Current time: 10-18-2017, 06:28 AM Hello There, Guest! (LoginRegister)

Post Reply 
 
Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Never mind the webpages, how do you fix the business??
01-29-2013, 03:08 AM
Post: #1
Exclamation Never mind the webpages, how do you fix the business??
Lots of the posts on these forums ask "how can i improve my test score", or "how can i make my website faster". I know what needs to be done to do those things, its quite trivial (from a technical perspective).

What i don't know how to do, is convince the business (those with the fitted carpets and potted plants) that *they* are killing the website. As conversion decreases, they feel compelled to add yet more 3rd party tracking tags. I know it's crazy. Its like saying, if i weigh my self more often I'll lose more weight!

Its got so bad that our end users actually have to down load 52% web page and 48% tracking tags!! (number of GETs, amount of data Kbytes)

Its so bad in fact, that the size of the cookies in the client request now fragments the packet. So a simple get for a 1x1 tracking pixel is fragmented across two packets. Not only that, but 1.3k was transmitted to fetch ~40 bytes!!

So, please tell me. How have you had this conversation, how did go down? Did you get far with quoting facts and figures from similar websites as yours?

Did you use the industry arguments: "Amazon: 100ms faster = 1% revenue increase", etc etc?

I look forward to your responses.

many thanks,

@hsiboy
Find all posts by this user
Quote this message in a reply
01-29-2013, 03:14 AM
Post: #2
RE: Never mind the webpages, how do you fix the business??
Are the tracking tags in the way of the user experience? If so, that's the first problem and should be solvable through engineering (though the cookies sound like they are out of control).

Do you have any ability to convince them to do an A/B test and measure the actual impact on business metrics? Joshua Bixby had a pretty good post/video on how to extract some of it even without running a test: http://www.webperformancetoday.com/2011/...e-website/

I assume that they are getting business value from each of the tracking beacons so it comes down to figuring out if they can get more business value without them or if the site can be engineered in such a way as to minimize the impact.
Visit this user's website Find all posts by this user
Quote this message in a reply
01-29-2013, 03:18 AM
Post: #3
RE: Never mind the webpages, how do you fix the business??
Hi Pat,

one of the problems *is* the A/B testing, supplied by Maxymiser. They inject so much code, and cookies.

I have used WPT to great effect (using scripting and video comparison) in the past by demonstrating how fast the site could be if 3rd party tags were removed. Still, they remain unconvinced.

The fact that they have tealeaf, GA and Webtrends sounds alarm bells for me, a deep seated FUD.

@hsiboy
Find all posts by this user
Quote this message in a reply
01-29-2013, 04:54 AM
Post: #4
RE: Never mind the webpages, how do you fix the business??
Hi Stu,

I'm aware that I'm not a disinterested party but this is one of the situations where external consultants can help, whether it's someone like me or someone like Intechnica.

For some reason the views of external consultants often seen to carry more weight than those of internal staff (I've been there and know how frustrating it can be).

If you want to convince people to do things differently my advice is you've got to give them the business benefits e.g. use RUM to determine if people with faster experiences drive more revenue.

Unfortunately it's not uncommon to see masses of 3rd party tags on pages, many of them collecting similar sorts of data and in someways tag management services seem to be making this worse (out of sight out of mind I suspect)

Perhaps another place to start might be to establish what data it being gathered by each of the third-party tags and what the business is using it for, work out the overlaps and sell it to the rest of the company as a cost saving exercise i.e. they don't need to pay for both services and so removing one saves money and page weight.

Cheers

Andy

Andy

Using WebPageTest - http://usingwpt.com/
Visit this user's website Find all posts by this user
Quote this message in a reply
01-29-2013, 06:43 AM (This post was last modified: 01-29-2013 06:45 AM by Chris Stark.)
Post: #5
RE: Never mind the webpages, how do you fix the business??
I would look at putting a dns or server log based analytics in place and comparing that with one javascript based analytics to see how far your drop off is from load times. The bigger the differential, the worse the problem is. If you can definitely say between dns request to your site and analytics tag triggering you're seeing an x% loss * income then you have a pretty strong argument for real ROI for making things quicker.

I also like using Torbit.com for RUM, if you setup the premium version you can actually correlate load times with conversion ratios. This further helps to put real numbers to the I just want things to be faster argument.

After I have these two things I can usually go to a c-level decision maker, say to them if I cut your site speed from 10 seconds to 3 seconds you're going to make X amount more money, it's going to cost you X amount to program the improvements and you're going to make a 500% return on investment in the first year. Even for mom and pop type places it's pretty simple to understand, here's my research, spend things in development, make this return, do you want to make more money? 9 times out of 10 it works for me. Some people are hopelessly dense and don't care thinking the internet is a waste of time, you're never going to convince those people of anything.

I have an example on my blog about how Walmart tackled the same question http://www.digitalmgmt.com/4-awesome-sli...lmart-com/
Visit this user's website Find all posts by this user
Quote this message in a reply
01-29-2013, 08:04 AM (This post was last modified: 01-29-2013 08:16 PM by hsiboy.)
Post: #6
RE: Never mind the webpages, how do you fix the business??
Hi Chris,

Thanks for the link to the wallmart case study, I think awesome slide #3 was what i was looking for, such a simple metric, yet so powerful. If the penny doesn't drop after seeing that, then there is no hope.

@hsiboy

(01-29-2013 06:43 AM)Chris Stark Wrote:  I would look at putting a dns or server log based analytics in place and comparing that with one javascript based analytics to see how far your drop off is from load times. The bigger the differential, the worse the problem is. If you can definitely say between dns request to your site and analytics tag triggering you're seeing an x% loss * income then you have a pretty strong argument for real ROI for making things quicker.

I also like using Torbit.com for RUM, if you setup the premium version you can actually correlate load times with conversion ratios. This further helps to put real numbers to the I just want things to be faster argument.

After I have these two things I can usually go to a c-level decision maker, say to them if I cut your site speed from 10 seconds to 3 seconds you're going to make X amount more money, it's going to cost you X amount to program the improvements and you're going to make a 500% return on investment in the first year. Even for mom and pop type places it's pretty simple to understand, here's my research, spend things in development, make this return, do you want to make more money? 9 times out of 10 it works for me. Some people are hopelessly dense and don't care thinking the internet is a waste of time, you're never going to convince those people of anything.

I have an example on my blog about how Walmart tackled the same question http://www.digitalmgmt.com/4-awesome-sli...lmart-com/


Hi Andy,

Yeah, I know. I have their ear for sure, but there is something else, something I'm missing perhaps. Its black and white to me, and maybe that's part of the problem. Because its so clear to me, I'm not demonstrating it clearly enough to them.

RUM is on the table, but honestly, Gomez has done so much damage...

@hsiboy


(01-29-2013 04:54 AM)andydavies Wrote:  Hi Stu,

I'm aware that I'm not a disinterested party but this is one of the situations where external consultants can help, whether it's someone like me or someone like Intechnica.

For some reason the views of external consultants often seen to carry more weight than those of internal staff (I've been there and know how frustrating it can be).

If you want to convince people to do things differently my advice is you've got to give them the business benefits e.g. use RUM to determine if people with faster experiences drive more revenue.

Unfortunately it's not uncommon to see masses of 3rd party tags on pages, many of them collecting similar sorts of data and in someways tag management services seem to be making this worse (out of sight out of mind I suspect)

Perhaps another place to start might be to establish what data it being gathered by each of the third-party tags and what the business is using it for, work out the overlaps and sell it to the rest of the company as a cost saving exercise i.e. they don't need to pay for both services and so removing one saves money and page weight.

Cheers

Andy
Find all posts by this user
Quote this message in a reply
Post Reply 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)