Spectrum Media Services’ core focus is on leveraging recent content consumption as a signal of consumer intent. The Spectrum Platform is our proprietary contextual & audience platform, which has the ability, in real time & at Web scale, to understand the meaning of a page at the topic level – not just through a few isolated keywords.
The Platform listens to content consumption event streams, reaching millions of users consuming billions of URLs. The Platform uses its topic-level understanding of content to score these media opportunities based on how likely they are to engage the consumer with your marketing message, producing page-level contextual or audience Content Targets for targeting, reach extension, & analytics.
ATTRIBUTES & DIFFERENTIATORS
Spectrum provides brands, agencies, data companies, & DSPs with a scaled, strong intent signal that leverages:
Targeting: Capture broad or focused targets with ease by using example URLs that match your intent
Relevance: Topic-based matching engine avoids the relevance pitfalls of keywords
Environment: Scoring & targeting of only article pages, not channel/home or social
Recency: True 30-day lookback window on content consumption
Reach: Qualified cross-tactic reach extension based on insights
Scale: 250M users consuming 3B URLs across 50K sites, with 350K new events/second
Transparency: Detail & summary on matches, helping maximize accuracy & brand safety
Insights: Endemic & non-endemic insights based on observed content consumption behavior
Portability: All of the above, configurable for your event source & topic model
THE CONTENT TARGET
The core concept of the Spectrum Platform is the Content Target. A Content Target (CT) is a persistent, reusable representation of a search process around a given theme, topic, or brand. Fundamental to the CT is the concept of the Target Set - a collection of URLs acting as ideal examples of web pages where ads for the CT topic would run. A CT, once built, can then be activated to materialize contextual or audience-based segments suitable for targeting in any buying platform.
AUDIENCE PLANNER TOOL
We use the concept of concentric circles to assemble targeting plans for clients. The circles are a guideline for producing data-driven scale, from endemic to non-endemic.
The Spectrum Audience Planner is able to compute unduplicated reach numbers across such a portfolio, giving clients the actual uniques they should expect in a variety of formats.
Here’s an example for a client targeting people consuming content about grocery shopping, foods/recipes, & entertaining (i.e. grocery shoppers). As you can see, the de-duplicated audience shows high unique MAID reach across the 5 segments in the plan.
Unduplicated reach computation is one form of overlap analysis that’s useful applied to a set of segments. Zooming in, between any two segments, overlap takes on additional meaning:
(1) High overlap between contextually related segments validates expected relationships and
(2) Low - but non-zero - overlap reflects the nature of readership on the Web, and drives non-endemic affinities
The Spectrum Audience Planner reports on segment-wise overlap in two ways - intra-plan, and as compared to our in-house taxonomy that covers a wide variety of topics. Below is the topic overlap analysis for the retiree audience.
This process often reveals both endemic and non-endemic insights. It’s intuitive and validating that topics like health insurance and the stock market are things a retiree might be reading about. Fashion shows, perhaps less so, and a potential new target for a marketer looking to reach their target audience in new ways.
Spectrum’s topic overlap analysis helps show clients that their targeting plans are well-fit, and where else they can target based on recent observed content consumption behavior.