop-outs or the Gerry Whites of the world spending his/her money in my store.

QUOTE
adibranch - agreed.. either that OR there is something fundamentally wrong with your navigaion, that customers resort to the search box
Yep. It's a big problem. It used to not be, but the owner of the company heard/read where users should only have to click no more than two or three times before getting to a product. That's generally a true statement. However, with the complicated aspect of our product line combined with there being many hundreds of similar but different products in most of the categories that philosophy makes it more difficult for them to quickly find a left-handed product thru navigation. Why? Because there are hundreds more right-handed products in the same category (which results in many more page clicks to find the left-handed stuff). That owner design change is exactly why I'm focusing on making our search function as user-friendly as possible. Perhaps one day the light will shine on the owner, but until then I'm making search as responsive and intuitive as possible.

QUOTE
It's purely a guess on my part, but I have a hunch this might be an important factor in how well it's being received right out of the box.
Good guess Randy. We intentionally built it that way because of the 2+ million/month SE referrals. Why not let the SE's train them to better use our search. LOL! ! ! hysterical.gif

QUOTE
IMO it's always a good idea when starting something new to at least try to model it (visually, if not on the back end) after something people are already used to. Simply because of previous conditioning it makes a tool easier for people to use, meaning they'll use it more often.
Couldn't have said it better.

QUOTE
Gerry White - I am interested in one thing, which analytics package did you use ? I typically use GA, but you can only do what your talking about with GA with some intelligent customisation....
Believe it or not we use GA, but not for analyzing our internal search. We use load-balanced search appliances and they have reporting capabilities. BTW, we tried to use ClickTracks, but it couldn't handle our 2+ gig log files very well. We had the server portion running on the latest quad-processor with 8 gig of ram and a tera-byte harddrive on a gigbyte network and it was simply not able to handle that much info very well. Very, very slow performance. However, GA does an amazing job for the price and it's within +/- 1% accuracy from what we've been able to determine. Most of the inaccuracy results from the latency of the updates.

Edited by BBCoach, 20 May 2009 - 01:48 PM.


#18 NASA

NASA

    HR 4

  • Active Members
  • PipPipPipPip
  • 183 posts

Posted 21 May 2009 - 04:00 AM

Warning... next comment is a joke for those who get offended easily....

BBCoach how can you possibly say these two things in the same sentence...
QUOTE
They're not idiots, well mostly they're not.

and then say

the average reading ability of Americans is at the 8th grade level or lower (which means they can't spell very well), not to mention English may be their second language.


double negative me thinks hysterical.gif


#19 adibranch

adibranch

    HR 5

  • Active Members
  • PipPipPipPipPip
  • 332 posts

Posted 21 May 2009 - 04:27 AM

i dont think all web users are idiots, but i always assume they are biggrin.gif

#20 NASA

NASA

    HR 4

  • Active Members
  • PipPipPipPip
  • 183 posts

Posted 21 May 2009 - 04:33 AM

lol.gif yup , just like driving, everyone else on the road is an idiot!

#21 Gerry White

Gerry White

    HR 2

  • Active Members
  • PipPip
  • 48 posts
  • Location:UK

Posted 21 May 2009 - 06:25 AM

@ adibranch - try looking at only people who convert, and see search usage there, I will bet money its a lot higher (not much money, though...)

#22 BBCoach

BBCoach

    HR 5

  • Moderator
  • 402 posts

Posted 21 May 2009 - 09:40 AM

QUOTE
NASA - BBCoach how can you possibly say these two things in the same sentence...
Easy. They're not idiots when spending money at my store. They're loyal, valued and trusted customers. thumbup1.gif

BTW, it's paragraph and not sentence. cheers.gif

#23 NASA

NASA

    HR 4

  • Active Members
  • PipPipPipPip
  • 183 posts

Posted 21 May 2009 - 09:49 AM

touche! appl.gif

#24 BBCoach

BBCoach

    HR 5

  • Moderator
  • 402 posts

Posted 21 May 2009 - 01:00 PM

One other point about internal search engines for all those concerned about keywords and PPCs. I've said this in a different thread before, but it bears repeating. Site searches will open the treasure trove of the "how and what" users are looking for and you can start making adjustments to page titles and copy on your site accordingly (to better match what's being typed in SEs). In addition, those PPCs will be more targeted to your world-wide audience resulting in a bigger ROI. So whether or not you decide to incorporate auto-complete, the real important thing is to have a good website search feature so that you can analyze how users phrase the search and what they're searching for while comparing those searches to the ones done at SEs.

#25 NASA

NASA

    HR 4

  • Active Members
  • PipPipPipPip
  • 183 posts

Posted 22 May 2009 - 03:48 AM

Do you beleive generally what ever someone types in your search box is what they will be typing in the search engines.

It's not how I generally use an SE vs site search, for example if I wanted some computer components, I'd use the SE's to search for websites that sell 'Cheap Computer Components" , but when i visit the site then i would use the search to find the CPU or graphics card i was looking for.

I would not look for product specific stuff in the SE search, as a rule.

But I know i'm not normal lol.gif so I can see where you're coming from, I always find it hard to second guess what my visitor wants or thinks like, I think like me not them!

What type of site search vs GOAN kwd research have you done that could give an indication on match %

#26 BBCoach

BBCoach

    HR 5

  • Moderator
  • 402 posts

Posted 22 May 2009 - 11:12 AM

QUOTE
NASA - Do you beleive generally what ever someone types in your search box is what they will be typing in the search engines.
No. That's why I stated:
QUOTE
the real important thing is to have a good website search feature so that you can analyze how users phrase the search and what they're searching for while comparing those searches to the ones done at SEs


What I've found is that there are many types of searchers (duh), but for my industry there are two large groups of users coming from SEs that stand out. One group will use more general words (phrased like you described along with using a manufacturer name) than they do on the internal site search where they use fewer words and more accurately named products. In my opinion, that's why I needed to help them spell the products as accurately as possible when conducting a search to get them there with as few frustrations as possible. It seems to be working nicely.

The other SE group specifically looks for products using manufacturer part numbers and/or exact product names and typically don't do as many "related" searches because the pages that they're hitting are exactly what they're looking for.

QUOTE
NASA - What type of site search vs GOAN kwd research have you done that could give an indication on match %
With 65,000+ SE referrals per day I would go nuts trying to use GA's kwd stats. Here's my secret sauce, but it's not for those that are allergic to zeros and ones. The search appliances record the query, datetime, and most importantly the IP address of all site searches. The web server log files likewise record the HTTPRefer, datetime and the IP. I wrote a program to parse the log files extracting the SE query from the HTTPRefer string, the datetime and IP sucking that info into a database. I then export the data (I don't have access to the raw logs) from the search appliances and upload that data to the same table as the other data. There's a column that identifies whether the data is from an SE or the site. I now have the capability to view something like this per user.

Query / PageHit / IP / DateTime / Source
"cheap fishing pole" / www.mysite.com/1234.aspx / 123.123.0.1 / 2009-05-22 09:08:04.417 / Google
"ugly stick" / www.mysite.com/3456.aspx / 123.123.0.1 / 2009-05-22 09:08:10.214 / Site

Originally, I analyzed the data daily. A bunch of work, but initially very important to set base-lines. Today, I do this once per week. What's exceedingly nice about my system is that I can easily pattern out each user individually, and also easily group them in a myriad of ways to get a better understanding of how/what the masses are searching for. This system also is used to track marketing campaigns (email/PPC) which helps to analyze their effectiveness. I can run a report on a kwd or phrase and group them to determine the most popular way it's spelled, other words used with kwd and the order of word usage, which usage converts best, revenue by kwd, popularity of product line (today, this month, this quarter and this year) and a hole bunch more gold star SEO/marketing analysis. Another mash up is to load all of the web server log file data in another table and join these two tables to view site usage info

QUOTE
NASA - But I know i'm not normal
Maybe not normal, but definitely in a large like minded group of searchers. biggrin.gif

#27 1dmf

1dmf

    Keep Asking, Keep Questioning, Keep Learning

  • Active Members
  • PipPipPipPipPipPipPip
  • 2,167 posts
  • Location:Worthing - England

Posted 22 May 2009 - 12:03 PM

cool, so where can I download this log file analysis tool wink1.gif

Would you say a site search is ONLY beneficial when selling products? or whre some type of category or itemisation is available.

Ho usefull is a site search for a site that only offers information, would those pages simply be found via the SE search anyway, would a sime G! siteseach facility be best when dealing with info only?

And how beneficial would you say it is to generate a product specific site map (XML) to each potential product you sell so you can get those pages indexed separately?

I guess that way you get as many possible bites of the cherry when they search the SE, and then back it up with a comprehensive site search facility.

#28 BBCoach

BBCoach

    HR 5

  • Moderator
  • 402 posts

Posted 22 May 2009 - 12:43 PM

QUOTE
1dmf - cool, so where can I download this log file analysis tool
You're funny 1dmf hysterical.gif
I will tell you that I use Log Parser 2.0 from MS with C#

QUOTE
1dmf - Would you say a site search is ONLY beneficial when selling products? or whre some type of category or itemisation is available. Ho usefull is a site search for a site that only offers information, would those pages simply be found via the SE search anyway, would a sime G! siteseach facility be best when dealing with info only?
Is there another reason? Doh! My hobby site needs a good kwd search to locate info pages, but that entire site is built to one day be a marketing/selling rock star. So yeah it's beneficial if you care about users finding your info and perhaps the relationships that may exist between the different pages, but a good site search is typically not a free plugin. Yes, the G search plugin will work, but you probably won't get the granularity of my system. It all depends on how deep your pocket is. Our search appliances cost $9,000 each and my time to write the applications is many times more than they cost.

QUOTE
1dmf - And how beneficial would you say it is to generate a product specific site map (XML) to each potential product you sell so you can get those pages indexed separately?
I either don't understand what you're asking or I would answer that all of my pages are accessible to the bots and I don't use sitemaps XML or otherwise.

Edited by BBCoach, 22 May 2009 - 02:04 PM.


#29 NASA

NASA

    HR 4

  • Active Members
  • PipPipPipPip
  • 183 posts

Posted 26 May 2009 - 04:02 AM

QUOTE
QUOTE
1dmf - And how beneficial would you say it is to generate a product specific site map (XML) to each potential product you sell so you can get those pages indexed separately?
I either don't understand what you're asking or I would answer that all of my pages are accessible to the bots and I don't use sitemaps XML or otherwise.


What that means is if you had a program that facilitated the dynamic display of products via a query string, is it advisable to cretae an XML sitemap for each product, by using myscript.cgi?Product=Guns , etc..

I guess you already have a separate page for each product so this is not relevant.

QUOTE
It all depends on how deep your pocket is. Our search appliances cost $9,000 each and my time to write the applications is many times more than they cost.


Jeez, what does this search do, do you have a demo I can check out, seems like alot of money, does it scan page content or query a backend DB?

#30 BBCoach

BBCoach

    HR 5

  • Moderator
  • 402 posts

Posted 26 May 2009 - 10:50 AM

QUOTE
Jeez, what does this search do, do you have a demo I can check out, seems like alot of money, does it scan page content or query a backend DB?
Just Google for search appliance and check them out. This one will crawl the site (which we want for many reasons) or it can be pointed to a dB. The cost is directly associated to the number of pages you want crawled. The more pages, the more money.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

We are now a read-only forum.
 
No new posts or registrations allowed.