Existential Advertising: Does it exist

Most advertising is temporary. It exists for a moment, then disappears. Does it exist?

Have you ever seen an ad -- banner, commercial, video, or other content -- and tried to find it later? Have you ever searched for a branded product, only to turn up reviews and retailers but not the ad or messaging? In other words, advertising is not indexed or fixed in a place, it lacks permanence and persistence, therefore its existence is that moment in time where it targets you.

In practice, existential advertising is typified by the one-and-done annual campaigns that most companies still plan, budget, and execute. In the old days these campaigns were dominated by TV and print with a big creative USP (unique selling proposition), while today they are fractured into data rich 360 integrated omni-channel campaigns using paid, earned, and owned media and multiple, targeted A/B test optimized messages.

The growth of content marketing is marginally changing this as the document is fixed in a place. Coupled with ontology and semantic mark-up technology, this content is permanent and discoverable in search. This permanence will be absolutely critical as artificial intelligence, natural language processing, and voice technology become common.

These technologies require a lot of data for training, testing and validation sets. What does this mean for your brand and the advertising messaging? How will the AI know you? How will Amazon Alexa, Microsoft Cortana, Apple SIRI, and Google Duplex going to respond when you ask for a product recommendation?

Without data in a machine readable form, you most likely will not get an answer. So what to do?

1) Collect all of your owned content. Besides your advertising, you probably have custom articles, white papers, customer service scripts, sales documents, and even legal/regulatory guidance (customers facing).

2) If you have the chops, put all this into a graph database and give everything a PermID or URI (Uniform Resource Identifier). If technology resources are limited, you will want to publish to Wikipedia which will provide the mark-up necessary to have your data available in the Linked Open data cloud (LOD), a method of publishing structured data so that it can be interlinked and read automatically by computers. In turn Wikipedia as the people facing resource is also available as dbPedia, which is machine readable and is one of the most popular training, testing, and validation sites.

3) Likewise for images, use ImageSnippets which will provide LOD markup . One of the benefits is for asset management and governance where third parties (like ad agencies) can access the one true brand logo.

4) Depending upon industry (like FIBO for finance), it might make sense to create an Ontology or Knowledge Graph. While complex and often expensive, ownership of a brand and category Ontology could be a strategic competitive advantage in the development of Artificial Intelligence and Natural Language Processing applications.

The net take away is to embed your brand into the fabric of the internet. Do this using semantic technology so machines are able to understand your brand and its association to other things and concepts.

Ask me anything. Click to read more about AI and branding.

Larry Smith

@lwsmith10011

linkedin.com/in/lwsmith10011