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

Help With Interpretation

Expand Messages
  • Stephanie Sims
    I have a couple questions about how to read all this data, how do I make recommendations on this data, what should I be looking for? We have the business
    Message 1 of 13 , Jun 28, 2006
      I have a couple questions about how to read all this data, how do I make recommendations on this data, what should I be looking for? We have the business goals/objectives mapped out along with KPI's and a excel report. But when my manager comes to me and says how do we increase conversion rate based on the data we have? I don't know what to say, I guess I am too new to the world of HBX and web analytics to really understand. If you have a document or have some time to talk about data interpretation and how to apply it to your business goals that would be much appreciated.

      I have posted before but really didn't get my questions answered. If any one knows how to explain this complicated, complex stuff to a complete novice I would be very grateful.

      Stephanie


      ---------------------------------
      Do you Yahoo!?
      Get on board. You're invited to try the new Yahoo! Mail Beta.

      ---------------------------------
      Yahoo! Music Unlimited - Access over 1 million songs.Try it free.

      [Non-text portions of this message have been removed]
    • Stephen Turner
      ... Stephanie, This is a great question. I would say that the key to making web analytics data useful is segmentation -- that means, looking at the same KPI
      Message 2 of 13 , Jun 29, 2006
        --- In webanalytics@yahoogroups.com, Stephanie Sims <s_simstmag@...>
        wrote:
        >
        > I have a couple questions about how to read all this data, how do
        > I make recommendations on this data, what should I be looking for?
        > We have the business goals/objectives mapped out along with KPI's
        > and a excel report. But when my manager comes to me and says how do
        > we increase conversion rate based on the data we have? I don't know
        > what to say
        >

        Stephanie,

        This is a great question. I would say that the key to making web
        analytics data useful is segmentation -- that means, looking at the
        same KPI for different groups of users and seeing which types of users
        are performing well and which are performing badly.

        For example, suppose that one of your KPIs is average revenue per
        visitor. You find that it's $3.21. That doesn't really tell you
        anything. But now, suppose that you segment according to which page
        visitors first saw on your site, and you find that some entry pages
        have a really high revenue per visitor and some stink. Now you can
        work on the bad pages, and/or redirect more visitors to the good
        pages. Or you could segment by whether or not they saw your special
        offer -- does it help or is it a waste of money? Or which
        pay-per-click ads they came from -- are some ads costing you lots of
        money but not converting? Or by which referrer they came from, or
        which search words they searched on, or what time of day they arrived,
        or any number of other things. Suddenly you haven't just got a number,
        you've got an understanding of your different types of visitors, what
        they're trying to accomplish, and which visitors are good for your
        business and which ones aren't. Then you can concentrate on attracting
        more of the good visitors, and not attracting the bad ones, or else
        satisfying the bad visitors' desires to turn them into good visitors.

        My second tip is to read these posts on Avinash Kaushik's blog:
        http://www.kaushik.net/avinash/2006/05/excellent-analytics-tip2-segment-absolutely-everything.html
        http://www.kaushik.net/avinash/2006/06/tips-for-web-analytics-success-for-small-businesses.html
        And my third tip is to go back and read everything else on his blog.
        There are loads of good tips and tricks in there.

        --
        Stephen Turner
        CTO, ClickTracks http://www.clicktracks.com/
        WINNER: ClickZ Best Web Analytics Tool 2003, 2004 & 2005
      • matpflum
        Stephanie, Besides reading Eric Peterson s excellent books Web Analytics Demystified and Web Analytics Measurement Hacks , I found it helpful to direct my
        Message 3 of 13 , Jun 29, 2006
          Stephanie,
          Besides reading Eric Peterson's excellent books "Web Analytics
          Demystified" and "Web Analytics Measurement Hacks", I found it
          helpful to direct my staff to
          http://www.bruceclay.com/web_analytics.htm for an overview. There
          may be other overviews like this, but this is the first one that came
          to mind. But these don't really explain how to be an analyst, which
          is what your manager is asking you to be. Analysis is a skill beyond
          manipulating web analytics data. It's been my experience that good
          analysts have been trained with graduate school classes in research
          and statistics, but I am sure there are some natural analysts out
          there. I am also asking my web analytics staff to step into true
          analyst roles, and so have had to set up a series of analytics labs
          for them with case studies. In short, I would advise you to
          construct hypotheses about why the data is the way it is, and then
          test those hypotheses. This will lead down a lot of dead ends, but
          in the process I hope you will begin to see the bigger picture of
          influencers on conversions or whatever.
          Maureen Pflum
          Interactivate, Inc.

          --- In webanalytics@yahoogroups.com, Stephanie Sims <s_simstmag@...>
          wrote:
          >
          > I have a couple questions about how to read all this data, how
          do I make recommendations on this data, what should I be looking for?
          We have the business goals/objectives mapped out along with KPI's and
          a excel report. But when my manager comes to me and says how do we
          increase conversion rate based on the data we have? I don't know what
          to say, I guess I am too new to the world of HBX and web analytics to
          really understand. If you have a document or have some time to talk
          about data interpretation and how to apply it to your business goals
          that would be much appreciated.
          >
          > I have posted before but really didn't get my questions
          answered. If any one knows how to explain this complicated, complex
          stuff to a complete novice I would be very grateful.
          >
          > Stephanie
          >
          >
          > ---------------------------------
          > Do you Yahoo!?
          > Get on board. You're invited to try the new Yahoo! Mail Beta.
          >
          > ---------------------------------
          > Yahoo! Music Unlimited - Access over 1 million songs.Try it free.
          >
          > [Non-text portions of this message have been removed]
          >
        • Stephanie Sims
          Thank you Maureen for your advice. I think building hypotheses on the data and why it is the way it is will be a good start. Like you said I will go through
          Message 4 of 13 , Jun 29, 2006
            Thank you Maureen for your advice. I think building hypotheses on the data and why it is the way it is will be a good start. Like you said I will go through trial and error but I think this will give me a little more insight. Thanks

            Stephanie

            matpflum <matpflum@...> wrote:
            Stephanie,
            Besides reading Eric Peterson's excellent books "Web Analytics
            Demystified" and "Web Analytics Measurement Hacks", I found it
            helpful to direct my staff to
            http://www.bruceclay.com/web_analytics.htm for an overview. There
            may be other overviews like this, but this is the first one that came
            to mind. But these don't really explain how to be an analyst, which
            is what your manager is asking you to be. Analysis is a skill beyond
            manipulating web analytics data. It's been my experience that good
            analysts have been trained with graduate school classes in research
            and statistics, but I am sure there are some natural analysts out
            there. I am also asking my web analytics staff to step into true
            analyst roles, and so have had to set up a series of analytics labs
            for them with case studies. In short, I would advise you to
            construct hypotheses about why the data is the way it is, and then
            test those hypotheses. This will lead down a lot of dead ends, but
            in the process I hope you will begin to see the bigger picture of
            influencers on conversions or whatever.
            Maureen Pflum
            Interactivate, Inc.

            --- In webanalytics@yahoogroups.com, Stephanie Sims <s_simstmag@...>
            wrote:
            >
            > I have a couple questions about how to read all this data, how
            do I make recommendations on this data, what should I be looking for?
            We have the business goals/objectives mapped out along with KPI's and
            a excel report. But when my manager comes to me and says how do we
            increase conversion rate based on the data we have? I don't know what
            to say, I guess I am too new to the world of HBX and web analytics to
            really understand. If you have a document or have some time to talk
            about data interpretation and how to apply it to your business goals
            that would be much appreciated.
            >
            > I have posted before but really didn't get my questions
            answered. If any one knows how to explain this complicated, complex
            stuff to a complete novice I would be very grateful.
            >
            > Stephanie
            >
            >
            > ---------------------------------
            > Do you Yahoo!?
            > Get on board. You're invited to try the new Yahoo! Mail Beta.
            >
            > ---------------------------------
            > Yahoo! Music Unlimited - Access over 1 million songs.Try it free.
            >
            > [Non-text portions of this message have been removed]
            >






            ---------------------------------
            Want to be your own boss? Learn how on Yahoo! Small Business.

            [Non-text portions of this message have been removed]
          • Avery Cohen
            Interesting question, Stephanie: How do we increase conversion rate based on the data we have? We have the business goals/objectives mapped out along with
            Message 5 of 13 , Jun 29, 2006
              Interesting question, Stephanie: "How do we increase conversion rate based
              on the data we have?"

              "We have the business goals/objectives mapped out along with KPI's and a
              excel report. ...I am new to the world of HBX."

              We might simplify and say: analysts basically have two tools to bring
              insight and one tool for action.

              The two tools for insight could be called "Process Decompisition" and
              "Segmentation". They will help you see "what is" and frame "what is
              possible".

              With "Process Decomposition", you can break down "conversion" into a series
              of processes. I don't know what you want to convert, which would make it
              easier for us to discuss specifics -- are you looking to sell, generate
              leads, or create opt-ins? We might combine "Process Decomposition" and
              "Segmentation" right from the start. Perhaps we could look at the
              acquisition process -- how are people reaching your web site (what are your
              referral sources)? This is Segmentation by Source. Where are they landing?
              This is mapping the Process Start Point and Segmenting the process by
              Landing Page. How many visits from each source are single-page-view visits?
              How many of the multi-page visits are "converted" to one of your success
              measures?

              Your tool for action is Testing Changes. After all, the only way to actually
              "increase conversion rate" is to change something. One might ask: "What can
              we change to get more multi-page visits?" You'll get more ideas if you are
              clear about what is the first page people are seeing and how they got to
              that page. One might ask: "Which of our Acquisition Sources are most
              effective?" and go down a different path.

              You can use HBX features to see visit characteristics and behavior based on
              referral source.

              But I'm already starting to make assumptions about your "business
              goals/objectives" and what "conversion" means to you. Perhaps this posting
              will help you get started using web analytics (and your HBX subscription)
              more effectivly. Feel free to respond with more specifics.

              Avery.
              --
              Avery J. Cohen
              Principal
              Metrist Partners
              http://www.metrist.com


              [Non-text portions of this message have been removed]
            • Stephanie Sims
              Avery, Thank you... Our site sells Performance Sports Apparel and Footwear. So one of our business goals is to increase reveune by 15% over last year. So in
              Message 6 of 13 , Jun 29, 2006
                Avery,

                Thank you... Our site sells Performance Sports Apparel and Footwear. So one of our business goals is to increase reveune by 15% over last year. So in doing so we have to convert our visitors more often, acquire more new visitors and make our returning visitors buy more per purchase.

                One of the most challenging questions that I hear on a daily basis is why is our site performing so badly today compared to yesterday or the day before? From here (tell me if I am wasting time or looking into the right data) I will then go and analyze the single access pages, exit pages and get a total of all of our visits and compare them to these other metrics. For example: he wanted to know why yesterday was so bad and after digging around I found that most people were going to out footwear pages and then to our flagship shoe. When the consumer got there they would leave or exit all together due to the product not being available, no inventory.

                So I am learning but it is just taking me time. Please give me anymore advice or tips that you may have.

                Thanks

                Stephanie

                Avery Cohen <avery.cohen@...> wrote:
                Interesting question, Stephanie: "How do we increase conversion rate based
                on the data we have?"

                "We have the business goals/objectives mapped out along with KPI's and a
                excel report. ...I am new to the world of HBX."

                We might simplify and say: analysts basically have two tools to bring
                insight and one tool for action.

                The two tools for insight could be called "Process Decompisition" and
                "Segmentation". They will help you see "what is" and frame "what is
                possible".

                With "Process Decomposition", you can break down "conversion" into a series
                of processes. I don't know what you want to convert, which would make it
                easier for us to discuss specifics -- are you looking to sell, generate
                leads, or create opt-ins? We might combine "Process Decomposition" and
                "Segmentation" right from the start. Perhaps we could look at the
                acquisition process -- how are people reaching your web site (what are your
                referral sources)? This is Segmentation by Source. Where are they landing?
                This is mapping the Process Start Point and Segmenting the process by
                Landing Page. How many visits from each source are single-page-view visits?
                How many of the multi-page visits are "converted" to one of your success
                measures?

                Your tool for action is Testing Changes. After all, the only way to actually
                "increase conversion rate" is to change something. One might ask: "What can
                we change to get more multi-page visits?" You'll get more ideas if you are
                clear about what is the first page people are seeing and how they got to
                that page. One might ask: "Which of our Acquisition Sources are most
                effective?" and go down a different path.

                You can use HBX features to see visit characteristics and behavior based on
                referral source.

                But I'm already starting to make assumptions about your "business
                goals/objectives" and what "conversion" means to you. Perhaps this posting
                will help you get started using web analytics (and your HBX subscription)
                more effectivly. Feel free to respond with more specifics.

                Avery.
                --
                Avery J. Cohen
                Principal
                Metrist Partners
                http://www.metrist.com

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






                ---------------------------------
                How low will we go? Check out Yahoo! Messenger’s low PC-to-Phone call rates.

                [Non-text portions of this message have been removed]
              • Ken
                I m not sure what kind of site you have, but assuming you are competitive with your product or service offering, I firmly believe that improving conversion is
                Message 7 of 13 , Jun 29, 2006
                  I'm not sure what kind of site you have, but assuming you are competitive with your product or service offering, I firmly believe that improving conversion is all about working to improve the visitors experience which relates to site design and usability.

                  Look at clickstream data, process and product abandonment to find out where your visitors are leaving you and then prioritize where to spend your efforts. Test, evaluate, make changes and retest. This is a continuous process.

                  Some good newsletters:

                  http://www.ecommerce-guide.com/
                  www.thegrok.com
                  http://www.internetretailer.com/


                  Ken Mitchell
                  Hanover Direct

                  ephanie Sims <s_simstmag@...> wrote:
                  I have a couple questions about how to read all this data, how do I make recommendations on this data, what should I be looking for? We have the business goals/objectives mapped out along with KPI's and a excel report. But when my manager comes to me and says how do we increase conversion rate based on the data we have? I don't know what to say, I guess I am too new to the world of HBX and web analytics to really understand. If you have a document or have some time to talk about data interpretation and how to apply it to your business goals that would be much appreciated.

                  I have posted before but really didn't get my questions answered. If any one knows how to explain this complicated, complex stuff to a complete novice I would be very grateful.

                  Stephanie


                  ---------------------------------
                  Do you Yahoo!?
                  Get on board. You're invited to try the new Yahoo! Mail Beta.

                  ---------------------------------
                  Yahoo! Music Unlimited - Access over 1 million songs.Try it free.

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





                  __________________________________________________
                  Do You Yahoo!?
                  Tired of spam? Yahoo! Mail has the best spam protection around
                  http://mail.yahoo.com

                  [Non-text portions of this message have been removed]
                • sfernandez0013
                  That s a pretty tall order, but is the exact question that everyone asks when first looking at web data, especially if you re using an analytics package.
                  Message 8 of 13 , Jun 29, 2006
                    That's a pretty tall order, but is the exact question that everyone
                    asks when first looking at web data, especially if you're using an
                    analytics package. You're sold on the idea that the package will
                    provide you all the answers. The truth is, it's just the means to
                    ask questions. What makes a good analyst is understanding the
                    situation to know what question to ask and how to ask it, asking
                    lots of questions. "What gives rise," a former professor used to
                    tell us.

                    While I can't answer your question directly, I'll give this bit of
                    wisdom. There usually isn't a magic bullet to increase conversion.
                    Not unless something is specifically broken. Your visitor will
                    usually tell you when that's the case. So, you start looking at
                    processes / funnels on your site. You use the data to start
                    measuring how visitors move through the site. You target were there
                    appear to be bottlenecks, and while the data can't tell you what to
                    do; it should give your own instincts a good hint on where to dig
                    further or maybe what to do next. The data can also take a while
                    before it says anything. I try not to offer analysis unless I have
                    at least a month of data. I work in retail and there are just too
                    many variables per week to take less than 4 weeks to say anything.
                    A quarter would be better, but manage can't wait that long. But, it
                    does take time to build data to see patterns. Without baselines and
                    trends, how can you know anything you do has real impact or just
                    fluke timing with a random event / variable.

                    Steve Fernandez


                    --- In webanalytics@yahoogroups.com, Stephanie Sims <s_simstmag@...>
                    wrote:
                    >
                    > I have a couple questions about how to read all this data, how
                    do I make recommendations on this data, what should I be looking
                    for? We have the business goals/objectives mapped out along with
                    KPI's and a excel report. But when my manager comes to me and says
                    how do we increase conversion rate based on the data we have? I
                    don't know what to say, I guess I am too new to the world of HBX and
                    web analytics to really understand. If you have a document or have
                    some time to talk about data interpretation and how to apply it to
                    your business goals that would be much appreciated.
                    >
                    > I have posted before but really didn't get my questions
                    answered. If any one knows how to explain this complicated, complex
                    stuff to a complete novice I would be very grateful.
                    >
                    > Stephanie
                    >
                    >
                    > ---------------------------------
                    > Do you Yahoo!?
                    > Get on board. You're invited to try the new Yahoo! Mail Beta.
                    >
                    > ---------------------------------
                    > Yahoo! Music Unlimited - Access over 1 million songs.Try it free.
                    >
                    > [Non-text portions of this message have been removed]
                    >
                  • Ken
                    Hi: Your example is a perfect one. Lack of inventory kills conversion. If I can t find the product I want on your site, it is pretty easy for me to find it
                    Message 9 of 13 , Jun 29, 2006
                      Hi:

                      Your example is a perfect one. Lack of inventory kills conversion. If I can't find the product I want on your site, it is pretty easy for me to find it somewhere else. It also may leave a bad impression on site visitors.

                      Obsessing over daily changes in consumer behavior is not productive. Better to aggregate data over time. This month vs. last month or trends that emerge over time are more important indicators.

                      You should try to manage the expectations of your boss while you work out KPI's and A/B test changes to your site. Are they looking for scapegoats in the site design when the real problem may be in the merchandising/buying/inventory area?

                      Avery made some great comments. Assuming the merchandise decisions are out of you control, you can only focus on making the site functions as clean and user friendly as they can be. Give the product abandonment info to your merchants to help them with their buying decisions, but focus your efforts on site processes. Is there a way to de-emphasize product that may be out of stock and feature other popular items that you have?

                      Ken


                      Stephanie Sims <s_simstmag@...> wrote:
                      Avery,

                      Thank you... Our site sells Performance Sports Apparel and Footwear. So one of our business goals is to increase reveune by 15% over last year. So in doing so we have to convert our visitors more often, acquire more new visitors and make our returning visitors buy more per purchase.

                      One of the most challenging questions that I hear on a daily basis is why is our site performing so badly today compared to yesterday or the day before? From here (tell me if I am wasting time or looking into the right data) I will then go and analyze the single access pages, exit pages and get a total of all of our visits and compare them to these other metrics. For example: he wanted to know why yesterday was so bad and after digging around I found that most people were going to out footwear pages and then to our flagship shoe. When the consumer got there they would leave or exit all together due to the product not being available, no inventory.

                      So I am learning but it is just taking me time. Please give me anymore advice or tips that you may have.

                      Thanks

                      Stephanie

                      Avery Cohen <avery.cohen@...> wrote:
                      Interesting question, Stephanie: "How do we increase conversion rate based
                      on the data we have?"

                      "We have the business goals/objectives mapped out along with KPI's and a
                      excel report. ...I am new to the world of HBX."

                      We might simplify and say: analysts basically have two tools to bring
                      insight and one tool for action.

                      The two tools for insight could be called "Process Decompisition" and
                      "Segmentation". They will help you see "what is" and frame "what is
                      possible".

                      With "Process Decomposition", you can break down "conversion" into a series
                      of processes. I don't know what you want to convert, which would make it
                      easier for us to discuss specifics -- are you looking to sell, generate
                      leads, or create opt-ins? We might combine "Process Decomposition" and
                      "Segmentation" right from the start. Perhaps we could look at the
                      acquisition process -- how are people reaching your web site (what are your
                      referral sources)? This is Segmentation by Source. Where are they landing?
                      This is mapping the Process Start Point and Segmenting the process by
                      Landing Page. How many visits from each source are single-page-view visits?
                      How many of the multi-page visits are "converted" to one of your success
                      measures?

                      Your tool for action is Testing Changes. After all, the only way to actually
                      "increase conversion rate" is to change something. One might ask: "What can
                      we change to get more multi-page visits?" You'll get more ideas if you are
                      clear about what is the first page people are seeing and how they got to
                      that page. One might ask: "Which of our Acquisition Sources are most
                      effective?" and go down a different path.

                      You can use HBX features to see visit characteristics and behavior based on
                      referral source.

                      But I'm already starting to make assumptions about your "business
                      goals/objectives" and what "conversion" means to you. Perhaps this posting
                      will help you get started using web analytics (and your HBX subscription)
                      more effectivly. Feel free to respond with more specifics.

                      Avery.
                      --
                      Avery J. Cohen
                      Principal
                      Metrist Partners
                      http://www.metrist.com

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


                      ---------------------------------
                      How low will we go? Check out Yahoo! Messenger’s low PC-to-Phone call rates.

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





                      __________________________________________________
                      Do You Yahoo!?
                      Tired of spam? Yahoo! Mail has the best spam protection around
                      http://mail.yahoo.com

                      [Non-text portions of this message have been removed]
                    • Michael C. Samec
                      Stephanie: Perhaps tracking along the following would help; This Monday vs. Last Monday This, the last Monday in June vs. Last Monday in June, 2005 - % change
                      Message 10 of 13 , Jun 30, 2006
                        Stephanie:

                        Perhaps tracking along the following would help;

                        This Monday vs. Last Monday
                        This, the last Monday in June vs. Last Monday in June, 2005 - % change up or
                        down

                        Then compare this week against last week and 'last week of June 2006'
                        against 'last week of June 2005'

                        Finally, June '06 vs. June '05

                        You'll get a perspective of your businesses "normal" ebb and flow. You may
                        find that a shift of 10% or even more might simply occur for "no reason"
                        (Okay, for reasons beyond the ability of your analytics tools to ID).

                        The next logical question is, 'what happened last day/week/year and
                        yesterday/this week/this year that affected results?' Here is where keeping
                        a diary of the changes to your site, the ads run, noticeable changes in what
                        competitors are doing, external market influence (war, changes in
                        legislation, etc) might have also had an impact.

                        Reviewing the timeline trends, you'll develop an intuition for what is
                        "normal" for your biz and will gain confidence in what is going on, how to
                        respond and where to look for possible flaws and/or new opportunities.

                        Review this regularly with this manager - He/she will

                        also get a gut feel of what is normal
                        gain confidence that you are "on the ball"

                        And since we are coming up on a (U.S.) holiday, now is a good time to remind
                        all of us - don't forget to make note of when a holiday occurs. I was
                        scratching my head at the end of March trying to figure out why traffic
                        spiked over last year when I remembered that Easter fell at the end of March
                        last year (doh...!).
                        _____

                        From: webanalytics@yahoogroups.com [mailto:webanalytics@yahoogroups.com] On
                        Behalf Of Stephanie Sims
                        Sent: Thursday, June 29, 2006 7:36 PM
                        To: webanalytics@yahoogroups.com
                        Subject: Re: [webanalytics] Help With Interpretation



                        Avery,

                        Thank you... Our site sells Performance Sports Apparel and Footwear. So one
                        of our business goals is to increase reveune by 15% over last year. So in
                        doing so we have to convert our visitors more often, acquire more new
                        visitors and make our returning visitors buy more per purchase.

                        One of the most challenging questions that I hear on a daily basis is why is
                        our site performing so badly today compared to yesterday or the day before?
                        From here (tell me if I am wasting time or looking into the right data) I
                        will then go and analyze the single access pages, exit pages and get a total
                        of all of our visits and compare them to these other metrics. For example:
                        he wanted to know why yesterday was so bad and after digging around I found
                        that most people were going to out footwear pages and then to our flagship
                        shoe. When the consumer got there they would leave or exit all together due
                        to the product not being available, no inventory.

                        So I am learning but it is just taking me time. Please give me anymore
                        advice or tips that you may have.

                        Thanks

                        Stephanie

                        Avery Cohen <avery.cohen@ <mailto:avery.cohen%40gmail.com> gmail.com> wrote:
                        Interesting question, Stephanie: "How do we increase conversion rate based
                        on the data we have?"

                        "We have the business goals/objectives mapped out along with KPI's and a
                        excel report. ...I am new to the world of HBX."

                        We might simplify and say: analysts basically have two tools to bring
                        insight and one tool for action.

                        The two tools for insight could be called "Process Decompisition" and
                        "Segmentation". They will help you see "what is" and frame "what is
                        possible".

                        With "Process Decomposition", you can break down "conversion" into a series
                        of processes. I don't know what you want to convert, which would make it
                        easier for us to discuss specifics -- are you looking to sell, generate
                        leads, or create opt-ins? We might combine "Process Decomposition" and
                        "Segmentation" right from the start. Perhaps we could look at the
                        acquisition process -- how are people reaching your web site (what are your
                        referral sources)? This is Segmentation by Source. Where are they landing?
                        This is mapping the Process Start Point and Segmenting the process by
                        Landing Page. How many visits from each source are single-page-view visits?
                        How many of the multi-page visits are "converted" to one of your success
                        measures?

                        Your tool for action is Testing Changes. After all, the only way to actually
                        "increase conversion rate" is to change something. One might ask: "What can
                        we change to get more multi-page visits?" You'll get more ideas if you are
                        clear about what is the first page people are seeing and how they got to
                        that page. One might ask: "Which of our Acquisition Sources are most
                        effective?" and go down a different path.

                        You can use HBX features to see visit characteristics and behavior based on
                        referral source.

                        But I'm already starting to make assumptions about your "business
                        goals/objectives" and what "conversion" means to you. Perhaps this posting
                        will help you get started using web analytics (and your HBX subscription)
                        more effectivly. Feel free to respond with more specifics.

                        Avery.
                        --
                        Avery J. Cohen
                        Principal
                        Metrist Partners
                        http://www.metrist <http://www.metrist.com> com

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


                        ---------------------------------
                        How low will we go? Check out Yahoo! Messenger's low PC-to-Phone call rates.

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






                        [Non-text portions of this message have been removed]
                      • Avery Cohen
                        Glad I could help, Stephanie. One of the tricks of the trade is to keep an eye on the high-level goals while getting more and more specific with our questions.
                        Message 11 of 13 , Jun 30, 2006
                          Glad I could help, Stephanie.

                          One of the tricks of the trade is to keep an eye on the high-level goals
                          while getting more and more specific with our questions. Focusing too much
                          on daily sales fluctuations can easily get you looking at trees and
                          forgetting the forrest. However, you have already found one problem:
                          "Unavailable Product Kills Visits". Great job!

                          Why were so many people looking at shoes that day? (Was that traffic pattern
                          actually unique?) Do you have a weekly "shoes" email? Are you buying terms
                          dealing with footwear? If so, you might consider refocusing your messaging
                          and spending when you don't have your flagship product (ouch!)

                          You said "why was yesterday so bad?" Based on your description, my
                          understanding is that traffic wasn't bad, the problem was the "conversion",
                          or purchase rate. If I was looking into daily traffic and conversion rate
                          fluctuation, I'd focus on visit sources first and foremost. If there's a lot
                          of variation, I'd coordinate with whoever is managing the emails and paid
                          search spend for those days and match that information, day-for-day. I would
                          expect this to quickly link actions and results. Complete the feedback with
                          daily sales by product category for those days. I think that will give you
                          and the marketing manager (assuming it's at least one other person) lots to
                          work with as far as testing ways to drive "invited" visits that are
                          effective at selling items you have in stock.

                          After that (and only after that), I'd got to HBX Content Analysis and look
                          at landing pages and multi-page visit drivers. We recommend (and, to be
                          fair, we may have picked this up from Peterson) that our clients invert the
                          usual statistic (single-page visits) and manage the (usually) lower number
                          -- multi-page visits. That way you are always trying to raise the KPI.

                          Finally, I would focus significant efforts on email -- measuring and
                          communicating the effectiveness of emails sent and getting people to opt-in
                          to our email list. Only because it gives you the most control for the least
                          expense. Lessons learned from inexpensive testing of email can be applied to
                          search.

                          Be sure to write back and let us know what is working for you!

                          Avery.
                          --
                          Avery J. Cohen
                          Principal
                          Metrist Partners
                          http://www.metrist.com


                          On 6/29/06, Stephanie Sims <s_simstmag@...> wrote:
                          >
                          > Avery,
                          >
                          > Thank you... Our site sells Performance Sports Apparel and Footwear. So
                          > one of our business goals is to increase reveune by 15% over last year. So
                          > in doing so we have to convert our visitors more often, acquire more new
                          > visitors and make our returning visitors buy more per purchase.
                          >
                          > One of the most challenging questions that I hear on a daily basis is why
                          > is our site performing so badly today compared to yesterday or the day
                          > before? From here (tell me if I am wasting time or looking into the right
                          > data) I will then go and analyze the single access pages, exit pages and get
                          > a total of all of our visits and compare them to these other metrics. For
                          > example: he wanted to know why yesterday was so bad and after digging around
                          > I found that most people were going to out footwear pages and then to our
                          > flagship shoe. When the consumer got there they would leave or exit all
                          > together due to the product not being available, no inventory.
                          >
                          > So I am learning but it is just taking me time. Please give me anymore
                          > advice or tips that you may have.
                          >
                          > Thanks
                          >
                          > Stephanie
                          >
                          >
                          > Avery Cohen <avery.cohen@... <avery.cohen%40gmail.com>> wrote:
                          > Interesting question, Stephanie: "How do we increase conversion rate based
                          > on the data we have?"
                          >
                          > "We have the business goals/objectives mapped out along with KPI's and a
                          > excel report. ...I am new to the world of HBX."
                          >
                          > We might simplify and say: analysts basically have two tools to bring
                          > insight and one tool for action.
                          >
                          > The two tools for insight could be called "Process Decompisition" and
                          > "Segmentation". They will help you see "what is" and frame "what is
                          > possible".
                          >
                          > With "Process Decomposition", you can break down "conversion" into a
                          > series
                          > of processes. I don't know what you want to convert, which would make it
                          > easier for us to discuss specifics -- are you looking to sell, generate
                          > leads, or create opt-ins? We might combine "Process Decomposition" and
                          > "Segmentation" right from the start. Perhaps we could look at the
                          > acquisition process -- how are people reaching your web site (what are
                          > your
                          > referral sources)? This is Segmentation by Source. Where are they landing?
                          > This is mapping the Process Start Point and Segmenting the process by
                          > Landing Page. How many visits from each source are single-page-view
                          > visits?
                          > How many of the multi-page visits are "converted" to one of your success
                          > measures?
                          >
                          > Your tool for action is Testing Changes. After all, the only way to
                          > actually
                          > "increase conversion rate" is to change something. One might ask: "What
                          > can
                          > we change to get more multi-page visits?" You'll get more ideas if you are
                          > clear about what is the first page people are seeing and how they got to
                          > that page. One might ask: "Which of our Acquisition Sources are most
                          > effective?" and go down a different path.
                          >
                          > You can use HBX features to see visit characteristics and behavior based
                          > on
                          > referral source.
                          >
                          > But I'm already starting to make assumptions about your "business
                          > goals/objectives" and what "conversion" means to you. Perhaps this posting
                          > will help you get started using web analytics (and your HBX subscription)
                          > more effectivly. Feel free to respond with more specifics.
                          >
                          > Avery.
                          > --
                          > Avery J. Cohen
                          > Principal
                          > Metrist Partners
                          > http://www.metrist.com
                          >


                          [Non-text portions of this message have been removed]
                        • Stephanie Sims
                          Avery, I think you gave me great information and I will indeed write back and let you know how your reommendations are working for me! Thanks Stephanie Avery
                          Message 12 of 13 , Jun 30, 2006
                            Avery,

                            I think you gave me great information and I will indeed write back and let you know how your reommendations are working for me! Thanks

                            Stephanie

                            Avery Cohen <avery.cohen@...> wrote:
                            Glad I could help, Stephanie.

                            One of the tricks of the trade is to keep an eye on the high-level goals
                            while getting more and more specific with our questions. Focusing too much
                            on daily sales fluctuations can easily get you looking at trees and
                            forgetting the forrest. However, you have already found one problem:
                            "Unavailable Product Kills Visits". Great job!

                            Why were so many people looking at shoes that day? (Was that traffic pattern
                            actually unique?) Do you have a weekly "shoes" email? Are you buying terms
                            dealing with footwear? If so, you might consider refocusing your messaging
                            and spending when you don't have your flagship product (ouch!)

                            You said "why was yesterday so bad?" Based on your description, my
                            understanding is that traffic wasn't bad, the problem was the "conversion",
                            or purchase rate. If I was looking into daily traffic and conversion rate
                            fluctuation, I'd focus on visit sources first and foremost. If there's a lot
                            of variation, I'd coordinate with whoever is managing the emails and paid
                            search spend for those days and match that information, day-for-day. I would
                            expect this to quickly link actions and results. Complete the feedback with
                            daily sales by product category for those days. I think that will give you
                            and the marketing manager (assuming it's at least one other person) lots to
                            work with as far as testing ways to drive "invited" visits that are
                            effective at selling items you have in stock.

                            After that (and only after that), I'd got to HBX Content Analysis and look
                            at landing pages and multi-page visit drivers. We recommend (and, to be
                            fair, we may have picked this up from Peterson) that our clients invert the
                            usual statistic (single-page visits) and manage the (usually) lower number
                            -- multi-page visits. That way you are always trying to raise the KPI.

                            Finally, I would focus significant efforts on email -- measuring and
                            communicating the effectiveness of emails sent and getting people to opt-in
                            to our email list. Only because it gives you the most control for the least
                            expense. Lessons learned from inexpensive testing of email can be applied to
                            search.

                            Be sure to write back and let us know what is working for you!

                            Avery.
                            --
                            Avery J. Cohen
                            Principal
                            Metrist Partners
                            http://www.metrist.com

                            On 6/29/06, Stephanie Sims <s_simstmag@...> wrote:
                            >
                            > Avery,
                            >
                            > Thank you... Our site sells Performance Sports Apparel and Footwear. So
                            > one of our business goals is to increase reveune by 15% over last year. So
                            > in doing so we have to convert our visitors more often, acquire more new
                            > visitors and make our returning visitors buy more per purchase.
                            >
                            > One of the most challenging questions that I hear on a daily basis is why
                            > is our site performing so badly today compared to yesterday or the day
                            > before? From here (tell me if I am wasting time or looking into the right
                            > data) I will then go and analyze the single access pages, exit pages and get
                            > a total of all of our visits and compare them to these other metrics. For
                            > example: he wanted to know why yesterday was so bad and after digging around
                            > I found that most people were going to out footwear pages and then to our
                            > flagship shoe. When the consumer got there they would leave or exit all
                            > together due to the product not being available, no inventory.
                            >
                            > So I am learning but it is just taking me time. Please give me anymore
                            > advice or tips that you may have.
                            >
                            > Thanks
                            >
                            > Stephanie
                            >
                            >
                            > Avery Cohen <avery.cohen@... <avery.cohen%40gmail.com>> wrote:
                            > Interesting question, Stephanie: "How do we increase conversion rate based
                            > on the data we have?"
                            >
                            > "We have the business goals/objectives mapped out along with KPI's and a
                            > excel report. ...I am new to the world of HBX."
                            >
                            > We might simplify and say: analysts basically have two tools to bring
                            > insight and one tool for action.
                            >
                            > The two tools for insight could be called "Process Decompisition" and
                            > "Segmentation". They will help you see "what is" and frame "what is
                            > possible".
                            >
                            > With "Process Decomposition", you can break down "conversion" into a
                            > series
                            > of processes. I don't know what you want to convert, which would make it
                            > easier for us to discuss specifics -- are you looking to sell, generate
                            > leads, or create opt-ins? We might combine "Process Decomposition" and
                            > "Segmentation" right from the start. Perhaps we could look at the
                            > acquisition process -- how are people reaching your web site (what are
                            > your
                            > referral sources)? This is Segmentation by Source. Where are they landing?
                            > This is mapping the Process Start Point and Segmenting the process by
                            > Landing Page. How many visits from each source are single-page-view
                            > visits?
                            > How many of the multi-page visits are "converted" to one of your success
                            > measures?
                            >
                            > Your tool for action is Testing Changes. After all, the only way to
                            > actually
                            > "increase conversion rate" is to change something. One might ask: "What
                            > can
                            > we change to get more multi-page visits?" You'll get more ideas if you are
                            > clear about what is the first page people are seeing and how they got to
                            > that page. One might ask: "Which of our Acquisition Sources are most
                            > effective?" and go down a different path.
                            >
                            > You can use HBX features to see visit characteristics and behavior based
                            > on
                            > referral source.
                            >
                            > But I'm already starting to make assumptions about your "business
                            > goals/objectives" and what "conversion" means to you. Perhaps this posting
                            > will help you get started using web analytics (and your HBX subscription)
                            > more effectivly. Feel free to respond with more specifics.
                            >
                            > Avery.
                            > --
                            > Avery J. Cohen
                            > Principal
                            > Metrist Partners
                            > http://www.metrist.com
                            >

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






                            ---------------------------------
                            How low will we go? Check out Yahoo! Messenger’s low PC-to-Phone call rates.

                            ---------------------------------
                            How low will we go? Check out Yahoo! Messenger’s low PC-to-Phone call rates.

                            [Non-text portions of this message have been removed]
                          • kenn gold
                            Hey Stephanie, I work for an online publisher, so not sure if an e-commerce site like yours would work the same way, but I am betting it does: When comparing
                            Message 13 of 13 , Jun 30, 2006
                              Hey Stephanie,

                              I work for an online publisher, so not sure if an
                              e-commerce site like yours would work the same way,
                              but I am betting it does:

                              When comparing traffic/conversion trends, teach your
                              company to not look day after day, but day over day.

                              For me, a Monday and a Tuesday are never going to be
                              comparable. So if I want to know why yesterday seemed
                              so bad or good, I look at it in comparison to previous
                              Thursdays, not the day before.

                              Kenn Gold
                              Director of Reporting and Analysis
                              TheStreet.com

                              --- Stephanie Sims <s_simstmag@...> wrote:

                              > Avery,
                              >
                              > I think you gave me great information and I will
                              > indeed write back and let you know how your
                              > reommendations are working for me! Thanks
                              >
                              > Stephanie
                              >
                              > Avery Cohen <avery.cohen@...> wrote:
                              > Glad I could help, Stephanie.
                              >
                              > One of the tricks of the trade is to keep an eye on
                              > the high-level goals
                              > while getting more and more specific with our
                              > questions. Focusing too much
                              > on daily sales fluctuations can easily get you
                              > looking at trees and
                              > forgetting the forrest. However, you have already
                              > found one problem:
                              > "Unavailable Product Kills Visits". Great job!
                              >
                              > Why were so many people looking at shoes that day?
                              > (Was that traffic pattern
                              > actually unique?) Do you have a weekly "shoes"
                              > email? Are you buying terms
                              > dealing with footwear? If so, you might consider
                              > refocusing your messaging
                              > and spending when you don't have your flagship
                              > product (ouch!)
                              >
                              > You said "why was yesterday so bad?" Based on your
                              > description, my
                              > understanding is that traffic wasn't bad, the
                              > problem was the "conversion",
                              > or purchase rate. If I was looking into daily
                              > traffic and conversion rate
                              > fluctuation, I'd focus on visit sources first and
                              > foremost. If there's a lot
                              > of variation, I'd coordinate with whoever is
                              > managing the emails and paid
                              > search spend for those days and match that
                              > information, day-for-day. I would
                              > expect this to quickly link actions and results.
                              > Complete the feedback with
                              > daily sales by product category for those days. I
                              > think that will give you
                              > and the marketing manager (assuming it's at least
                              > one other person) lots to
                              > work with as far as testing ways to drive "invited"
                              > visits that are
                              > effective at selling items you have in stock.
                              >
                              > After that (and only after that), I'd got to HBX
                              > Content Analysis and look
                              > at landing pages and multi-page visit drivers. We
                              > recommend (and, to be
                              > fair, we may have picked this up from Peterson) that
                              > our clients invert the
                              > usual statistic (single-page visits) and manage the
                              > (usually) lower number
                              > -- multi-page visits. That way you are always trying
                              > to raise the KPI.
                              >
                              > Finally, I would focus significant efforts on email
                              > -- measuring and
                              > communicating the effectiveness of emails sent and
                              > getting people to opt-in
                              > to our email list. Only because it gives you the
                              > most control for the least
                              > expense. Lessons learned from inexpensive testing of
                              > email can be applied to
                              > search.
                              >
                              > Be sure to write back and let us know what is
                              > working for you!
                              >
                              > Avery.
                              > --
                              > Avery J. Cohen
                              > Principal
                              > Metrist Partners
                              > http://www.metrist.com
                              >
                              > On 6/29/06, Stephanie Sims <s_simstmag@...>
                              > wrote:
                              > >
                              > > Avery,
                              > >
                              > > Thank you... Our site sells Performance Sports
                              > Apparel and Footwear. So
                              > > one of our business goals is to increase reveune
                              > by 15% over last year. So
                              > > in doing so we have to convert our visitors more
                              > often, acquire more new
                              > > visitors and make our returning visitors buy more
                              > per purchase.
                              > >
                              > > One of the most challenging questions that I hear
                              > on a daily basis is why
                              > > is our site performing so badly today compared to
                              > yesterday or the day
                              > > before? From here (tell me if I am wasting time or
                              > looking into the right
                              > > data) I will then go and analyze the single access
                              > pages, exit pages and get
                              > > a total of all of our visits and compare them to
                              > these other metrics. For
                              > > example: he wanted to know why yesterday was so
                              > bad and after digging around
                              > > I found that most people were going to out
                              > footwear pages and then to our
                              > > flagship shoe. When the consumer got there they
                              > would leave or exit all
                              > > together due to the product not being available,
                              > no inventory.
                              > >
                              > > So I am learning but it is just taking me time.
                              > Please give me anymore
                              > > advice or tips that you may have.
                              > >
                              > > Thanks
                              > >
                              > > Stephanie
                              > >
                              > >
                              > > Avery Cohen <avery.cohen@...
                              > <avery.cohen%40gmail.com>> wrote:
                              > > Interesting question, Stephanie: "How do we
                              > increase conversion rate based
                              > > on the data we have?"
                              > >
                              > > "We have the business goals/objectives mapped out
                              > along with KPI's and a
                              > > excel report. ...I am new to the world of HBX."
                              > >
                              > > We might simplify and say: analysts basically have
                              > two tools to bring
                              > > insight and one tool for action.
                              > >
                              > > The two tools for insight could be called "Process
                              > Decompisition" and
                              > > "Segmentation". They will help you see "what is"
                              > and frame "what is
                              > > possible".
                              > >
                              > > With "Process Decomposition", you can break down
                              > "conversion" into a
                              > > series
                              > > of processes. I don't know what you want to
                              > convert, which would make it
                              > > easier for us to discuss specifics -- are you
                              > looking to sell, generate
                              > > leads, or create opt-ins? We might combine
                              > "Process Decomposition" and
                              > > "Segmentation" right from the start. Perhaps we
                              > could look at the
                              > > acquisition process -- how are people reaching
                              > your web site (what are
                              > > your
                              > > referral sources)? This is Segmentation by Source.
                              > Where are they landing?
                              > > This is mapping the Process Start Point and
                              > Segmenting the process by
                              > > Landing Page. How many visits from each source are
                              > single-page-view
                              > > visits?
                              > > How many of the multi-page visits are "converted"
                              > to one of your success
                              > > measures?
                              > >
                              > > Your tool for action is Testing Changes. After
                              > all, the only way to
                              > > actually
                              > > "increase conversion rate" is to change something.
                              > One might ask: "What
                              > > can
                              > > we change to get more multi-page visits?" You'll
                              > get more ideas if you are
                              > > clear about what is the first page people are
                              > seeing and how they got to
                              > > that page. One might ask: "Which of our
                              > Acquisition Sources are most
                              > > effective?" and go down a different path.
                              > >
                              > > You can use HBX features to see visit
                              > characteristics and behavior based
                              > > on
                              > > referral source.
                              > >
                              > > But I'm already starting to make assumptions about
                              > your "business
                              > > goals/objectives" and what "conversion" means to
                              > you. Perhaps this posting
                              > > will help you get started using web analytics (and
                              > your HBX subscription)
                              > > more effectivly. Feel free to respond with more
                              > specifics.
                              > >
                              > > Avery.
                              > > --
                              > > Avery J. Cohen
                              >
                              === message truncated ===
                            Your message has been successfully submitted and would be delivered to recipients shortly.