Big Data Meets Ad Tech

By
  • Facebook
  • Google Plus
  • Twitter
  • LinkedIn

Data-Driven Thinking"Data-Driven Thinking" is a column written by members of the media community and containing fresh ideas on the digital revolution in media.

Today's column is written by Tim Chang, Managing Director, Mayfield Fund, a venture investment firm, and Tim Hanlon, CEO, Vertere Group, a consulting group.

The brand ad market is massive (multi-multi $B), with 67% of worldwide ad spend in branding.

However only 25% of digital ad spend is in branding and the online brand ad market is still relatively primitive for these reasons:

  • No standards or consistent measures of “success” other than outdated or inadequate metrics like CPM and CTR (clickthrough rate)
  • Limited real-time intelligence
  • Unsuitable display ad formats (still mostly banners)
  • Lack of creativity in formats

Today, marketers and their agencies are overwhelmed by media choices, and departments are not optimized for a data-driven future of real-time bidding in the digital world. Online advertising is driving the need for a new breed of marketer who can handle an exponential blizzard of data sets (external and proprietary) and derive actionable marketing insights/decisions. As a result, CMOs and CIOs are cross-pollinating, driven by CRM/loyalty programs tied to increasingly granular media measures.

However, some in the ad industry lament that the awkward growth of the online display market is pushing things too far into the "math men" camp, and a "correction" of sorts is overdue, especially as some of the data measurement tenets start to come to other media forms – this drives a real need for an "informed" brand metric (e.g. brand metrics "2.0"). This new era of "big data" is not ready to fully replace traditional metrics from "tiny data" such as Nielsen-like TV sampling projections (e.g. national ratings based on sampling only 40k homes), but is primed to supplement and enhance traditional measures. A good analogy is what has happened in investment bank ratings: S&P, Fitch, and Moody's legacy ratings are now quickly supplemented with multiple sources of granular data, ending sole reliance on proprietary incumbent systems, and adding more depth and sophistication of understanding.

The current world of online advertising is effectively dominated by the single metric of click-through rate as the key measurement of ad effectiveness. However, clicks are not an effective measure of online advertising effectiveness, as even effective online branding and display ads do not naturally lead to click-through like eCommerce-oriented direct response ads do. The predominant CPM-based approach has been driven by legacy TV formats centered upon static “impressions” and do not capture the interactivity and complexity of today’s digital formats. The ad tech industry is evolving new approaches to media expenditures - including what industry leaders are calling a "Brand Equity Index," which aims to equate and more accurately depict the actual contribution and effectiveness of various forms of "media" expenditures against brands and consumers.

To support these new approaches, today’s brand marketers, publisher ad sales teams, and agencies have very few tools to understand their online brand equity – the ad tech market has not yet had simple Google Analytics or Omniture-style SaaS solutions, or even Google Search-style tools to dynamically lookup or index online ad performance and metrics. We believe this void presents a dramatic opportunity for a hybrid agency/SaaS company that combines the simplicity of Google Image Search, specialized tastes of Pinterest users and the back-end power of a Google Analytics on steroids. With the emergence of companies that deliver such solutions, ad tech can benefit from real brand intelligence and brand analytics, at the same level of granularity and depth that all other online businesses, from gaming to eCommerce to social media now enjoy.

Follow Tim Chang (@timchange), Tim Hanlon (@timhanlon), Mayfield Fund (@mayfieldfund), Vertere Group (@verteregroup) and AdExchanger (@adexchanger) on Twitter.

  • Facebook
  • Google Plus
  • Twitter
  • LinkedIn

Email This Post Email This Post

By on at

7 Responses to “Big Data Meets Ad Tech”


  1. Tim, thanks for a thoughtful article. The problem of measurement is may not be as big an issue as we think right now. Companies whose products and channels make it easier for them to trace the effects of online media to an ultimate business action are developing custom metrics to understand the right mix of targeting, creative and destination techniques. In my experience, these metrics include too much proprietary business intelligence to make the resulting metrics public. This makes it harder for us to discuss the medium publicly, but it doesn't seem to hurt my discussions with clients as we make decisions about how to use online media.

  2. Bart Myers says:

    YES! Not only are the tools lacking, basic education is as well. Explaining the prevalence of this trend to business leaders in the ad/marketing space is extraordinarily difficult and missing some basic industry wide tools. The IAB's resources on data driven buying is a start, but insufficient. There is an opportunity here, I can smell it.

  3. Adi Orzel says:

    This article is right on the money! The solutions needed would transform the brand metrics used in traditional media channels to the online world. No more click-throughs, but actual measurement of brand metrics elevation throughout a campaign flight (a survey at the end is not enough). Add to that real analysis and optimization capabilities and you've got your 2.0 brand metrics product.

  4. Joe Pych says:

    TV does not have standard measures of success nor real-time intelligence, yet it takes in a large and growing chunk of media spend. Shifting dollars to digital is more than a analytics tools and data problem. Your second two points on ad formats and creativity were not explored in this article, but are perhaps more significant than the first two. The iab just rolled out some new standard "big" ad formats that look promising.

  5. I particularly like this notion of Google Image search as an input. When measuring television in particular, the quality of the flight data continues to be opaque in terms of actual exposure. Being able to accurately input into models the specific creative and flight-time would improve the effectiveness of the recommendations considerably.

    Is there anyone doing that yet as a service? We could certainly use that data.

  6. Nick Choat says:

    I work for a large media firm that has now (recently) fully "big data enabled" our digital space. Our strategy is to look for the "unknown unknowns" that will translate into innovative ways to measure performance. In order to truly accomplish this mission you unfortunately need to employ data scientists (math men.) Once you've identified these innovative metrics, however, you can then operationalize the metrics for more simple, but effective, daily consumption. This is very early stage work for us as well as the industry, but the promise is large.

  7. Jimmy Bogroff says:

    Paralysis by analysis.

    The problem is we're trying to take a quantitative approach to a qualitative issue.

    How do you subjectively value each user, the location they viewed the ad, the format of the ad, and then try to align a metric to their response?

    Its the equivalent of trying to determine a coefficient for the emotional response each unique user has when exposed to the brand.

    Stickiness and context play so much into brand perception. While the brand can certainly control their message, they are finding it more difficult to control the context online.

    It is important that the experience be truly interactive (and this doesn't necessarily just mean "online). Its not about how many impressions, or clicks, or fans, or even online sales (unless you only sell online); it is about creating uniquely connected experiences.

    Are your online stores, webpages, storefronts, packaging, kiosks, and etc. all promoting the same experience?

    Or is it a cluster of disjointed messaging, sloppily executed campaigns (each with a different execution strategy by channel),and conflicting promotions that leave your audience confused about your brand?

    Because of the platform fragmentation, it doesn't surprise me that only 25% of digital dollars are classified as brand dollars.

Leave a Reply