Forum Moderators: buckworks

Message Too Old, No Replies

Good UI v ease of maintaing large data on e-com store

         

Whitey

2:59 am on Dec 24, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



I'm not sure if I'm posting to the right forum, but this is about an e-commerce problem that many may face with their administration and developers. We solved one problem, now have two more.

Background
We had an e-commerce site with a taxonomy of 15,000 locations and 90,000 business names provided in an xml / api feed. The site called up live pricing on specific locations and business names. We stored and cached the location / business names on our server and optimized the scripts to give fast response times for the pricing.

The background issue
The providers of the xml / api data would change the location and business names frequently, which would make maintenance a resource nightmare, where we had to continually monitor changes, such that it didn't get maintained properly and we were consequently optimizing discontinued business names, and not showing other new ones.

User preference filtering was easy, as returned searches on our api could be stored our side giving instant refinement - standard e-commerce stuff.

We built a new site
We then went on to build a new site, which does not store the business names on our side, [ which has increased to over 1million ] thus reducing data maintenance. This makes a call into the xml/api supplier and provides business names with pricing on demand - we no longer SEO business names as they have no ranking benefit to us.

The new problem
By eliminating the problem of maintenance, we have encountered 2 major problems:

- Speed With over 5,000 business names in some locations, the new site has become unworkable - taking up to 90 seconds to return results. Industry standard with the big players is 10-15 seconds.

- Bad UI / Filtering When filtering preferences we have to put in an additional call through the api, because it is not stored on our side. This is a bad UI experience

Question:
Many of you would be involved in the operation of large e-commerce online stores similar to this. For business' operating on a low maintenance budget, what is the answer for good UI versus the ease of managing data on big e-commerce sites?

How can we get the UI/filtering and speed roaring without getting into a data maintenance nightmare?

fathom

3:52 am on Dec 24, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



I can't help on the tangible issues with speeding things up or improving UX but I don't believe you're having ranking issues from either, or. If this is part of your concern.

IMHO ranking begins below position #1000, not only in the top 10, contrary to common belief. When Google denotes 2.3 million results, everything above 2.3 million is ranked and being in the top #1000 is in the top 0.004%.

That said, if you are in the top #1000 some factors that Google eludes, likely had their positive impact already.

Whitey

4:30 am on Dec 24, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



@fathom - thanks. The question isn't an SEO one at this stage. Sure the consequences can play into it, but that's not where I'm coming from.

I'm hoping some folks who manage/have knowledge of big data stores might chime in at some stage for help.

fathom

5:19 am on Dec 24, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



Ya... I noticed you never mentioned that but like I hinted I'm not a programmer, designer, or anything else that might help you other than bumping the thread with a little semi-unrelated BS. :-)

Whatever gets the job done!

Merry Christmas

[edited by: fathom at 6:00 am (utc) on Dec 24, 2015]

Whitey

5:37 am on Dec 24, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



Merry Christmas

@fathom - Likewise. And also to anyone else who chimes in before Christmas !

piatkow

11:07 am on Dec 26, 2015 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



As an old mainframe dinosaur I would suggest going back to basics with some old fashioned entity- relationship and process modelling to work out the optimal structure for your data.