Quantifying the engagement impact of Display Advertising

speedo_4__76509Display Advertising is a normative experiential aspect of web sites nowadays and the majority of site visitors do not take offense to the presence of advertising real-estate. As a mater-of-fact: after reviewing user feedback (UGC), Call-center tickets, and direct customer conversations concerning advertising. Most users stated that they understood and accepted the business necessity as long as it didn’t interfere with the User Experience (UX).

There is a common misconception among Publishers that Display Advertising is an endless well of “free money.” With this misconception comes a naivety that the sites content is so compelling as to outweigh any negative UX of intrusive Advertising. That being said, there is a ray of truth in that last statement; some sites (notably News) can leverage the perceived value into supporting more Adverting. At the end of the day – there is no defined number of ads or ad types. You will want to let your engagement and attribution metrics communicate what your customers and traffic sources feel is appropriate. The following are three fundamentals to aid in shaping the ideal, Ad Supported User Experience (UX.)

1. ACTIVELY MONITOR AD/RESOURCE FOOTPRINT

Advertising (especially Rich) can add a tremendous resource tax to your website in the form of page load time and page size. Furthermore, I have seen flash-based video ads (NDN, Taboola) peg CPU utlitization, bringing even the mightiest workstation to its knees. The scatter chart below showcases the change in page load time and size by just removing the advertising.

news_site_loadtimes
Competitive site load analysis – highlighting the impact of Ads on user experience

Replicating bad Ad/Ad Network behavior is difficult, but not impossible to quantify with behavioral, geographic and programmatic targeting altering the experience at the user level. However, our goal should not be in chasing individual ad units, but in identifying the Ad Networks which don’t do an adequate job in filtering them out. Here are some ways to continuously monitor your advertising to minimize impact:

  • Understand how you compete (page resource-wise) against your competition. If you’re too far out of line, you are giving your customers a chance to churn to a competitor.
  • Establish guidelines with your Ad Network partners to limit Ad types with high footprints (e.g. video and flash-based ads)
  • Enable RUM (Real User Monitoring) and/or Synthetic monitoring to identify changes in page load time.
    • Make sure that you track this over time and add notations when you add or modify Ad networks in your waterfall.
    • Create a hidden (no crawl/no robots), Ad-free version of the page you’re measuring to baseline against. Remember, changes in content or site UI/UX can increase page load time as well. We want to isolate the Ad impact against variations in the normative content
    • Make sure to periodically track and benchmark against your top competitors.

Remember: Your goal is not to have the fastest website ever – but one that is as fast or faster than your competitors.

  • If your Ad waterfall is too complex to wrangle (like the one shown below), consider work with services like Ghostery Trackermap and/or MCM to help identify Ad networks that are too resource intensive. They may appear expensive up-front but it will save you money in the long run.

  • Develop a cohort-based, regression model to better understand the relationship between page/site load times and engagement/attribution
    • Why? Because this will let you know if you’ve got too much advertising – or not enough.

2. QUALITY, QUALITY, QUALITY

Shady Ad Networks as well as aggressive experiences (such as pop-unders, auto-start video, auto-audio and takeovers) may offer initially higher eRPM amounts, but will come with a cost of decreased key engagement metrics (time-on-page, pages/session). This is a great way to kill your site’s traffic in a very short period.

Always remember your customer and your brand promise to them. If they’re visiting your site at work: the last thing your customer wants is the rest of the cube farm hearing an auto-play audio ad talking about hemorrhoid treatments. Now, what do you think is the likelihood that customer will revisit your site again?

Most business owners don’t understand that Facebook and Google pay very close attention to the bounce rates on your content until its too late. A persistently poor experience will negatively impact Social network attribution. And with Social referrals contributing up to 50% of traffic for a good site, this isn’t trivial. Furthermore, your Marketing department may be paying for traffic – which you’re throwing away money on (via bounce rate penalties or abandonment).

If you don’t like the experience you’re seeing, you can be guaranteed the users wont either. Its always easier to walk away from a bad Ad Network than trying to win back lost customers.

 

Moral of the story, Closely monitor customer feedback, attribution and engagement metrics. 

3. BE VERY CAREFUL ABOUT IMPLEMENTING ADS ON PAID/SUBSCRIPTION SITES

Most users view this as form of corporate greed and assume that, by paying for the service, they should enjoy an Ad-free experience. Adding advertising WILL cause an increase in subscription churn.  Measuring the impact of Advertising on adjusted Lifetime-Value (LTV) is crucial.

  • Develop a model to closely follow Key-Performance-Indicators (KPI) Deltas on a Daily basis
    • Churn (total cancellation or cancel of auto-renewal)
    • Changes/Trend in Pages/Visit by cohort (to remove variable influences of attribution sources and devices)
    • Changes/Trend in Visitation by Unique
    • New Sub acquisitions and conversion rates, post-trial-period, if applicable
    • New Display Advertising (DA) revenues
    • Total site revenues (subscription and advertising)
    • Lastly, don’t forget to include the additional OpEx, CapEx costs of running Advertising on your property (Ad Server costs, trafficking expenses, Hiring a sales staff, producing sales materials, resource opportunity costs)

Have fun, listen to your customers and remember to keep your brand promise

David

Article by David

Your host and an experienced analyst dedicated to evangelizing the quantifiable, data-driven measurement of business with an emphasis on the customer relationship.

Leave a Reply

Your email address will not be published. Required fields are marked *