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

ecommerce sites. Conversion rate before and after redesign anyone??

Expand Messages
  • mr_awesome77
    Hi I am in the process of re-architecting a big ecommerce group s sites. the question I keep getting asked by those concerned is what will our conversion rate
    Message 1 of 10 , Sep 24, 2009
      Hi

      I am in the process of re-architecting a big ecommerce group's sites.

      the question I keep getting asked by those concerned is "what will our conversion rate be with the new site"

      I am looking at a lot of targetting and merchandising improvements... and the site is currently AWFUL. in every single way.

      Does anyone have any experiences they can share?

      I am banking on a 5% increase in conversion (and a 5% increase in basket size).. eg if we are doing 100 orders a day then with the new site we would do 105 (the real number is much bigger than this of course).. dont know if this is conservative

      I realise there are a lot of case studies around from various vendors.. but I trust these are far as I can throw them .. generally speaking.

      It is a very difficult one to predict or model.

      cheers all
      jon
    • damn_wright
      A/B test with the two sites running simultaneously.
      Message 2 of 10 , Sep 24, 2009
        A/B test with the two sites running simultaneously.
      • mr_awesome77
        ... sure, I understand that.. but management asks the question of me: If we are investing x hundred thousand in this.. how much extra revenue will be generate?
        Message 3 of 10 , Sep 24, 2009
          --- In webanalytics@yahoogroups.com, "damn_wright" <daniel@...> wrote:
          >
          > A/B test with the two sites running simultaneously.
          >


          sure, I understand that.. but management asks the question of me:
          If we are investing x hundred thousand in this.. how much extra revenue will be generate?

          They dont want to know after the money is spent.

          jon
        • Miles Bennett (Targetstone)
          Couldn t you look at the faults in the conversion process eg drop offs, number of orders that have to be completed offline and also work out a perceived value
          Message 4 of 10 , Sep 25, 2009
            Couldn't you look at the faults in the conversion process eg drop offs, number of orders that have to be completed offline and also work out a perceived value per click or page and give a profit and loss statement based on that?

            Miles
            Miles Bennett
            TargetStone
            M: 07545822786
            W: www.targetstone.com
            B: http://blog.targetstone.com
            L: http://www.linkedin.com/in/milesbennett

            -----Original Message-----
            From: "mr_awesome77" <jon.bovard@...>
            Date: Fri, 25 Sep 2009 00:32:27
            To: <webanalytics@yahoogroups.com>
            Subject: [webanalytics] Re: ecommerce sites. Conversion rate before and after redesign anyone??

            --- In webanalytics@yahoogroups.com, "damn_wright" <daniel@...> wrote:
            >
            > A/B test with the two sites running simultaneously.
            >


            sure, I understand that.. but management asks the question of me:
            If we are investing x hundred thousand in this.. how much extra revenue will be generate?

            They dont want to know after the money is spent.

            jon




            [Non-text portions of this message have been removed]
          • Jared Spool
            ... Hi Jon, This is a different question than the one you asked before. Conversion rates are very different (and often unrelated) to revenues. It s frequently
            Message 5 of 10 , Sep 25, 2009
              On Sep 24, 2009, at 8:32 PM, mr_awesome77 wrote:

              > sure, I understand that.. but management asks the question of me:
              > If we are investing x hundred thousand in this.. how much extra
              > revenue will be generate?
              >
              > They dont want to know after the money is spent.

              Hi Jon,

              This is a different question than the one you asked before.

              Conversion rates are very different (and often unrelated) to revenues.
              It's frequently meaningless to talk about conversion rates, since they
              can go up while revenues go down (being that they're a ratio).
              Revenues is the right metric to be looking at when talking about
              design changes. (Conversion rates are only really meaningful when
              you're talking about a specific campaign, but even then, they suffer
              because you haven't compensated for lead quality. But, that's a whole
              other discussion.)

              I think the reason that you didn't get an immediate response is that,
              in my company's research, we've found most major redesigns of e-
              commerce sites result in a 20% (or greater) loss in revenues.
              Organizations that redesign completely and lose 10% or less are rare
              and should consider themselves very lucky.

              In the past 18 months, one of the top-20 retailers did a full redesign
              of their site, only to see revenues overall drop by about 28%. (Their
              highest margin, most popular products dropped revenues by 45%. Profits
              sank big with the redesign.)

              A few years back, a major retailer spent $100,000,000 on a major
              redesign to see a 20% reduction in sales. (Imagine what they could've
              achieved had they only spent $50,000,000.) It took them 3.5 years to
              bring revenues back to the level they were at before the redesign.

              I think your management is asking the right question. And I think
              you'll find the right answer is to not do a redesign at all. At least,
              not all at once. (Wrote about this a long time ago here: http://is.gd/3F679
              It's still true today.)

              The most successful sites do incremental changes, the smaller the
              better. When was the last time Amazon launched a redesign? Or Netflix?
              Avid users of these sites can't tell you because the sites never have.
              Instead they change on a weekly (in the case of Amazon, daily) basis.
              Small changes. (Our research shows that one such small change on
              Amazon's site looks like it resulted in an annual revenue increase of
              $2.7 billion: http://is.gd/3F6j1)

              The key question to answer your management's question is: How much
              time have you spent watching real shoppers shop for their desired
              products on your site? If you're doing this effectively, you can
              estimate, fairly accurately, the revenues you'll gain by specific
              design changes. (We call this metric "money left on the table".)

              If a site hasn't had much attention spent to its design, there's often
              low-hanging fruit that can see a good increase in revenues without
              much effort. Because it's not a redesign, you can make the change
              quickly and don't run the risk of the huge reductions in revenues that
              redesigns typically bring. (If, for some reason you get it wrong and
              do reduce the revenues as a result, because it was a small change, you
              can usually roll it back easily.)

              Hope that helps,

              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 Twitter: @jmspool



              [Non-text portions of this message have been removed]
            • eran_savir
              I think that it really depends on the difference in the design between the old and the new website. I would suggest running a feedback form on your website
              Message 6 of 10 , Sep 28, 2009
                I think that it really depends on the difference in the design between the old and the new website.
                I would suggest running a feedback form on your website before the redesign, this will allow you to see what people are talking about, what's not working well, and what do people really like.
                You can try installing http://www.kampyle.com on your website. You'll start getting answers from your users immediately.

                Eran.
              • mr_awesome77
                hmmmm understand... But what if I told you two things.. 1. The business is a Direct marketing business.. and about 95% of its daily traffic is new .. ie. the
                Message 7 of 10 , Sep 28, 2009
                  hmmmm understand...

                  But what if I told you two things..

                  1. The business is a Direct marketing business.. and about 95% of its daily traffic is "new".. ie. the repeat rate of its traffic/customers is very low relative to the web universe - and especially very different to large established etailers with large % of repeat customers and traffic? (ignoring the fact that yes - new customer acqusition as a revenue model emphasis - is a very inefficient way to run a business!). Would this change the proposition of a redesign? ....given that we arent "as" concerned as we would be if we had a LOT of regular customers/visitors returning find the house has been knocked down... and somebody has built an entirely new site.

                  2. The business has thus far generated literally.. multiple millions of dollars business on an off the shelf ecommerce "platform" called Shopfactory. This monstrosity of a platform cannot even allow the end user/developer to move a button, change a link, resize an image, touch the underlying code - one...single.. bit.... crazy stuff. So MVT/AB testing is not even an option.

                  Its a nightmare but also an opportunity.

                  Yes, its a strange world we live in, but those two things above I have inherited as part of a new job :)


                  cheers
                  jon



                  >
                  > Hi Jon,
                  >
                  > This is a different question than the one you asked before.
                  >
                  > Conversion rates are very different (and often unrelated) to revenues.
                  > It's frequently meaningless to talk about conversion rates, since they
                  > can go up while revenues go down (being that they're a ratio).
                  > Revenues is the right metric to be looking at when talking about
                  > design changes. (Conversion rates are only really meaningful when
                  > you're talking about a specific campaign, but even then, they suffer
                  > because you haven't compensated for lead quality. But, that's a whole
                  > other discussion.)
                  >
                  > I think the reason that you didn't get an immediate response is that,
                  > in my company's research, we've found most major redesigns of e-
                  > commerce sites result in a 20% (or greater) loss in revenues.
                  > Organizations that redesign completely and lose 10% or less are rare
                  > and should consider themselves very lucky.
                  >
                  > In the past 18 months, one of the top-20 retailers did a full redesign
                  > of their site, only to see revenues overall drop by about 28%. (Their
                  > highest margin, most popular products dropped revenues by 45%. Profits
                  > sank big with the redesign.)
                  >
                  > A few years back, a major retailer spent $100,000,000 on a major
                  > redesign to see a 20% reduction in sales. (Imagine what they could've
                  > achieved had they only spent $50,000,000.) It took them 3.5 years to
                  > bring revenues back to the level they were at before the redesign.
                  >
                  > I think your management is asking the right question. And I think
                  > you'll find the right answer is to not do a redesign at all. At least,
                  > not all at once. (Wrote about this a long time ago here: http://is.gd/3F679
                  > It's still true today.)
                  >
                  > The most successful sites do incremental changes, the smaller the
                  > better. When was the last time Amazon launched a redesign? Or Netflix?
                  > Avid users of these sites can't tell you because the sites never have.
                  > Instead they change on a weekly (in the case of Amazon, daily) basis.
                  > Small changes. (Our research shows that one such small change on
                  > Amazon's site looks like it resulted in an annual revenue increase of
                  > $2.7 billion: http://is.gd/3F6j1)
                  >
                  > The key question to answer your management's question is: How much
                  > time have you spent watching real shoppers shop for their desired
                  > products on your site? If you're doing this effectively, you can
                  > estimate, fairly accurately, the revenues you'll gain by specific
                  > design changes. (We call this metric "money left on the table".)
                  >
                  > If a site hasn't had much attention spent to its design, there's often
                  > low-hanging fruit that can see a good increase in revenues without
                  > much effort. Because it's not a redesign, you can make the change
                  > quickly and don't run the risk of the huge reductions in revenues that
                  > redesigns typically bring. (If, for some reason you get it wrong and
                  > do reduce the revenues as a result, because it was a small change, you
                  > can usually roll it back easily.)
                  >
                  > Hope that helps,
                  >
                  > 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 Twitter: @jmspool
                  >
                  >
                  >
                  > [Non-text portions of this message have been removed]
                  >
                • Jared Spool
                  Good questions. The answer is still that I believe you don t want to do a total redesign. That 95% of your business comes from new customers doesn t really
                  Message 8 of 10 , Sep 29, 2009
                    Good questions.

                    The answer is still that I believe you don't want to do a total
                    redesign.

                    That 95% of your business comes from new customers doesn't really
                    matter here. The important assumption I'm making is that you and the
                    team haven't spent enough time watching real customers shop on your
                    site. (I came to that conclusion *because* you want to do a redesign.
                    Folks who regularly watch their shoppers -- a minimum of 4 hours of
                    observation each month -- rarely go that route.)

                    Assuming I'm correct and you haven't been watching your shoppers, it
                    means that you probably don't have a good handle on (a) what the
                    problems are and (b) what's been working to generate your multiple
                    million dollars of business. So, the odds are great that you'll break
                    something before you'll get it right. The less you know about how
                    people actually shop, the harder it will be to return to the current
                    level once you've broken something.

                    Avoiding redesigns is only partially about upsetting the current
                    customer base. It's about minimizing risk. The smaller the changes you
                    make, the more you can focus on collecting data.

                    Say you took a single product line -- not your highest volume, but
                    something with a small number of SKUs that currently gets decent
                    enough sales -- and just redesigned that. You'd want to start by
                    studying how shoppers buy those products today. What's working and
                    what isn't. Then you redesign just those pages, even using a new
                    platform (to test trial an alternative to Shopfactory).

                    The resulting project will be much smaller than a site-wide redesign.
                    It'll probably have less stakeholders (only the merchandisers and
                    management of that product line) to contend with. And, if you screw it
                    up (the odds are good the first few times), it's probably easy to roll
                    back to your existing system while you figure out what to do better
                    next time.

                    I understand the desire in a new job to want to make a big change
                    quickly. And there's a lot to be said about rolling the dice and going
                    for the big win. However, experience tells me that a conservative,
                    minimized-risk approach will get better returns in the long run.

                    That's my opinion, which is worth exactly what you paid for it.

                    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 Twitter: @jmspool



                    On Sep 28, 2009, at 8:16 AM, mr_awesome77 wrote:

                    > hmmmm understand...
                    >
                    > But what if I told you two things..
                    >
                    > 1. The business is a Direct marketing business.. and about 95% of
                    > its daily traffic is "new".. ie. the repeat rate of its traffic/
                    > customers is very low relative to the web universe - and especially
                    > very different to large established etailers with large % of repeat
                    > customers and traffic? (ignoring the fact that yes - new customer
                    > acqusition as a revenue model emphasis - is a very inefficient way
                    > to run a business!). Would this change the proposition of a
                    > redesign? ....given that we arent "as" concerned as we would be if
                    > we had a LOT of regular customers/visitors returning find the house
                    > has been knocked down... and somebody has built an entirely new site.
                    >
                    > 2. The business has thus far generated literally.. multiple millions
                    > of dollars business on an off the shelf ecommerce "platform" called
                    > Shopfactory. This monstrosity of a platform cannot even allow the
                    > end user/developer to move a button, change a link, resize an image,
                    > touch the underlying code - one...single.. bit.... crazy stuff. So
                    > MVT/AB testing is not even an option.
                    >
                    > Its a nightmare but also an opportunity.
                    >
                    > Yes, its a strange world we live in, but those two things above I
                    > have inherited as part of a new job :)



                    [Non-text portions of this message have been removed]
                  • fredeilam
                    First, to answer your initial question, based on work I ve been personally involved with, I ve seen conversion improvements from zero to 50%. As others have
                    Message 9 of 10 , Sep 29, 2009
                      First, to answer your initial question, based on work I've been personally involved with, I've seen conversion improvements from zero to 50%.

                      As others have said before me - it depends.

                      To put this in perspective, tell the client that it's like asking a broker if we switch our 401k from plan A to plan B how much additional money will I make?

                      You're comment though that split testing is not an option is not true.

                      I understand that you have no control over the existing eCommerce platform, but you can always do an A/B test where basically you're comparing the entire site (or at least any eCommerce pages) with the existing platform vs. the new one. This should work with most split testing platforms (even Google's free website optimizer).

                      I've personally conducted such a test.
                      It's not a trivial undertaking, but it *is* possible.

                      Another option is to use a split testing solutions such as SiteSpect. They put a proxy server between your web server and the Internet, allowing you to test just about ANYTHING without changing a a single line of code :)

                      Good luck!

                      - Ophir
                      http://www.POP.us

                      --- In webanalytics@yahoogroups.com, "mr_awesome77" <jon.bovard@...> wrote:
                      >
                      > hmmmm understand...
                      >
                      > But what if I told you two things..
                      >
                      > 1. The business is a Direct marketing business.. and about 95% of its daily traffic is "new".. ie. the repeat rate of its traffic/customers is very low relative to the web universe - and especially very different to large established etailers with large % of repeat customers and traffic? (ignoring the fact that yes - new customer acqusition as a revenue model emphasis - is a very inefficient way to run a business!). Would this change the proposition of a redesign? ....given that we arent "as" concerned as we would be if we had a LOT of regular customers/visitors returning find the house has been knocked down... and somebody has built an entirely new site.
                      >
                      > 2. The business has thus far generated literally.. multiple millions of dollars business on an off the shelf ecommerce "platform" called Shopfactory. This monstrosity of a platform cannot even allow the end user/developer to move a button, change a link, resize an image, touch the underlying code - one...single.. bit.... crazy stuff. So MVT/AB testing is not even an option.
                      >
                      > Its a nightmare but also an opportunity.
                      >
                      > Yes, its a strange world we live in, but those two things above I have inherited as part of a new job :)
                      >
                      >
                      > cheers
                      > jon
                      >
                      >
                      >
                      > >
                      > > Hi Jon,
                      > >
                      > > This is a different question than the one you asked before.
                      > >
                      > > Conversion rates are very different (and often unrelated) to revenues.
                      > > It's frequently meaningless to talk about conversion rates, since they
                      > > can go up while revenues go down (being that they're a ratio).
                      > > Revenues is the right metric to be looking at when talking about
                      > > design changes. (Conversion rates are only really meaningful when
                      > > you're talking about a specific campaign, but even then, they suffer
                      > > because you haven't compensated for lead quality. But, that's a whole
                      > > other discussion.)
                      > >
                      > > I think the reason that you didn't get an immediate response is that,
                      > > in my company's research, we've found most major redesigns of e-
                      > > commerce sites result in a 20% (or greater) loss in revenues.
                      > > Organizations that redesign completely and lose 10% or less are rare
                      > > and should consider themselves very lucky.
                      > >
                      > > In the past 18 months, one of the top-20 retailers did a full redesign
                      > > of their site, only to see revenues overall drop by about 28%. (Their
                      > > highest margin, most popular products dropped revenues by 45%. Profits
                      > > sank big with the redesign.)
                      > >
                      > > A few years back, a major retailer spent $100,000,000 on a major
                      > > redesign to see a 20% reduction in sales. (Imagine what they could've
                      > > achieved had they only spent $50,000,000.) It took them 3.5 years to
                      > > bring revenues back to the level they were at before the redesign.
                      > >
                      > > I think your management is asking the right question. And I think
                      > > you'll find the right answer is to not do a redesign at all. At least,
                      > > not all at once. (Wrote about this a long time ago here: http://is.gd/3F679
                      > > It's still true today.)
                      > >
                      > > The most successful sites do incremental changes, the smaller the
                      > > better. When was the last time Amazon launched a redesign? Or Netflix?
                      > > Avid users of these sites can't tell you because the sites never have.
                      > > Instead they change on a weekly (in the case of Amazon, daily) basis.
                      > > Small changes. (Our research shows that one such small change on
                      > > Amazon's site looks like it resulted in an annual revenue increase of
                      > > $2.7 billion: http://is.gd/3F6j1)
                      > >
                      > > The key question to answer your management's question is: How much
                      > > time have you spent watching real shoppers shop for their desired
                      > > products on your site? If you're doing this effectively, you can
                      > > estimate, fairly accurately, the revenues you'll gain by specific
                      > > design changes. (We call this metric "money left on the table".)
                      > >
                      > > If a site hasn't had much attention spent to its design, there's often
                      > > low-hanging fruit that can see a good increase in revenues without
                      > > much effort. Because it's not a redesign, you can make the change
                      > > quickly and don't run the risk of the huge reductions in revenues that
                      > > redesigns typically bring. (If, for some reason you get it wrong and
                      > > do reduce the revenues as a result, because it was a small change, you
                      > > can usually roll it back easily.)
                      > >
                      > > Hope that helps,
                      > >
                      > > 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 Twitter: @jmspool
                      > >
                      > >
                      > >
                      > > [Non-text portions of this message have been removed]
                      > >
                      >
                    • lee385329
                      For what it is worth I recently conducted a complete redesign of our website. It wasn t something we considered lightly and was very nervous about carrying it
                      Message 10 of 10 , Oct 1, 2009
                        For what it is worth I recently conducted a complete redesign of our website. It wasn't something we considered lightly and was very nervous about carrying it out in light of other reports of fallen revenue. However, we did study the way our customers used the site and along with the fact that we found it difficult to implement functionality incrementally due to the very old design, we concluded that a complete redesign was the best option.

                        Our initial conversion (order conversion) dipped slightly, 7.9% to 7.4%, which we were expecting as the redesign also saw more visitors come from another of our sites (we consolidated two sites into one with the redesign), however, the conversion from users entering the basket to placing an order increased from an average of 30% to 37%.

                        You will need to look at the goals and objectives of the site and make sure your redesign allows users to complete those goals with ease. In a simplistic sense if you make it easier for the user to complete a goal then the conversion will increase, if not then it will decrease. If 95% of your traffic is "new" then I don't think you have too much of a problem with the complete redesign per se.

                        thanks
                        Lee

                        --- In webanalytics@yahoogroups.com, "mr_awesome77" <jon.bovard@...> wrote:
                        >
                        > hmmmm understand...
                        >
                        > But what if I told you two things..
                        >
                        > 1. The business is a Direct marketing business.. and about 95% of its daily traffic is "new".. ie. the repeat rate of its traffic/customers is very low relative to the web universe - and especially very different to large established etailers with large % of repeat customers and traffic? (ignoring the fact that yes - new customer acqusition as a revenue model emphasis - is a very inefficient way to run a business!). Would this change the proposition of a redesign? ....given that we arent "as" concerned as we would be if we had a LOT of regular customers/visitors returning find the house has been knocked down... and somebody has built an entirely new site.
                        >
                        > 2. The business has thus far generated literally.. multiple millions of dollars business on an off the shelf ecommerce "platform" called Shopfactory. This monstrosity of a platform cannot even allow the end user/developer to move a button, change a link, resize an image, touch the underlying code - one...single.. bit.... crazy stuff. So MVT/AB testing is not even an option.
                        >
                        > Its a nightmare but also an opportunity.
                        >
                        > Yes, its a strange world we live in, but those two things above I have inherited as part of a new job :)
                        >
                        >
                        > cheers
                        > jon
                        >
                      Your message has been successfully submitted and would be delivered to recipients shortly.