ROI research on RIA

22 Feb 2005 - 8:47am
9 years ago
3 replies
908 reads
Welie, Martijn van
2005

Hi,

I am doing some research RIA applications and one of the aspects that are
often mentioned is the ROI aspect. Macromedia and other pro-RIA 'activists'
claim that when you do RIA instead of HTML sites, the ROI will be higher.

However, I cannot find ANY scientific evidence that supports this
assertion!!!! All documents that I can find make suggestive conclusions in
favour of RIA's and even without comparing the RIA to an alternative.

For example, it is said that when Mini launched their RIA Car Configurator
it boosted their sales so much that their sales targets were exceeded by
25%. That is very suggestive reasoning to me. Where is the evidence? Did
they have a car configurator before? Did they launch a marketing campaign at
the same time?

If RIAs are better than HTML I'd expect stories where an HTML application
was redesigned while keeping the same functionality and the ROI was actually
higher. And of course all other variables kept constant.

Another angle is that I wonder how the ROI is actually improved most. Is it
because of higher revenues or lower costs?

Does anybody know of solid info on ROI stuff and RIAs?

Regards,

Martijn van Welie

Comments

22 Feb 2005 - 11:21am
Manu Sharma
2003

Martijn van Welie:
> Does anybody know of solid info on ROI stuff and RIAs?

Molecular consultants recently claimed that a shopping checkout RIA
solution developed by them dramatically raised conversion rate. Here's
the case study:

TJX Treats Customers to Better Shopping Cart Technology; Increases
Conversion Rates by 50%
http://www.molecular.com/clients/case_studies/tjx.aspx

Molecular Expands User Experience Design Services; Announces New
Solution to Tackle Shopping Cart Abandonment
http://www.molecular.com/news/press_releases/100604_singlescreencheckout.aspx
(a search on Molecular and shopping cart will reveal news stories about
this.)

The 50% increase in conversion claim, if true, is really remarkable for
any shopping site. I wasn't very impressed though when I tried it on TJ
Maxx website about a month ago.[1] It's Flash based and it takes some
time to load.

Manu.

[1] Looks like TJ Maxx has already taken it off the site! Compare the
video on Molecular
http://www.molecular.com/singlescreencheckout/video.aspx with the
actual solution on TJ Maxx http://www.tjmaxx.com/home.jsp

23 Feb 2005 - 11:20am
Welie, Martijn van
2005

>The 50% increase in conversion claim, if true, is really remarkable for any
shopping site. I wasn't very >impressed though when I tried it on TJ Maxx
website about a month ago.[1] It's Flash based and it takes
>some time to load.

Ofcourse it is nice for the developers of this solution to say that there
was a 50% increase in conversion. But how was that measured? What did the
HTML version look like and to what extent was that a bad/good solution? What
was happening with the site around that time? Was the 50% increase a
permanent increase or what it one particular month? And so on and so on?
What if you would build the current design using HTML and page refreshes?

Such statements are pure sales statements if you ask me, not scientific at
least.

I guess it is difficult to do real objective (scientific) assesments.
However, it does make me feel that improve ROI is a very argument for
RIAs....mainly because it depends on so many more things than the RIA
itself...

Martijn

23 Feb 2005 - 12:56pm
Manu Sharma
2003

Martijn:
> it is nice for the developers of this solution to say that there was
a 50% increase in conversion. But how was that measured? What did the
HTML version look like and to what extent was that a bad/good solution?
What was happening with the site around that time? Was the 50% increase
a permanent increase or what it one particular month? And so on and so
on? What if you would build the current design using HTML and page
refreshes?

Exactly. That's why I used the phrases "they claimed..." and "if
true..." One cannot rely on numbers produced by the marketing
department of a firm that provides little information about the
methodology adopted for measurement and other details and that stands
to gain by publishing such results. It's little to do with using a
scientific methodology though and more about bias and transparency.

That said, their solution is impressive and I can imagine a significant
decrease in cart abandonment rates. I also agree however, that the same
functionality could have been implemented using DHTML or what folks at
AP are calling "Ajax technologies." [1]

Manu.

* http://www.adaptivepath.com/

Syndicate content Get the feed