3 May 2012 in ExchangeWire EMEA 7 Comments

Understanding Engagement: The Final Frontier

Ollie Bath is head of Client solutions at IgnitionOne UK. Here he discusses the importance of understanding the user journey and properly measuring engagement in the path to conversion.

Let’s be honest, measuring online media remains a tricky subject. But it’s so important to understand.
Here’s why; money talks. Nearly £5 billion was spent on online media in 2011 – accounting for more than a quarter of all UK advertising. However, marketers must collate data from across all digital outlets to get the best return on their online spend. The resulting insight can then be applied to each channel, increasing accuracy across the online customer journey.

Last click metrics do not mirror the full conversion journey story accurately. Online purchase decisions are based on numerous actions, rather than a single one. Having visibility into user paths to conversion and distributing credit to contributing exposures is relatively straightforward, with the right tools in place. But understanding what drives online engagement is the final frontier for many marketers. So, as marketers, we need to equip ourselves with the right tools.

And the market is already swamped with a seemingly endless deluge of these tools – from sentiment analysis to algorithmic powered demand-side platforms. To truly understand the overall impact of online marketing, marketers should look to tie the full user journey together by integrating platforms, not only individual point solutions.

It’s now possible to take online data, interpret it in real-time, and alter the way ads and content are presented to create a better, more personalised user experience, ultimately increasing the likelihood of a conversion. Custom attribution models enable marketers to distribute fractions of action credit to exposures in a user path. To date however, this approach has only assessed latency, exposure sequence and associated revenue. This completely disregards the value of the amount of engagement that the website visit created, leading to the conversion.

What’s needed is an additional layer of information that collates data across all channels – SEO, paid search, display, email, affiliates and social media – to understand how users engage as a result of seeing media relative to their stage in the conversion funnel.

Touchpoints that are indirectly related to the end conversion also have an impact. This is precisely why it’s important to assign a value to each media exposure, in order to determine which ones had the most positive effect.
Understanding that if users are highly engaged but not yet converted, helps marketers interpret the true value of the purchase funnel and the media these users have been exposed to, rather than the value of the media relative to the end conversion.

Numerous brands are already realising the benefits of measuring engagement. At IgnitionOne, we’ve been working with the printing and personal publishing business, PhotoBox, for several months, measuring Facebook’s impact on the brand’s marketing performance. This involves monitoring three sources of traffic:

- Fan Pages (people coming directly from the Fan Page to the website)
- Paid Clicks (people coming directly from paid ads on Facebook to the website)
- Viral Clicks (people coming directly from any other Facebook page other than the Fan Page or a paid advertisement to the website)

Through analysis, it’s evident that Facebook paid ads deliver more new visitors than viral or Fan Pages. However, viral visitors stay on the site longer and are almost twice as engaged on the site. The key finding was, taking into account the cost, conversion rate and engagement scoring, Facebook and paid search ads offered comparable value to aiding conversion.

Research by Microsoft and comScore further proves that campaigns with high engagement boost advertisers’ site traffic by 69% and improve brand engagement – increasing page views and time spent on the brand’s site.
Overall, engagement insight gives a clear indication of where to direct online marketing spend, with a greater degree of clarity and accuracy.



Latest jobs on Exchangewire


  • HarleyNorrgren

    Hi there,

    I like the atricle and agree that measuring cross platform is the key for attribution.
    I’d like to know how you’re defining engagement as it’s not really explained above, and if you have a link to that Microsoft paper?

    For me, site engagement works from a brand campaign point of view, but is just another proxy for conversion from a DR perspective as tracking shares, fanpage sign ups and posts on a user level is currently unavailable on Facebook.

    I may have missed the point of your article but how do you feel brand engagement adds to a sales focussed attribution model, which already tracks interactions?

  • http://twitter.com/Bathallot Ollie Bath

    Hi Harley,

     

    Thanks for your input and excuse the delayed response…bank holiday weekend.   

     

    We see engagement as synonymous with ‘propensity
    to convert’. The algorithm is self learning, comparing the behavior of the
    visitor on the site to all of the visitors before them and gauging the
    likelihood that they will become a converter (propensity scoring). There are
    120 variables that are looked across recency, frequency and monetary value. The
    key thing is we are predicting future behavior by considering that likelihood
    as compared to other converters and non converters. This method is in contrast
    to business rules that are based on assumptions (often unjustified or
    outdated). 

     

    I see brand behavior is essentially activity
    at the top of the funnel. The problem is that most optmisation decisions are based
    only on converting data. The general perspective is to find converters and give
    them the best chance at success with budget, etc and find the point of
    diminishing return.  This represents 3% (converters) of the data available
    and our perspective is that there is tremendous untapped value in the 97%
    (non-converters) to better understand the value of their behavior (engagement)
    to lead to clear ways to prioritize and act to increase conversions. The
    technique of Engagement Optimization is about focusing more holistically across
    the value of actions (conversions) and activity (engagement) leading to
    conversion.

     

    The idea here is to not only look at
    exposure path and latency, but to assess each media ‘experience’ . How did the user
    engage on the advertiser’s real-estate off the back of specific media he/she encountered.
    Did someone coming off a Facebook add behavior differently onsite to someone
    who came off a PPC brand term.  It’s
    adding an engagement metric that is undoubtedly bringing greater insights into
    the value the media has contributed in pushing users down the conversion funnel.
      

     

    Ollie 

     

  • http://www.infectiousmedia.com/ Harley Norrgren

    Hi Ollie,
    Thanks for the reply, I see your model looks at engagement as a touchpoint rather than a user behaviour metric, this would make it much easier to optimise!
    What are your opinions on non-view trackable channels, such as PPC and Facebook and whether exposures via these channels could be considered media experiences, similar to display views? If so, should we be pushing these channels to release view level data back to advertisers in order to eliminate attribution bias?

    Harley

  • http://twitter.com/Bathallot Ollie Bath

    Hi Harley,

     

    No problem.

     

    We absolutely do measure engagement/propensity
    to buy  as a metric, which is how we can (a)
    optimise it (b) align it to managed media. If an advertiser is managing their  media via our media management suite we are
    able to align the ‘engagement metric’ to referring media. Doing this enables an
    advertiser to assess the value of the media (PPC /FB click) based on engagement
    on their site – Adding the engagement metric to the equation is what’s helping
    the advertisers better solve the attribution.

    In reference to your view/click point –  Views are slightly more challenging as we have
    to aggregate. Impressions that a user has been exposed to between sessions as an
    arrival can’t be directly associated with one singe impression.

    Hope this answers your question.

     

    Ollie

  • Matty

    Hey Ollie,
    This bleeding edge method of understanding the attribution funnel is exciting. However, do you feel there are enough clients taking such an approach on board? In my experience the use of such approaches is the domain of the econometric analyst and so gets ignored in the day 2 day marketing meetings where the bottom line CPA’s rule the roost. Creating the debate, as you are doing here, is a great way to move this into the company boardroom and away from the excel spreadsheets. Another way might be to share examples of some top brands that are successfully adopting this approach and seeing a bottom line improvement in conversion efficiency.Great stuff.

  • http://www.abercrombieandfitchonsale.co.uk/ Abercrombie and Fitch

    It was very useful for me. Keep sharing such ideas in the future as well. This was actually what I was looking for, and I am glad to came here! Thanks for sharing the such information with us.

  • http://twitter.com/Bathallot Ollie Bath

    Hi Matty,

     

    To your question: Unfortunately not enough,
    this should be mandatory given what we as digital marketers know and have access
    to.  Saying this we are seeing an
    increase as the tools and ‘know how’ advances. The brands that we’re working
    with that are taking the necessary steps to move off legacy last click are undoubtedly
    seeing the effect of their hard work pay off.

     

    Ollie