Marketing Science Starts in Physics Class

A slightly altered version of the content in this article has been published by martechseries.com.

Some of the most important lessons for programmatic advertising are taught in high school physics classes. It’s quite clear to me that we’ll all do well to drown out some of the noise, buzz and hype in our space as the world of marketing becomes more data-driven. Let's talk about the electron.

In school most of us learned that an electron is in fixed orbit around the nucleus of an atom. The assumption made is such that there is a definitive place where that electron exists.

Some in the quantum physics community have contended that an electron is not in some definitive place. “According to Bohr, the electrons revolve around the nucleus in certain well defined circular orbits. But the idea of uncertainty in position and velocity is said to overrule Bohr's idea of fixed circular orbits.” Ambiguity surrounding the relevant positions and velocities of electrons have caused some quantum physicists to point out the corruption of the resulting data set. i.e., by trying to fit the electron into a category of existence you misunderstand the nature of its reality. Since the reality of the situation is misunderstood the numbers used in calculation are faulty. This debate is ongoing.

Semasio shares quantum physicists’ desire for revolution and exactness.

Advertising technology companies build audiences for advertisers. Some of these audiences are behavioral, and are categorized broadly as “home buyer” or “new home offer.” If an online user visits a certain number of URLs (frequency), over a certain period of time (recency), that user is bucketed into a “home intender” audience. URLs are labeled broadly, thus polluting the integrity of the profile. Demographics, age, income, etc. are often laced into the segmenting of these audiences via Boolean Logic; the use of “and” or “or.”

If a consumer’s demographic or income level puts them in both “finance” and “home intender” buckets then they are served advertisements. The information utilized is often outdated, the data’s origins are often opaque, and the settings (frequency/recency) are often unknown. These audiences fall whim to taxonomy and make assumptions about users’ affinities towards goods and services based on vague behaviors associated with blanket URLs and demo/socio-graphics. Then, this polluted data pool is bucketed and sold as a segment. Thus, advertisers purchase segments that are built on outdated &/ hidden traits. Therefore, this approach is oversimplified and data destructive; similar to the argument brought to the table by quantum physicists. What’s more, since the original data pool is destroyed, the true interests of the end users are not kept in mind -- regardless of the creatives used.

Semasio is built on Semantic Behavioral technology. We perform a semantic analysis of the content on pages and associate that content with the users who have consumed it. This forms a user’s profile, visually represented as a word cloud. These segments adapt in real time to account for eternal change. In terms of the quantum physics revolution: The electron is better seen as an entity that is somewhere in the atom’s cloud – not an entity in a fixed and static position.

In our attempt to democratize data we empower more than 300 blue chip advertisers to define the tradeoff between semantic relativity and reach in their segments. Advertisers define all of the settings that define their segments, which are sourced from more than 95% of the internet population across over 30 languages. Thus, advertisers who work with Semasio can target current audiences of users who:

• Share semantic relativity with those who’ve performed a certain action (via first party data) or,
• Have consumed significant keywords and phrases.

As with many things in marketing, the classics still have resonance. And, a sound approach that recalls science can pay dividends if you focus on the basics and common sense.

Prefer a linguists view on the matter? View this article from my colleague and friend Jeffrey Adler.