Timing is Everything: Time Segmentation as the Future of Data

This is a follow-up for the last post I wrote, Time Segmentation for Streaming Music Radio, which explores the concept of time segmentation in a detailed, product-specific environment. This post will take a broader, industry-unspecific point of view, focusing more on the theories and narratives behind them that may be better suited for strategic planning and vision-building.

A time-segmented analysis of media usage from The Nielsen Company. Highly recommended reading!

The 5 Ws — Who, What, When, Where, Why — are the basic questions people ask before starting any research or data-gathering project. Internet-based software products were initially heavily focused on the Whats (search/keywords), moved onto the Whos (social media), then the Wheres (mobile/geolocation) as a historical progression. Over the years, the methods that were devised around these information sources have gotten more sophisticated and integrated, leading to many of the products and platforms that we see today.

In the software space, we’re starting to see more experiments done in the When category now, but as of today its developments are still largely in its infancy stages. If developed with deliberation and intent, however, time-segmentation could potentially become “the next big thing” on the Internet because it represents a big part of data science that has remained largely untapped.


A few feature ideas, to start:

Business: Time-Segmentation Marketing, Seasonal Campaign Normalization, Incremental Price-Changes

Design: Time-Sensitive UX/UI/Design (e.g. Google Maps’ Day/Night Cycle)

Engineering: Time-Segmented Content Targeting, Incremental/Controlled Software Updates

Data: Additional Segmentation Options and Data Points

Community Management: “Crash-Only” Designs for Democratically Operated Communities

The nice thing about time-segmentation is that unlike keywords and profiles, time data is more or less objective, making it easier to arrive at conclusions and agreements among team members involved with specific projects. And because the process is just a matter of extracting data from time-stamp information, in most cases new features can be built without the need for expensive upgrades or additions to existing infrastructure.


Content Targeting

Of all of the possible feature ideas mentioned above, improved content targeting is the most potent, likely to yield the most effective results in the long run. The information from the first 4 Ws all lead to a better understanding of the 5th — the Whys of customer behavior and needs. The Whys allow both advertisers and content creators to focus their targeting efforts, greatly increasing the odds of a match in interests and needs.

Current targeting methodologies tend to leave out the Whens, however, leading to an incomplete or misrepresented picture of the customers’ story and journey. Your relationship with a school changes the moment after you graduate, evolving into something even more different years down the line. Your shopping habits during certain holiday seasons are unlikely to be similar to the ones for the rest of the year. The reasons why you might listen to a certain playlist in the morning isn’t the same ones at night, even if the songs themselves are the same.

Time-segmentation basically allows you to fill in the missing pieces of the story with concrete data, reducing the amount of assumptions you have to make for any given campaign. It’s no secret to anyone that current content targeting methods aren’t perfect: people have come to expect misplaced ads and suggestions as part of the online experience, rather than the other way around as it should be. Time is missing key element in representing the complexity and richness of a person’s journey: how it changes throughout the day, and how it evolves over time as part of their personal narrative and history.

Timing is Everything

Because of the limited amount of pixels available on-screen, content and advertising in digital marketing spaces have become extremely competitive in recent years, leading to saturations in the market and limited business development opportunities throughout. Current content targeting methods allow you to target users by keyword, demographics, geolocation, device/distribution type, and maybe a few other tweaks here and there. Time segmentation adds another layer to that process, allowing creators the power to determine when they want their content to display to the user.

In practice, the control panel for content targeting practices will simply include another question: When do you want your content to display to your users? (Premium rates may apply for high-traffic/high-volume times.)

Many software platforms operate under the “open 24/7” principal, not really accounting for the Whens of their Whys. As they say, timing is everything — even if you have all the right cards, if you don’t play them at the right time it won’t make much of a difference in the end.

In future iterations of software products, however, we’re likely going to see a more sophisticated, targeted approach to the idea of time. It’ll be interesting to see because unlike traditional time-segmentation practices done on TV and radio, these segmentations can be automated and personalized for every user, even if they travel a lot or have erratic schedules that doesn’t neatly fall into the 9–5pm work schedule. If the “gig economy” were to become a thing in the future, it might even become a necessity, since we wouldn’t be able to rely on traditional time models anymore, even if we wanted to.

Has anyone noticed other examples of time-segmentation used in software platforms of today’s products? They’re spread throughout the internet in small widgets here and there, but they can be pretty hard to spot since its functions are often hidden in the background amidst the other things running in the front. Always interested in hearing what people think.