Deep interaction and interest metric

Understand behaviour in context, walk away with clear ideas how to improve your site

proposal wireframe

Highlights

Existing web analytics solutions can provide clues which help site owners to narrow down possibilities. But they fail short to offer concrete answers about shortcomings of their sites. They leave too much open for interpretation and meaning often get lost in translation. Based on my MA research on how information architecture affects the efficiency of communication on the web I propose the Deep interaction and engagement metric which has a potential to improve existing metrics standards, but more importantly displays data in relation to its context.

  • Structured data heat map

    Structured data vs. pages

    One of the biggest handicaps of existing analytics tool is measuring sites performance in relation to pages. Bounce rate, average visit duration, pages per visit and scroll depth can illustrate that the audience is not paying attention, but they don’t provide any feedback on why or how to improve it. Heat maps can be more informing however they rely on clicks. But if users are not engaged enough to click than no information about their interest is recorded. Adding a layer of data based on user interaction in relation to semantically structured data would allow site owners to clearly see the difference between what they think is important and that which their users do, helping them make decisions on how to improve the information architecture and design accordingly.

  • Keyword matching

    Interest Speculation Algorithm

    To understand users’ interest and behaviour it is necessary to take into consideration their experience prior to the visit. Visitors coming from organic search are driven by specific incentives, those coming from referrals are conditioned by the content within referral pages and direct traffic is a mystery. The Interest Speculation Algorithm builds an index of keywords for each section of the website and compares them to keywords from referral pages or organic search in order to speculate the section users are probably looking for. Being able to clearly see differences in interest and behaviour from different sources as well as similarities when analysing all traffic is a priceless feedback for optimising your site for all three scenarios.

    Keyword matching results
  • Interest KPIs

    Understanding engagement and conversion through interest

    Tracking and analysing users’ scrolling behaviour in relation to the position of content sections would produce precise data on what and how much is being read, skimmed through or ignored. This can be done for individual landing pages as well as all pages visited during a session. Based on this it would also be possible to generate additional data on average reading speed and actual reading time. Clearly visible connections between interest, engagement and goals (i.e. conversion rate) is an incredibly useful feedback for optimising content and its design for optimal results.

  • See data in context

    Analyse performance in relation to its context

    Context really is everything. While users’ demographics and time are well known factors, our understanding of behaviour in relation to the context has been even more challenged with the emergence of smart phones and computer tablets. Seeing data in direct relation to the set of conditions in which it was recorded provides a much deeper insight into the users’ experience and allows for more precise interpretations. I.e. if countries where English is not a native language represent a significant amount of total traffic but results show drastically lower average reading speed and level of reading interest you might consider a bilingual approach or making your copy less demanding. Or if your website is mostly visited during the day, but the interest and engagement levels are low, you might consider making the content shorter and easier to digest and move in depth information to separate pages.

Further reading

  • Research report draft

    This proposal is based on the theory from MA research paper titled Information Architecture gospel according to ELIZA which interrogates the question "how does the structure of information affect the efficiency of communication?".

  • Documentation and specifications

    Within this document you will find a break down of the project scope, descriptions and rationale behind statistical parameters for the metric as well as an outline of possible methods and logic for tracking them.

About the author

The concepts of this proposal were conceived and designed by Natan Nikolič. He designs digital products and services that are striving for efficient performance of their purpose and becoming an irreplaceable part of their users’ lives. The latter being an overly ambitious mission he had yet to accomplish with his work. Most of the year he can be found in Slovenia London, where he is currently finishing a MA Communication Design course at Central Saint Martins College of Art and Design and helping a handful of cheeky clients achieve their business goals.

Please do not hesitate to get in touch