1st Impression Score

16 comments 14,344 views

On the average Web page or blog post, users may read about 20 percent. The first 10 seconds are critical. This score supports your efforts in keeping  readers longer.

The 1st Impression Score (you are here) is a story-analysis algorithm that assigns a numerical weighting to each blog post  to obtain the score ranging from 0 (worst) through 100 (best).

The ComMetrics 1st Impression Score  is part of what makes up the CyTRAP BlogRankThe latter is calculated by using FIVE indices: Headline ScoreEngagement Score Text Complexity ScoreRipple Score (Google+, Twitter AND Facebook) and the 1st Impressions Score (you are here).  

1 – screen-and-glean score

Research reports that the time users spend on a Web page follows a Weibull distribution.

Hence, people scan things and decide often within five seconds if they intend to read more or switch over to another page or blog post. Nielsen has proposed that  educated readers may be able to read about 250 words in 1 minute.   If you structure your post well so that people can screen-and-glean a lot of information within about 30 seconds, chances are greater that they will read on than otherwise.

Based on  the literature much might happen within about 5 seconds, 10, 20 and 30.  Of course, if your audience has read about 125 words and finds things still interesting, the chances are they will read on.  Nielsen suggests that in-depth content provides more value in less time for business people than numerous superficial postings.

Using headers, bold text as well as paragraphs helps increase the chances that people continue. Hence, considering that your blog attracts literate users, visitors are able to read about the following number of words:

– Sum A = 1st 5 seconds on your web site or blog = reading about 20 words,
– Sum B = 10 seconds spent on Web site or blog = reading about 40 words in total,
– Sum C = 20 seconds about 80 words, and
– Sum D = 30 seconds about 125 words.

On the average Web page or blog post, users may read about 20 percent. Therefore, the  first 10 seconds are critical.(see Jakob Nielsen: How little do users read?)

Hence, using bulleted lists and highlighted keywords, chunking the material and using descriptive headings, subheadings as well as hyperlinks helps.

Sum A – 20 words = [Headline, bold or italic text (each word)/numbers, paragraphs, URLs]
Sum B – 21 – 40 words =   [Headline, bold or italic text (each word)/#, paragraphs, URLs]
Sum C – 41 – 80 words =   [Headline, bold or italic text (each word)/#, paragraphs, URLs]
Sum D – 81 – 125 words = [Headline, bold or italic text (each word)/#, paragraphs, URLs]

PS: A headline counts once, if text is bold and italic it counts once, each paragraph counts and so does each URL.

2 – identifiability score

Research indicates that victim identification tends to generate more aid than any other type of pitch for donations. This might explain why charities employ a poster child, since this helps raise money for a general cause. This builds on the notion that certain stimuli evoke more affect than others.

Of course, showing how a donation can make a difference in a child’s life does help. As does if we show an image with a graphic or we embed a short video in a blog post that illustrates  how a customer’s problem can be solved with one of your products.

– 10 points = 2 to 3  images  (e.g., picture, graphic, or table) or else 2 videos, 1 or 2 images and 1 video
–   7.5 points =  2 images, 1 image and 1 video clip
–   5.0 points = 1 image (e.g., picture, graphic, or table) and/or 1 video clip
–     0 points = other

Selectively use images to reinforce your message, as images often take less time to understand than words.

My experience has been that if there are too many videos and pictures, something gets lost in the story. One reason being that the online reader (i.e. she is scanning the text on a mobile screen and/or laptop) is getting somewhat overwhelmed with too many images and/or videos.  See for instance, ComMetrics weekly review – social media going’s on.

This post has 4 short videos embedded and 2 images.  Based on this and other examples we decided to choose as top score the 2 images and 1 video or just 2 videos. To illustrate this press release has one graphic as an image: Sicherheit und Bedienungsfreundlichkeit sind Top-Kriterien für soziale Netzwerke.  It would probably benefit a great deal and improve the quality of the story if it also had a 1:45 min video embedded with this press release.

How can we score your story

So how will we score a title like this one: Sicherheit und Bedienungsfreundlichkeit sind Top-Kriterien für soziale Netzwerke

Sum[(A x .30) +  ( B x .275) + (C x .225)  (D x .20)]/4 =
1 + 1.375 + 0 + 0 = 2.375 

In the above story, there is also one graphic, which results in 5 points

We weigh the glean-and-screen score with .60 and the identifiability score with .40

Addordingl, the above results in: [(2.375  x .60) + (5.0 x .4) ] = 3.425

You can download the text analysis here

Another example is the beginning of this blog post which ends up to score as follows:

Sum[(A x .30) +  ( B x .275) + (C x .225)  (D x .20)]/4 =
           (6 x .30) + (9 x .275) + (9 x .225) (13 x .20) =

 The above results in:  2      +   2.475   + 2.025 + 2.6 =  9.1

Since there is no graphic in the above story, it gets a 0 for the Identifiability Score.

You can download the text analysis here

Just to repeat, the numbers we get above are then added with the following weights: Sum[(screen-and-glean score x .60) +  ( identifiability score  x .40)] = CyTRAP BlogRank 1st Impression  Score which we can then use to rank your blog compared to others.

The z-scores for the above indicators are added up to get an overall z-score.  This information is calculated into an overall score. Click here to find out how we process the raw data.

The actual ComMetrics Mnemonic  Score number is used in the ComMetrics algorithm to help determine the ComMetrics Footprint of the blog, website or other social media effort being benchmarked.

At this point, the overall scores are compared and rescaled using 100 as the top score.

Browser displays and mobile screen sizes

The size of the screen does also affect what you can do. The screen width and height is getting ever smaller due to the ever greater use of mobile devices for surfing (global stats – screen resolutions).

Browser displays and mobile screen sizes can be measured by:

1) the size, and
2) the resolution called pixel density.

The size is generally measured diagonally in inches, and the resolution is on the number of pixels displayed on the screen. (from Helena – screen sizes)

Many webpages use up about 10cm (i.e. more than half of a browser window’s vertical size) for the header and spacing (check with your PC or mobile to see what I mean)  before a blog post even starts. Hence, the 20% of people that scroll down the webpage before it finishes loading will, of course, not see the header and might miss the headline as well.

Recent research reports that the “very top of the top of the page actually has about a 20% lower view rate than slightly farther down.” see: Schwartz, Josh (August 12, 2013). Scroll behavior across the web. (Blog Post – Chartbeat). Retrieved, November 4, 2013 from http://blog.chartbeat.com/2013/08/12/scroll-behavior-across-the-web/#comment-1107895121


Chao Liu, Ryen W. White, and Susan Dumais. (2010). Understanding web browsing behaviors through Weibull analysis of dwell time. In Proceeding of the 33rd international ACM SIGIR conference on Research and development in information retrieval (SIGIR ’10), pp. 379-386. New York: ACM. DOI: 10.1145/1835449.1835513. Retrieved August 23, 2011, from http://doi.acm.org/10.1145/1835449.1835513

Small, Deborah, A., Loewenstein, George, Slovic, Paul (2007). Sympathy and callousness: The impact of deliberative thought on donations to identifiable and statistical victims. Organizational Behavior and Human Decision Processes, 102, pp. 143-153. Retrieved December 23, 2011, from http://www.juststoryit.com/articles?mode=PostView&bmi=778807

Weinreich, Harald, Obendorf, hartmut, Herder, Elco, and Mayer, Matthias. (February, 2008). Not quite the average: An empirical study of web use. ACM Transactions on the Web (TWEB), 2(1). Doi: 10.1145/1326561.1326566. Retrieved December 23, 2011, from http://dl.acm.org/citation.cfm?doid=1835449.1835513

This also benefitted from: Karen Dietz’s guest post that addresses the conceptual challenges of story telling: Terms of the art: What every discipline needs

  • Stan Albers

    This is interesting, Urs. Of course, objectivation of such complex entities as stories, in a weighing sense on top (good vs bad stories) is cumbersome. Do we really need to objectify it? Or do we just need story telling skills and measure whatever can be measured without so much effort?

    • http://info.drkpi.ch/articles/social-media-marketing-fur-den-kmu Urs E. Gattiker

      Thanks @1e2a9d6ba5f25a44b98efc121b057c77:disqus Very good point, do we need it.
      Well yes and no.  For instance, if I can operationalize the story component and measure it.  What would you say if the score you get is positively correlated with the amount of engagement – blog comments, tweets etc. your blog post gets?

      I have some data where we tried to test it with our blogs and some of our clients and story has an impact on what kind of engagement you get.  Of course… so if I can then find an algorithm that helps simplyfying this difficult task of analyzing these difficult data, voila.  Not bad.

      I am still revising this post but it is coming along nicely.

      • Stan Albers

        “What would you say if …” Of course that would be a valuable tool. I just ask myself, who is going to do this operationalizing? Many if not most other metrics are generated automatically. Here you must do a weighing of data all by yourself. But please go on! It is interesting!

        • http://info.drkpi.ch/articles/social-media-marketing-fur-den-kmu Urs E. Gattiker

          @1e2a9d6ba5f25a44b98efc121b057c77:disqus Thanks for your reply.
          Well I tried to operationalize part of it above and I have measured it with our various blog posts using the http://My.ComMetrics.com tool.

          There are still some fine-tuning efforts we have to undertake but it seems to be working…. more engagement, more conversation, increased pageviews… , more successful reaching out to targeted audience.

          Not bad for a start?  But based on some comments I hope to get we can still improve this.

          Thanks for sharing and Happy New Year to you and your loved ones from us at @ComMetrics:twitter 

  • http://ashshepherd.com Ash Shepherd

    Very comprehensive breakdown of your methodology and scoring system.  What I appreciate the most is that while you go into great detail and give solid examples of the work you do, others can also gain a top level understanding of what elements have what sort of impact on the “footprint”.

    Very exciting work you are doing here.

    • http://info.drkpi.ch/articles/social-media-marketing-fur-den-kmu Urs E. Gattiker

      Dear @ash_shepherd:disqus 

      Thanks so much for stopping by and taking the time to have a look at this material.  Of course, I appreciate your feedback very much and I hope that we continue to provide you with useful material.

      What I am really curious about is also to receive suggestions on how we can improve this with a comment or using e-mail such as info at CyTRAP.eu  where I will surely get it :-)

      I appreciate your feedback and wish you a Happy New Year!  Ciao @WebUrs:disqus 

  • http://twitter.com/kdietz Karen Dietz

    Fascinating work here Urs. So sorry it took me so long to review this.  I really like your algorithms and definitely think they can be helpful. Of course with the story element, what you are doing is simply counting if/how many images are present in support of the text, correct? From what I have been reading, yes — 3 visual images linked together in a way that makes sense (tells a visual story) is optimal.  Your algorithms focusing on 2-3 photos, or the mix of video/photos works great for me.

    Now the real tough piece to measure would be on the quality of the ‘story’ that’s on the website. As you know, not all photos, photo essays, or videos turn into conversions because they end up being poorly told stories. Same with the text.  But what I like about what you are doing is that by giving a score based on your algorithms, you are indicating that with a low score, your stories obviously are not working like they should. So back to the drawing board!

    I think you are really onto something here.  I would love for you to share more examples as you collect them. Many thanks for this piece of work. I’m excited about it.

    • http://howto.drkpi.ch/bang/storytelling/ Urs E. Gattiker

      @twitter-11253892:disqus  Thanks so much for your feedback.  I really appreciate it.
      I have done a bit of thinking to figure out what I can do here.

      1st. You are right I am still trying to get my handle on this and our team has thrown out a few definitions and attempts of clarifying things from my end. 

      The index does more than just count pictures and images although that is part of it.  It also addresses the screen and glean score that I borrowed some ideas from Jakob Nielsen on this.  It looks at bold text, italic text, paragraphs considering the Weibull distribution to figure it out as explained above.
      Basically, if you cannot get my attention in the first section, you may possibly not go on and read a bit….  (see my yellow highlighted summary agove)
      Therefore, most people leave within the first 5 seconds.  

      Beyond storytelling
      We have another index that addresses the language – difficulty, etc. being used to get your story across here:
      ====>  ComMetrics Mnemonic Score 

      Thanks Karen for answering.  Urs @ComMetrics:twitter 

  • Pingback: What's Your Website Story Score? ComMetrics TriageHazard Score | Just Story It | Scoop.it

  • Pingback: What's Your Website Story Score? ComMetrics TriageHazard Score | Local Search Marketing Ideas | Scoop.it

  • Pingback: What's Your Website Story Score? ComMetrics TriageHazard Score | Tessa Winship.com Children's Picture Books | Scoop.it

  • Pingback: Why do infographics fail? » Data visualization and ROI return on investement - Part 1, recognize a great infographic, Seven sure-fire suggestions, infographics, 91 percent, six basic issues you need to address, , Keep it simple Stupid » ComMetrics | me

  • Pingback: Die 10 besten Karriereblogs der Schweiz » » Dr KPI: ROI, Efficiency, KMU, Kennzahlen, Strategien

  • Pingback: Top 50 UK fashion blogs » February March 2014, Who are the top fashion bloggers, criteria do fashion magazines use, How does a fashion brand decide, Keywords, Top 50 US fashion blogs, Top 50 UK fashion blogs  you are here, By the way » ComMetrics | me

  • Pingback: Top 50 US fashion blogs » February March 2014, Who are the top fashion bloggers, criteria do fashion magazines use, How does a fashion brand decide, Keywords, Top 50 US fashion blogs you are here, Top 50 UK fashion blogs, Top 50 Modeblogs in Österreich

  • Pingback: Top 50: Modeblogs in Österreich » » Dr KPI: ROI, Efficiency, KMU, Kennzahlen, Strategien