Loading ...
Sorry, an error occurred while loading the content.

Re: [webanalytics] Relationship between page load time and drop-off rate?

Expand Messages
  • Apurba Sen
    Hey Michelle, Reminds me of 8 second rule :) i.e if users have to wait longer than 8 seconds to download a Web page, they will go elsewhere.So any page load
    Message 1 of 14 , Jul 1, 2007
    • 0 Attachment
      Hey Michelle,

      Reminds me of 8 second rule :) i.e if users have to wait longer than 8
      seconds to download a Web page, they will go elsewhere.So any page load time
      greater than or equal to 8 sec should ideally mean 100 % drop off :)

      Some related links:

      http://en.wikipedia.org/wiki/Network_performance
      http://www.pcmag.com/encyclopedia_term/0,2542,t=8-second+rule&i=37197,00.asp

      Regds
      Apurba


      On 6/28/07, Michelle Runyan <michelle.runyan@...> wrote:
      >
      > Has anyone done analysis or seen data that shows the relationship
      > between load time and drop-off rate? I have seen a few reports that
      > present what people say they are ok with (for example
      > http://www.akamai.com/dl/reports/Site_Abandonment_Final_Report.pdf),
      > but the real value is in what the data says.
      >
      > We have tried analyzing our raw data and can't find a wide enough
      > spectrum of load times to derive that X% drop off at 3-4 seconds and X%
      > drop off at 5-6 seconds ...
      >
      > If you have any thought or ideas please let me know.
      >
      > Thanks,
      > Michelle
      >
      >
      >



      --
      Cheers!!
      Apurba Sen
      Product Development
      रेडीफ़.कॉम इंडिया लिमिटेड
      NASDAQ:REDF
      web:www.rediff.com
      http://apurba.rediffiland.com
      Phone :91-22 2444 9144 * 313
      GSM :91-98670- 85718

      "Champions aren't made in the gyms. Champions are made from something they
      have deep inside them -- a desire, a dream, a vision."

      --Muhammad Ali


      [Non-text portions of this message have been removed]
    • Jared M. Spool
      ... Except the 8-second rule is a complete fabrication with no basis in actual user behavior. http://www.uie.com/articles/download_time/ Turns out that users
      Message 2 of 14 , Jul 1, 2007
      • 0 Attachment
        On Jul 1, 2007, at 5:49 AM, Apurba Sen wrote:

        > Reminds me of 8 second rule :) i.e if users have to wait longer than 8
        > seconds to download a Web page, they will go elsewhere.So any page
        > load time
        > greater than or equal to 8 sec should ideally mean 100 % drop off :)

        Except the "8-second rule" is a complete fabrication with no basis in
        actual user behavior.

        http://www.uie.com/articles/download_time/

        Turns out that users will wait for very long times, *if* they think
        they are going to complete their tasks. Not only will they wait,
        they'll rate the site as being fast.

        Jared

        Jared M. Spool
        User Interface Engineering
        510 Turnpike St., Suite 102, North Andover, MA 01845
        e: jspool@... p: +1 978 327 5561
        http://uie.com Blog: http://uie.com/brainsparks



        [Non-text portions of this message have been removed]
      • Steve
        ... That s not what that article states. The article is talking about a hypothesis of sites with faster download times would be more usable than slower
        Message 3 of 14 , Jul 1, 2007
        • 0 Attachment
          On 7/2/07, Jared M. Spool <jspool@...> wrote:
          > On Jul 1, 2007, at 5:49 AM, Apurba Sen wrote:
          > > Reminds me of 8 second rule :) i.e if users have to wait longer than 8
          > > seconds to download a Web page, they will go elsewhere.So any page
          > > load time
          > > greater than or equal to 8 sec should ideally mean 100 % drop off :)
          >
          > Except the "8-second rule" is a complete fabrication with no basis in
          > actual user behavior.
          >
          > http://www.uie.com/articles/download_time/
          >
          > Turns out that users will wait for very long times, *if* they think
          > they are going to complete their tasks. Not only will they wait,
          > they'll rate the site as being fast.


          That's not what that article states.
          The article is talking about a hypothesis of "sites with faster
          download times would be more usable than slower sites".

          Nothing to do with abandonment - which was the original question.

          While I don't have anything concrete to hand, our own internal
          analytics & testing has shown that when we decreased the speed of each
          page load from ~ 5-7 seconds down to sub second, we did get a quite
          respectable increase in traffic. Something like 20-50% more from
          memory.


          Being able to survive all the idiots who would spider us into oblivion
          was an obvious benefit to usability and reducing abandonment for our
          *other* users. ;-)


          Cheers!
          - Steve
        • Tim Wilson
          Hmmm. These articles both seem pretty thin when it comes to any data to validate the theory. To me, it’s always seemed like page load time would be a much
          Message 4 of 14 , Jul 2, 2007
          • 0 Attachment
            Hmmm. These articles both seem pretty thin when it comes to any data to validate the theory.



            To me, it’s always seemed like page load time would be a much bigger deal for sites where there are lots of other alternative sites that can be expected to offer largely similar content: cnn.com vs. msnbc.com, lowes.com vs. homedepot.com, espn.com vs. sportsillustrated.cnn.com. In a B2C world, for sites that I’m going to get into a pattern of visiting, persistently slow load times are going to drive me to look elsewhere. In the B2B world – higher price point products/services, more niches, longer buying cycle – it seems like small changes in page load times would not have much of an impact. I would happily wait for 10 seconds for a www.webanalyticsassociation.org <http://www.webanalyticsassociation.org/> page to load (not that it takes that long!).



            It’s an interesting challenge for folks on the IT side of the house – new functionality rolling out that may require more server crunching to deliver it. How much $$$ should they spend to maintain or reduce load times in the light of new functionality (that, presumably, adds more value for the visitor)?



            (Sorry, Michelle – I don’t have an answer by any means…but I’d love to see some insights on this from this group!)



            Tim



            ________________________________

            From: webanalytics@yahoogroups.com [mailto:webanalytics@yahoogroups.com] On Behalf Of Apurba Sen
            Sent: Sunday, July 01, 2007 4:50 AM
            To: webanalytics@yahoogroups.com
            Subject: Re: [webanalytics] Relationship between page load time and drop-off rate?



            Hey Michelle,

            Reminds me of 8 second rule :) i.e if users have to wait longer than 8
            seconds to download a Web page, they will go elsewhere.So any page load time
            greater than or equal to 8 sec should ideally mean 100 % drop off :)

            Some related links:

            http://en.wikipedia.org/wiki/Network_performance <http://en.wikipedia.org/wiki/Network_performance>
            http://www.pcmag.com/encyclopedia_term/0,2542,t=8-second+rule&i=37197,00.asp <http://www.pcmag.com/encyclopedia_term/0,2542,t=8-second+rule&i=37197,00.asp>

            Regds
            Apurba

            On 6/28/07, Michelle Runyan <michelle.runyan@... <mailto:michelle.runyan%40ni.com> > wrote:
            >
            > Has anyone done analysis or seen data that shows the relationship
            > between load time and drop-off rate? I have seen a few reports that
            > present what people say they are ok with (for example
            > http://www.akamai.com/dl/reports/Site_Abandonment_Final_Report.pdf <http://www.akamai.com/dl/reports/Site_Abandonment_Final_Report.pdf> ),
            > but the real value is in what the data says.
            >
            > We have tried analyzing our raw data and can't find a wide enough
            > spectrum of load times to derive that X% drop off at 3-4 seconds and X%
            > drop off at 5-6 seconds ...
            >
            > If you have any thought or ideas please let me know.
            >
            > Thanks,
            > Michelle
            >
            >
            >

            --
            Cheers!!
            Apurba Sen
            Product Development
            रेडीफ़.कॉम इंडिया लिमिटेड
            NASDAQ:REDF
            web:www.rediff.com
            http://apurba.rediffiland.com <http://apurba.rediffiland.com>
            Phone :91-22 2444 9144 * 313
            GSM :91-98670- 85718

            "Champions aren't made in the gyms. Champions are made from something they
            have deep inside them -- a desire, a dream, a vision."

            --Muhammad Ali

            [Non-text portions of this message have been removed]





            [Non-text portions of this message have been removed]
          • Apurba Sen
            Hey Tim!! Let me answer your query from marketing standpoint,to be specific through Brand Loyalty vs Satisfaction. Using Jones & Sasser framework, customers
            Message 5 of 14 , Jul 2, 2007
            • 0 Attachment
              Hey Tim!!

              Let me answer your query from marketing standpoint,to be specific through
              Brand Loyalty vs Satisfaction.

              Using Jones & Sasser framework, customers can be plotted in 4 quadrants of
              Loyalty [Y Axis] vs Satisfaction [X Axis]

              A.Loyalty Low -Satisfaction Low :Terrorist

              B.Loyalty Low-Satisfaction High :Mercenaries

              C.Loyalty High-Satisfaction Low :Hostages

              D.Loyalty High-Satisfaction High :Apostles


              Customers at the lowest and highest ends of the satisfaction scale tend to
              have intense feelings about a brand and its products /services. Terrorists
              actively attack the brand telling others not to buy from the company.
              Apostles are satisfied and loyal and talk favourably about the brand.

              Using your examples...

              Terrorist or Mercenaries:Page load time would be a somehow bigger deal where
              plenty of alternative sites are available.They can be found in abundance in
              [cnn.com vs. msnbc.com, lowes.com vs. homedepot.com, espn.com vs.
              sportsillustrated.cnn.com]

              Apostles or Hostages:B2B world ,niches even
              www.webanalyticsassociation.org.Thinking abt "Hostages" in this group..what
              are the other options avl with them :-)

              Hope this was enlightening enough.

              Reducing page delivery time should always be a goal while designing the
              page/information layout e.g. progressive download with relatively static
              content being cached.

              Regds

              Apurba Sen

              Related Article :

              Jones, T., & Sasser, E., 'Why Satisfied Customers Defect' - *Harvard
              Business Review*, Nov-Dec 1995







              On 7/2/07, Tim Wilson <twilson@...> wrote:
              >
              > Hmmm. These articles both seem pretty thin when it comes to any data to
              > validate the theory.
              >
              >
              >
              > To me, it's always seemed like page load time would be a much bigger deal
              > for sites where there are lots of other alternative sites that can be
              > expected to offer largely similar content: cnn.com vs. msnbc.com,
              > lowes.com vs. homedepot.com, espn.com vs. sportsillustrated.cnn.com. In a
              > B2C world, for sites that I'm going to get into a pattern of visiting,
              > persistently slow load times are going to drive me to look elsewhere. In the
              > B2B world – higher price point products/services, more niches, longer buying
              > cycle – it seems like small changes in page load times would not have much
              > of an impact. I would happily wait for 10 seconds for a
              > www.webanalyticsassociation.org <http://www.webanalyticsassociation.org/>
              > page to load (not that it takes that long!).
              >
              >
              >
              > It's an interesting challenge for folks on the IT side of the house – new
              > functionality rolling out that may require more server crunching to deliver
              > it. How much $$$ should they spend to maintain or reduce load times in the
              > light of new functionality (that, presumably, adds more value for the
              > visitor)?
              >
              >
              >
              > (Sorry, Michelle – I don't have an answer by any means…but I'd love to see
              > some insights on this from this group!)
              >
              >
              >
              > Tim
              >
              >
              >
              > ________________________________
              >
              > From: webanalytics@yahoogroups.com <webanalytics%40yahoogroups.com>[mailto:
              > webanalytics@yahoogroups.com <webanalytics%40yahoogroups.com>] On Behalf
              > Of Apurba Sen
              > Sent: Sunday, July 01, 2007 4:50 AM
              > To: webanalytics@yahoogroups.com <webanalytics%40yahoogroups.com>
              > Subject: Re: [webanalytics] Relationship between page load time and
              > drop-off rate?
              >
              >
              >
              > Hey Michelle,
              >
              > Reminds me of 8 second rule :) i.e if users have to wait longer than 8
              > seconds to download a Web page, they will go elsewhere.So any page load
              > time
              > greater than or equal to 8 sec should ideally mean 100 % drop off :)
              >
              > Some related links:
              >
              > http://en.wikipedia.org/wiki/Network_performance <
              > http://en.wikipedia.org/wiki/Network_performance>
              >
              > http://www.pcmag.com/encyclopedia_term/0,2542,t=8-second+rule&i=37197,00.asp<
              > http://www.pcmag.com/encyclopedia_term/0,2542,t=8-second+rule&i=37197,00.asp>
              >
              >
              > Regds
              > Apurba
              >
              > On 6/28/07, Michelle Runyan <michelle.runyan@...<michelle.runyan%40ni.com><mailto:
              > michelle.runyan%40ni.com> > wrote:
              > >
              > > Has anyone done analysis or seen data that shows the relationship
              > > between load time and drop-off rate? I have seen a few reports that
              > > present what people say they are ok with (for example
              > > http://www.akamai.com/dl/reports/Site_Abandonment_Final_Report.pdf <
              > http://www.akamai.com/dl/reports/Site_Abandonment_Final_Report.pdf> ),
              > > but the real value is in what the data says.
              > >
              > > We have tried analyzing our raw data and can't find a wide enough
              > > spectrum of load times to derive that X% drop off at 3-4 seconds and X%
              > > drop off at 5-6 seconds ...
              > >
              > > If you have any thought or ideas please let me know.
              > >
              > > Thanks,
              > > Michelle
              > >
              > >
              > >
              >
              > --
              > Cheers!!
              > Apurba Sen
              > Product Development
              > रेडीफ़.कॉम इंडिया लिमिटेड
              > NASDAQ:REDF
              > web:www.rediff.com
              > http://apurba.rediffiland.com <http://apurba.rediffiland.com>
              > Phone :91-22 2444 9144 * 313
              > GSM :91-98670- 85718
              >
              > "Champions aren't made in the gyms. Champions are made from something they
              >
              > have deep inside them -- a desire, a dream, a vision."
              >
              > --Muhammad Ali
              >
              > [Non-text portions of this message have been removed]
              >
              >
              >
              >
              > [Non-text portions of this message have been removed]
              >
              >
              >



              --
              Cheers!!
              Apurba Sen
              Product Development
              रेडीफ़.कॉम इंडिया लिमिटेड
              NASDAQ:REDF
              web:www.rediff.com
              http://apurba.rediffiland.com
              Phone :91-22 2444 9144 * 313
              GSM :91-98670- 85718

              "Champions aren't made in the gyms. Champions are made from something they
              have deep inside them -- a desire, a dream, a vision."

              --Muhammad Ali


              [Non-text portions of this message have been removed]
            • justo_ibarra
              Hi everybody: I just read a recent article about the 8 seconds rule. Altough develops a more theoretical approach on the subject maybe results helpfull.
              Message 6 of 14 , Jul 3, 2007
              • 0 Attachment
                Hi everybody:

                I just read a recent article about the 8 seconds rule. Altough
                develops a more theoretical approach on the subject maybe results
                helpfull.

                http://www.silverpop.com/landing/emarketer/24.html

                You can download it for free.

                Regards,

                Justo Ibarra.
                jibarra@...

                --- In webanalytics@yahoogroups.com, "Apurba Sen" <sensession@...>
                wrote:
                >
                > Hey Tim!!
                >
                > Let me answer your query from marketing standpoint,to be specific
                through
                > Brand Loyalty vs Satisfaction.
                >
                > Using Jones & Sasser framework, customers can be plotted in 4
                quadrants of
                > Loyalty [Y Axis] vs Satisfaction [X Axis]
                >
                > A.Loyalty Low -Satisfaction Low :Terrorist
                >
                > B.Loyalty Low-Satisfaction High :Mercenaries
                >
                > C.Loyalty High-Satisfaction Low :Hostages
                >
                > D.Loyalty High-Satisfaction High :Apostles
                >
                >
                > Customers at the lowest and highest ends of the satisfaction scale
                tend to
                > have intense feelings about a brand and its products /services.
                Terrorists
                > actively attack the brand telling others not to buy from the
                company.
                > Apostles are satisfied and loyal and talk favourably about the
                brand.
                >
                > Using your examples...
                >
                > Terrorist or Mercenaries:Page load time would be a somehow bigger
                deal where
                > plenty of alternative sites are available.They can be found in
                abundance in
                > [cnn.com vs. msnbc.com, lowes.com vs. homedepot.com, espn.com vs.
                > sportsillustrated.cnn.com]
                >
                > Apostles or Hostages:B2B world ,niches even
                > www.webanalyticsassociation.org.Thinking abt "Hostages" in this
                group..what
                > are the other options avl with them :-)
                >
                > Hope this was enlightening enough.
                >
                > Reducing page delivery time should always be a goal while designing
                the
                > page/information layout e.g. progressive download with relatively
                static
                > content being cached.
                >
                > Regds
                >
                > Apurba Sen
                >
                > Related Article :
                >
                > Jones, T., & Sasser, E., 'Why Satisfied Customers Defect' - *Harvard
                > Business Review*, Nov-Dec 1995
                >
                >
                >
                >
                >
                >
                >
                > On 7/2/07, Tim Wilson <twilson@...> wrote:
                > >
                > > Hmmm. These articles both seem pretty thin when it comes to any
                data to
                > > validate the theory.
                > >
                > >
                > >
                > > To me, it's always seemed like page load time would be a much
                bigger deal
                > > for sites where there are lots of other alternative sites that
                can be
                > > expected to offer largely similar content: cnn.com vs. msnbc.com,
                > > lowes.com vs. homedepot.com, espn.com vs.
                sportsillustrated.cnn.com. In a
                > > B2C world, for sites that I'm going to get into a pattern of
                visiting,
                > > persistently slow load times are going to drive me to look
                elsewhere. In the
                > > B2B world â€" higher price point products/services, more niches,
                longer buying
                > > cycle â€" it seems like small changes in page load times would
                not have much
                > > of an impact. I would happily wait for 10 seconds for a
                > > www.webanalyticsassociation.org
                <http://www.webanalyticsassociation.org/>
                > > page to load (not that it takes that long!).
                > >
                > >
                > >
                > > It's an interesting challenge for folks on the IT side of the
                house â€" new
                > > functionality rolling out that may require more server crunching
                to deliver
                > > it. How much $$$ should they spend to maintain or reduce load
                times in the
                > > light of new functionality (that, presumably, adds more value for
                the
                > > visitor)?
                > >
                > >
                > >
                > > (Sorry, Michelle â€" I don't have an answer by any means…but
                I'd love to see
                > > some insights on this from this group!)
                > >
                > >
                > >
                > > Tim
                > >
                > >
                > >
                > > ________________________________
                > >
                > > From: webanalytics@yahoogroups.com <webanalytics%
                40yahoogroups.com>[mailto:
                > > webanalytics@yahoogroups.com <webanalytics%40yahoogroups.com>] On
                Behalf
                > > Of Apurba Sen
                > > Sent: Sunday, July 01, 2007 4:50 AM
                > > To: webanalytics@yahoogroups.com <webanalytics%40yahoogroups.com>
                > > Subject: Re: [webanalytics] Relationship between page load time
                and
                > > drop-off rate?
                > >
                > >
                > >
                > > Hey Michelle,
                > >
                > > Reminds me of 8 second rule :) i.e if users have to wait longer
                than 8
                > > seconds to download a Web page, they will go elsewhere.So any
                page load
                > > time
                > > greater than or equal to 8 sec should ideally mean 100 % drop
                off :)
                > >
                > > Some related links:
                > >
                > > http://en.wikipedia.org/wiki/Network_performance <
                > > http://en.wikipedia.org/wiki/Network_performance>
                > >
                > > http://www.pcmag.com/encyclopedia_term/0,2542,t=8-
                second+rule&i=37197,00.asp<
                > > http://www.pcmag.com/encyclopedia_term/0,2542,t=8-
                second+rule&i=37197,00.asp>
                > >
                > >
                > > Regds
                > > Apurba
                > >
                > > On 6/28/07, Michelle Runyan <michelle.runyan@...<michelle.runyan%
                40ni.com><mailto:
                > > michelle.runyan%40ni.com> > wrote:
                > > >
                > > > Has anyone done analysis or seen data that shows the
                relationship
                > > > between load time and drop-off rate? I have seen a few reports
                that
                > > > present what people say they are ok with (for example
                > > >
                http://www.akamai.com/dl/reports/Site_Abandonment_Final_Report.pdf <
                > >
                http://www.akamai.com/dl/reports/Site_Abandonment_Final_Report.pdf> ),
                > > > but the real value is in what the data says.
                > > >
                > > > We have tried analyzing our raw data and can't find a wide
                enough
                > > > spectrum of load times to derive that X% drop off at 3-4
                seconds and X%
                > > > drop off at 5-6 seconds ...
                > > >
                > > > If you have any thought or ideas please let me know.
                > > >
                > > > Thanks,
                > > > Michelle
                > > >
                > > >
                > > >
                > >
                > > --
                > > Cheers!!
                > > Apurba Sen
                > > Product Development
                > > रेडीफ़.कॉम इंडिया
                लिमिटेड
                > > NASDAQ:REDF
                > > web:www.rediff.com
                > > http://apurba.rediffiland.com <http://apurba.rediffiland.com>
                > > Phone :91-22 2444 9144 * 313
                > > GSM :91-98670- 85718
                > >
                > > "Champions aren't made in the gyms. Champions are made from
                something they
                > >
                > > have deep inside them -- a desire, a dream, a vision."
                > >
                > > --Muhammad Ali
                > >
                > > [Non-text portions of this message have been removed]
                > >
                > >
                > >
                > >
                > > [Non-text portions of this message have been removed]
                > >
                > >
                > >
                >
                >
                >
                > --
                > Cheers!!
                > Apurba Sen
                > Product Development
                > रेडीफ़.कॉम इंडिया
                लिमिटेड
                > NASDAQ:REDF
                > web:www.rediff.com
                > http://apurba.rediffiland.com
                > Phone :91-22 2444 9144 * 313
                > GSM :91-98670- 85718
                >
                > "Champions aren't made in the gyms. Champions are made from
                something they
                > have deep inside them -- a desire, a dream, a vision."
                >
                > --Muhammad Ali
                >
                >
                > [Non-text portions of this message have been removed]
                >
              • Apurba Sen
                Hey Justo, Thanks a lot.My estimation was 100% ..its somewhere close to 50% :) -Apurba ... -- Cheers!! Apurba Sen Product Development
                Message 7 of 14 , Jul 3, 2007
                • 0 Attachment
                  Hey Justo,

                  Thanks a lot.My estimation was 100% ..its somewhere close to 50% :)

                  -Apurba


                  On 7/3/07, justo_ibarra <jibarra@...> wrote:
                  >
                  > Hi everybody:
                  >
                  > I just read a recent article about the 8 seconds rule. Altough
                  > develops a more theoretical approach on the subject maybe results
                  > helpfull.
                  >
                  > http://www.silverpop.com/landing/emarketer/24.html
                  >
                  > You can download it for free.
                  >
                  > Regards,
                  >
                  > Justo Ibarra.
                  > jibarra@... <jibarra%40latin3.com>
                  >
                  > --- In webanalytics@yahoogroups.com <webanalytics%40yahoogroups.com>,
                  > "Apurba Sen" <sensession@...>
                  > wrote:
                  > >
                  > > Hey Tim!!
                  > >
                  > > Let me answer your query from marketing standpoint,to be specific
                  > through
                  > > Brand Loyalty vs Satisfaction.
                  > >
                  > > Using Jones & Sasser framework, customers can be plotted in 4
                  > quadrants of
                  > > Loyalty [Y Axis] vs Satisfaction [X Axis]
                  > >
                  > > A.Loyalty Low -Satisfaction Low :Terrorist
                  > >
                  > > B.Loyalty Low-Satisfaction High :Mercenaries
                  > >
                  > > C.Loyalty High-Satisfaction Low :Hostages
                  > >
                  > > D.Loyalty High-Satisfaction High :Apostles
                  > >
                  > >
                  > > Customers at the lowest and highest ends of the satisfaction scale
                  > tend to
                  > > have intense feelings about a brand and its products /services.
                  > Terrorists
                  > > actively attack the brand telling others not to buy from the
                  > company.
                  > > Apostles are satisfied and loyal and talk favourably about the
                  > brand.
                  > >
                  > > Using your examples...
                  > >
                  > > Terrorist or Mercenaries:Page load time would be a somehow bigger
                  > deal where
                  > > plenty of alternative sites are available.They can be found in
                  > abundance in
                  > > [cnn.com vs. msnbc.com, lowes.com vs. homedepot.com, espn.com vs.
                  > > sportsillustrated.cnn.com]
                  > >
                  > > Apostles or Hostages:B2B world ,niches even
                  > > www.webanalyticsassociation.org.Thinking abt "Hostages" in this
                  > group..what
                  > > are the other options avl with them :-)
                  > >
                  > > Hope this was enlightening enough.
                  > >
                  > > Reducing page delivery time should always be a goal while designing
                  > the
                  > > page/information layout e.g. progressive download with relatively
                  > static
                  > > content being cached.
                  > >
                  > > Regds
                  > >
                  > > Apurba Sen
                  > >
                  > > Related Article :
                  > >
                  > > Jones, T., & Sasser, E., 'Why Satisfied Customers Defect' - *Harvard
                  > > Business Review*, Nov-Dec 1995
                  > >
                  > >
                  > >
                  > >
                  > >
                  > >
                  > >
                  > > On 7/2/07, Tim Wilson <twilson@...> wrote:
                  > > >
                  > > > Hmmm. These articles both seem pretty thin when it comes to any
                  > data to
                  > > > validate the theory.
                  > > >
                  > > >
                  > > >
                  > > > To me, it's always seemed like page load time would be a much
                  > bigger deal
                  > > > for sites where there are lots of other alternative sites that
                  > can be
                  > > > expected to offer largely similar content: cnn.com vs. msnbc.com,
                  > > > lowes.com vs. homedepot.com, espn.com vs.
                  > sportsillustrated.cnn.com. In a
                  > > > B2C world, for sites that I'm going to get into a pattern of
                  > visiting,
                  > > > persistently slow load times are going to drive me to look
                  > elsewhere. In the
                  > > > B2B world �€" higher price point products/services, more niches,
                  > longer buying
                  > > > cycle �€" it seems like small changes in page load times would
                  > not have much
                  > > > of an impact. I would happily wait for 10 seconds for a
                  > > > www.webanalyticsassociation.org
                  > <http://www.webanalyticsassociation.org/>
                  > > > page to load (not that it takes that long!).
                  > > >
                  > > >
                  > > >
                  > > > It's an interesting challenge for folks on the IT side of the
                  > house �€" new
                  > > > functionality rolling out that may require more server crunching
                  > to deliver
                  > > > it. How much $$$ should they spend to maintain or reduce load
                  > times in the
                  > > > light of new functionality (that, presumably, adds more value for
                  > the
                  > > > visitor)?
                  > > >
                  > > >
                  > > >
                  > > > (Sorry, Michelle �€" I don't have an answer by any means…but
                  > I'd love to see
                  > > > some insights on this from this group!)
                  > > >
                  > > >
                  > > >
                  > > > Tim
                  > > >
                  > > >
                  > > >
                  > > > ________________________________
                  > > >
                  > > > From: webanalytics@yahoogroups.com <webanalytics%40yahoogroups.com><webanalytics%
                  > 40yahoogroups.com>[mailto:
                  > > > webanalytics@yahoogroups.com <webanalytics%40yahoogroups.com><webanalytics%40yahoogroups.com>] On
                  > Behalf
                  > > > Of Apurba Sen
                  > > > Sent: Sunday, July 01, 2007 4:50 AM
                  > > > To: webanalytics@yahoogroups.com <webanalytics%40yahoogroups.com><webanalytics%40yahoogroups.com>
                  > > > Subject: Re: [webanalytics] Relationship between page load time
                  > and
                  > > > drop-off rate?
                  > > >
                  > > >
                  > > >
                  > > > Hey Michelle,
                  > > >
                  > > > Reminds me of 8 second rule :) i.e if users have to wait longer
                  > than 8
                  > > > seconds to download a Web page, they will go elsewhere.So any
                  > page load
                  > > > time
                  > > > greater than or equal to 8 sec should ideally mean 100 % drop
                  > off :)
                  > > >
                  > > > Some related links:
                  > > >
                  > > > http://en.wikipedia.org/wiki/Network_performance <
                  > > > http://en.wikipedia.org/wiki/Network_performance>
                  > > >
                  > > > http://www.pcmag.com/encyclopedia_term/0,2542,t=8-
                  > second+rule&i=37197,00.asp<
                  > > > http://www.pcmag.com/encyclopedia_term/0,2542,t=8-
                  > second+rule&i=37197,00.asp>
                  > > >
                  > > >
                  > > > Regds
                  > > > Apurba
                  > > >
                  > > > On 6/28/07, Michelle Runyan <michelle.runyan@...<michelle.runyan%
                  > 40ni.com><mailto:
                  > > > michelle.runyan%40ni.com> > wrote:
                  > > > >
                  > > > > Has anyone done analysis or seen data that shows the
                  > relationship
                  > > > > between load time and drop-off rate? I have seen a few reports
                  > that
                  > > > > present what people say they are ok with (for example
                  > > > >
                  > http://www.akamai.com/dl/reports/Site_Abandonment_Final_Report.pdf <
                  > > >
                  > http://www.akamai.com/dl/reports/Site_Abandonment_Final_Report.pdf> ),
                  > > > > but the real value is in what the data says.
                  > > > >
                  > > > > We have tried analyzing our raw data and can't find a wide
                  > enough
                  > > > > spectrum of load times to derive that X% drop off at 3-4
                  > seconds and X%
                  > > > > drop off at 5-6 seconds ...
                  > > > >
                  > > > > If you have any thought or ideas please let me know.
                  > > > >
                  > > > > Thanks,
                  > > > > Michelle
                  > > > >
                  > > > >
                  > > > >
                  > > >
                  > > > --
                  > > > Cheers!!
                  > > > Apurba Sen
                  > > > Product Development
                  > > > रेडीफ़.कॉम इंडिया
                  > लिमिटेड
                  > > > NASDAQ:REDF
                  > > > web:www.rediff.com
                  > > > http://apurba.rediffiland.com <http://apurba.rediffiland.com>
                  > > > Phone :91-22 2444 9144 * 313
                  > > > GSM :91-98670- 85718
                  > > >
                  > > > "Champions aren't made in the gyms. Champions are made from
                  > something they
                  > > >
                  > > > have deep inside them -- a desire, a dream, a vision."
                  > > >
                  > > > --Muhammad Ali
                  > > >
                  > > > [Non-text portions of this message have been removed]
                  > > >
                  > > >
                  > > >
                  > > >
                  > > > [Non-text portions of this message have been removed]
                  > > >
                  > > >
                  > > >
                  > >
                  > >
                  > >
                  > > --
                  > > Cheers!!
                  > > Apurba Sen
                  > > Product Development
                  > > रेडीफ़.कॉम इंडिया
                  > लिमिटेड
                  > > NASDAQ:REDF
                  > > web:www.rediff.com
                  > > http://apurba.rediffiland.com
                  > > Phone :91-22 2444 9144 * 313
                  > > GSM :91-98670- 85718
                  > >
                  > > "Champions aren't made in the gyms. Champions are made from
                  > something they
                  > > have deep inside them -- a desire, a dream, a vision."
                  > >
                  > > --Muhammad Ali
                  > >
                  > >
                  > > [Non-text portions of this message have been removed]
                  > >
                  >
                  >
                  >



                  --
                  Cheers!!
                  Apurba Sen
                  Product Development
                  रेडीफ़.कॉम इंडिया लिमिटेड
                  NASDAQ:REDF
                  web:www.rediff.com
                  http://apurba.rediffiland.com
                  Phone :91-22 2444 9144 * 313
                  GSM :91-98670- 85718

                  "Champions aren't made in the gyms. Champions are made from something they
                  have deep inside them -- a desire, a dream, a vision."

                  --Muhammad Ali


                  [Non-text portions of this message have been removed]
                • Pedro
                  Some recent data I saw from a B2C website that I worked with found that 4 seconds was the tipping point. You lost a large percentage of your users at the 4
                  Message 8 of 14 , Jul 5, 2007
                  • 0 Attachment
                    Some recent data I saw from a B2C website that I worked with found
                    that 4 seconds was the tipping point.

                    You lost a large percentage of your users at the 4 second mark. The
                    interesting thing to note was that optimizing a page from a 10 seconds
                    load time to 6 seconds made almost no difference, whereas
                    optimizations from 6 seconds to 4 seconds had a huge impact.

                    Then there's the issue of perceived load time vs. actual load time. If
                    pages load important elements very quickly, users didn't seem to have
                    a problem waiting longer for the full page to load.

                    Hope that helps a little bit.
                  • Apurba Sen
                    Absolutely on the mark Pedro. Important elements should be called first..then the corresponding ads..the basic idea of progressive download of a page.:-) Regds
                    Message 9 of 14 , Jul 6, 2007
                    • 0 Attachment
                      Absolutely on the mark Pedro.

                      Important elements should be called first..then the corresponding ads..the
                      basic idea of progressive download of a page.:-)


                      Regds
                      Apurba



                      On 7/5/07, Pedro <pedro@...> wrote:
                      >
                      > Some recent data I saw from a B2C website that I worked with found
                      > that 4 seconds was the tipping point.
                      >
                      > You lost a large percentage of your users at the 4 second mark. The
                      > interesting thing to note was that optimizing a page from a 10 seconds
                      > load time to 6 seconds made almost no difference, whereas
                      > optimizations from 6 seconds to 4 seconds had a huge impact.
                      >
                      > Then there's the issue of perceived load time vs. actual load time. If
                      > pages load important elements very quickly, users didn't seem to have
                      > a problem waiting longer for the full page to load.
                      >
                      > Hope that helps a little bit.
                      >
                      >
                      >



                      --
                      Cheers!!
                      Apurba Sen
                      Product Development
                      रेडीफ़.कॉम इंडिया लिमिटेड
                      NASDAQ:REDF
                      web:www.rediff.com
                      http://apurba.rediffiland.com
                      Phone :91-22 2444 9144 * 313
                      GSM :91-98670- 85718

                      "Champions aren't made in the gyms. Champions are made from something they
                      have deep inside them -- a desire, a dream, a vision."

                      --Muhammad Ali


                      [Non-text portions of this message have been removed]
                    • Brandon Hartness
                      How are you tracking page load time? Are you loading CSS, Javascript, images, etc. as a factor as well? Is there a way to do this with Omniture? Brandon ...
                      Message 10 of 14 , Jul 6, 2007
                      • 0 Attachment
                        How are you tracking page load time? Are you loading CSS, Javascript,
                        images, etc. as a factor as well? Is there a way to do this with Omniture?

                        Brandon


                        On 7/5/07, Pedro <pedro@...> wrote:
                        >
                        > Some recent data I saw from a B2C website that I worked with found
                        > that 4 seconds was the tipping point.
                        >
                        > You lost a large percentage of your users at the 4 second mark. The
                        > interesting thing to note was that optimizing a page from a 10 seconds
                        > load time to 6 seconds made almost no difference, whereas
                        > optimizations from 6 seconds to 4 seconds had a huge impact.
                        >
                        > Then there's the issue of perceived load time vs. actual load time. If
                        > pages load important elements very quickly, users didn't seem to have
                        > a problem waiting longer for the full page to load.
                        >
                        > Hope that helps a little bit.
                        >
                        >
                        >


                        [Non-text portions of this message have been removed]
                      • Ravi Pathak
                        When I start thinking about page load time, I think they have two dimensions to be looked at. 1. Actual page load time 2. Perceived page load time by users Is
                        Message 11 of 14 , Jul 7, 2007
                        • 0 Attachment
                          When I start thinking about page load time, I think they have two dimensions
                          to be looked at.

                          1. Actual page load time
                          2. Perceived page load time by users

                          Is anyone aware if we can actually reduce drop off by reducing 'perceived
                          page load time' then actual page load time.

                          Thanks,
                          Ravi
                          On 7/6/07, Apurba Sen <sensession@...> wrote:
                          >
                          > Absolutely on the mark Pedro.
                          >
                          > Important elements should be called first..then the corresponding ads..the
                          > basic idea of progressive download of a page.:-)
                          >
                          > Regds
                          > Apurba
                          >
                          > On 7/5/07, Pedro <pedro@... <pedro%40sostreassoc.com>> wrote:
                          > >
                          > > Some recent data I saw from a B2C website that I worked with found
                          > > that 4 seconds was the tipping point.
                          > >
                          > > You lost a large percentage of your users at the 4 second mark. The
                          > > interesting thing to note was that optimizing a page from a 10 seconds
                          > > load time to 6 seconds made almost no difference, whereas
                          > > optimizations from 6 seconds to 4 seconds had a huge impact.
                          > >
                          > > Then there's the issue of perceived load time vs. actual load time. If
                          > > pages load important elements very quickly, users didn't seem to have
                          > > a problem waiting longer for the full page to load.
                          > >
                          > > Hope that helps a little bit.
                          > >
                          > >
                          > >
                          >
                          > --
                          > Cheers!!
                          > Apurba Sen
                          > Product Development
                          > रेडीफ़.कॉम इंडिया लिमिटेड
                          > NASDAQ:REDF
                          > web:www.rediff.com
                          > http://apurba.rediffiland.com
                          > Phone :91-22 2444 9144 * 313
                          > GSM :91-98670- 85718
                          >
                          > "Champions aren't made in the gyms. Champions are made from something they
                          > have deep inside them -- a desire, a dream, a vision."
                          >
                          > --Muhammad Ali
                          >
                          > [Non-text portions of this message have been removed]
                          >
                          >
                          >


                          [Non-text portions of this message have been removed]
                        • Apurba Sen
                          Lol :D If consumer sees a blank page ,the perceived time would be much higher. A status update like Loading in next N seconds def helps to cater to the
                          Message 12 of 14 , Jul 7, 2007
                          • 0 Attachment
                            Lol :D

                            If consumer sees a blank page ,the perceived time would be much higher.

                            A status update like "Loading in next N seconds" def helps to cater to the
                            anxiety.




                            On 7/7/07, Ravi Pathak <ravipathak1@...> wrote:
                            >
                            > When I start thinking about page load time, I think they have two
                            > dimensions
                            > to be looked at.
                            >
                            > 1. Actual page load time
                            > 2. Perceived page load time by users
                            >
                            > Is anyone aware if we can actually reduce drop off by reducing 'perceived
                            > page load time' then actual page load time.
                            >
                            > Thanks,
                            > Ravi
                            > On 7/6/07, Apurba Sen <sensession@... <sensession%40gmail.com>>
                            > wrote:
                            > >
                            > > Absolutely on the mark Pedro.
                            > >
                            > > Important elements should be called first..then the corresponding
                            > ads..the
                            > > basic idea of progressive download of a page.:-)
                            > >
                            > > Regds
                            > > Apurba
                            > >
                            > > On 7/5/07, Pedro <pedro@... <pedro%40sostreassoc.com><pedro%40sostreasso
                            > c.com>> wrote:
                            > > >
                            > > > Some recent data I saw from a B2C website that I worked with found
                            > > > that 4 seconds was the tipping point.
                            > > >
                            > > > You lost a large percentage of your users at the 4 second mark. The
                            > > > interesting thing to note was that optimizing a page from a 10 seconds
                            > > > load time to 6 seconds made almost no difference, whereas
                            > > > optimizations from 6 seconds to 4 seconds had a huge impact.
                            > > >
                            > > > Then there's the issue of perceived load time vs. actual load time. If
                            > > > pages load important elements very quickly, users didn't seem to have
                            > > > a problem waiting longer for the full page to load.
                            > > >
                            > > > Hope that helps a little bit.
                            > > >
                            > > >
                            > > >
                            > >
                            > > --
                            > > Cheers!!
                            > > Apurba Sen
                            > > Product Development
                            > > रेडीफ़.कॉम इंडिया लिमिटेड
                            > > NASDAQ:REDF
                            > > web:www.rediff.com
                            > > http://apurba.rediffiland.com
                            > > Phone :91-22 2444 9144 * 313
                            > > GSM :91-98670- 85718
                            > >
                            > > "Champions aren't made in the gyms. Champions are made from something
                            > they
                            > > have deep inside them -- a desire, a dream, a vision."
                            > >
                            > > --Muhammad Ali
                            > >
                            > > [Non-text portions of this message have been removed]
                            > >
                            > >
                            > >
                            >
                            > [Non-text portions of this message have been removed]
                            >
                            >
                            >



                            --
                            Cheers!!
                            Apurba Sen
                            Product Development
                            रेडीफ़.कॉम इंडिया लिमिटेड
                            NASDAQ:REDF
                            web:www.rediff.com
                            http://apurba.rediffiland.com
                            Phone :91-22 2444 9144 * 313
                            GSM :91-98670- 85718

                            "Champions aren't made in the gyms. Champions are made from something they
                            have deep inside them -- a desire, a dream, a vision."

                            --Muhammad Ali


                            [Non-text portions of this message have been removed]
                          • Wandering Dave Rhee
                            Hi, Michelle and all, Here s my late contribution to this thread... I believe page load has much less to do with the site, or the individual page within the
                            Message 13 of 14 , Jul 7, 2007
                            • 0 Attachment
                              Hi, Michelle and all,

                              Here's my late contribution to this thread...

                              I believe page load has much less to do with the site, or the individual
                              page within the site, and everything to do with the expectation of your
                              audience for that page.

                              For example, on a landing page, the expectations are that my page will load
                              within, say, 20% of the time of my competitor's similar landing pages. And
                              since landing pages typically load quickly, mine had also better load
                              quickly (e.g., sub-4 seconds or whatever).

                              But if I'm visiting my bank account info, and they say, "Please wait while
                              we retrieve your account information," and give me a little
                              clock-spinner-graphic, I expect to wait 10-20 seconds, and won't abandon.
                              If, on the other hand, another site of mine usually has no graphic, and no
                              "please wait," and my account details load within 4 seconds, then suddenly
                              I'm upset if it takes 8-10, even though it may still be faster than my other
                              account's very similar site (in terms of functionality and market).

                              So if you load as quickly as your visitors expect, you're fine. If you load
                              more slowly, and don't lead your customers to expect it, then that's a Bad
                              Thing.

                              Even for a single content page, you may have a mix of visitors new to your site, and veterans -- they may have different expectations from each other for the same page.

                              How do you tell what your visitor expectations are? That's part of the
                              value of doing testing with hidden cameras, or interviewing test subjects.
                              More expensive, but you're not as likely to get good survey data otherwise
                              about expectations of performance, by page.

                              WDave


                              [Non-text portions of this message have been removed]
                            Your message has been successfully submitted and would be delivered to recipients shortly.