CB Insights Trends is software that analyzes millions of media articles to programmatically identify and understand the rate of adoption of emerging technologies and innovations.

In a time where we have 3D printed rockets and genome sequencing for $1000, why are our methods for identifying and understanding emerging technology trends so abysmally antiquated?

A prime example is in the below graphic of industry analyst estimates of 2 markets — virtual reality and drones.

pundit industrial complex

A quick review of the above estimates highlights the following:

  • The drone market could be $1.27 billion (very oddly precise) by 2020 up to $10 billion.
  • The AR/VR market estimates start at $1.06 billion (yes, that’s $1 billion 60 million) and go all the way up to $150 billion.

For some reason, we’re forced to settle for the random guesses of the Pundit Industrial Complex. You might compare it to throwing darts, but that would be insulting to darts.

Even harder than forecasting the size of a market people already know about is the early identification of emerging technologies you should be aware of.

Here the data is even less available so you’re forced to rely on anecdotally-driven frameworks like the Hype Cycle which purports to “give you a view of how a technology or application will evolve over time.”

Simplistic frameworks like these have created their own trough of disillusionment as can be seen here, here, here, here, here, here, and here.

We’ve seen baseball get reinvented with sabermetrics or the debunking of political forecasting pundits by the likes of Nate Silver and the use of predictive analytics in industries ranging from financial services to healthcare to oil & gas.

Isn’t it time our understanding and identification of emerging technology trends also begin to rely on software and probability and not snake-oil and punditry?

Introducing CB Insights Trends.

Using media attention to predict technology trends

CB Insights Trends applies machine learning to a massive corpus of media articles to enable a data-driven, real-time method to discover, predict, and plot the arc of rising expectations and excitement (and potential eventual disappointment) for emerging technologies. We call this arc the Media Attention Path (MAP), which we’ll discuss in greater detail below where we dig into why media attention is worth tracking.

But first, let’s start with a few examples of how CB Insights Trends works and what it can uncover about emerging industries as well as corporate strategy.

Have clinical wearables overtaken general fitness wearables?

Using Trends, we can see the growing excitement around wearables starting in 2013 followed by a more recent flattening of interest. Does this flattening mean that wearables are over?

wearables media


Over this same timeframe, Trends reveals that the zeitgeist around the “quantitative self” movement has abated dramatically as can be seen below.

quantified self


Quantified self was generally associated with wearable devices focused on fitness or exercise. So while general purpose fitness wearables have lost steam, we can use Trends to discover that disease-specific or clinical wearables (used for Parkinson’s or diabetes for example) are actually seeing increased momentum.

clinical wearable

With Trends, we can discover the shifting attitudes and areas of interest within a sector as shown above.


Has the Internet of Things eclipsed Big Data?

Let’s now look at two of technology’s buzziest areas — Big Data and the Internet of Things (IoT).

And as we can see, IoT has taken over from Big Data when it comes to media interest.

big data versus internet of things


But the Internet of Things is one of those seemingly generic terms that means everything and nothing at the same time. So within IoT, is there any sub-area that is gaining interest alongside the general IoT space?

Trends reveals that the Industrial IoT movement continues to gain steam. The Industrial IoT refers to asset-heavy industries like manufacturing, logistics, mining, oil, utilities, and agriculture which have also begun to apply IoT technology to improve efficiency and develop new business models.

industrial IoT


But again, we uncover another layer here within Industrial IoT and can see that sensors eclipses chatter around machine-to-machine or M2M indicating this is one area for which media commentary continues to increase.

Industrial IoT sensor, machine to machine, m2m


Again, with Trends, you can quickly identify a macro trend but also drill into sub-elements of it to understand the market at a more nuanced level highlighting the underlying trends and areas of interest within the industry.

Is Google still dominating autonomous vehicles?

Trends’ versatility also provides an interesting vantage point into corporate strategy and competitive dynamics which we can illustrate through a look at autonomous vehicles.

First, there are a few terms used in the industry that are synonyms for autonomous vehicles.

Which one is most common? Is it autonomous, self-driving, or driverless?

Trends reveals that autonomous nudges out self-driving while it is apparent we’ve moved on from the term driverless.

autonomous vs driverless vs self-driving

In the next graphic from Trends, we also plotted media mentions of autonomous vehicles that also included a mention of Google which was a pioneer in this area.

What Trends illustrates is that discussion of autonomous vehicles was almost synonymous with Google last April (2015) but this has changed in the last year as demonstrated by the growing gap between the blue and orange lines. With 30 corporations ranging from Baidu to Tesla to General Motors working on driverless vehicles, competition in the space has intensified and so media attention has become diffused among multiple players who all have initiatives around autonomous vehicles. Tracking media mentions highlights the growing level of competitive intensity in the space.

autonomous vehicles google

Speaking of new entrants including GM, we see the company is making moves in the autonomous space, actually pulling ahead of Apple whose autonomous vehicle efforts remain in stealth or rudderless depending on your perspective. Here, we see Trends illuminate corporate strategy. Company strategy is best seen by looking at what companies are actually doing (investments, acquisitions, partnerships, products) versus what they’re just saying. Media attention provides a view into actions — not just rhetoric.

GM Apple Google driverless car



Why the media illuminates tech trends

When properly organized, mined, and quantified, media attention serves as an in-house market research panel on steroids — a pre-qualified and willing group of “respondents” whose writing can be surveyed on a real-time basis to discover, predict, and plot technology trends and innovations.

Our analysis of millions of news articles reveals that tech trends tend to follow a few patterns or what we think of as a Media Attention Path (MAP).

The first phase of the MAP for most technology trends is the Discovery phase.

In the Discovery phase, niche trade publications or enthusiast blogs start writing about a technology and shine a light on still nascent technologies and innovations. These are often folks described as early adopters or enthusiasts or innovators and tracking what they are talking about is a great predictor of emerging technology trends. See the below examples for virtual reality (VR) and artificial intelligence (AI).

virtual reality artificial intelligence


What we see in the Discovery phase in the case of VR and AI is steady growth of news mentions over time. A couple of notes on the Discovery phase:

  • Different durations – emerging technologies stay in the Discovery phase for different amounts of time, i.e., there is no typical duration for the Discovery phase. Some remain a niche topic for enthusiasts, subject-matter experts, and vertical media outlets for a while and gradually become more mainstream while some explode into our collective awareness due to some event-based trigger.
  • Some tech trends never get past Discovery – In the case of VR and AI, we see a steadily increasing slope of media mentions. However, some technologies never get past the Discovery phase. They remain niche and fail to gain adoption or they get subsumed into a larger technology trend that takes off.

After the Discovery phase, things get more interesting as technology trends don’t all follow the same MAP. We’ve identified 3 typical MAPs and provide examples of each below.

  1. Boom & Plateau
  2. The Steady Climb
  3. The Rolling Hill

Note: While we understand the allure of nicely stylized frameworks like the hype cycle that put all technologies into a curve that follow the same path, unfortunately, we’ve found that technology trends do not behave so consistently.


MAP #1 – Boom & Plateau

Sometimes, tech trends tend to gain adoption through a series of event-driven booms. Virtual reality is an example of a MAP that follows the boom & plateau pattern. We see Facebook’s announced acquisition of Oculus drive interest in the space (the boom) followed by a pullback (the plateau).

virtual reality boom and plateau


In a Boom & Plateau MAP, there is often a dominant company in the space. When a dominant player exists, the MAP moves in lockstep with events centered on that company. Virtual reality is a prime example of this as can be seen below where the technology was essentially synonymous with Oculus initially. Over time, as more players entered, attention diffused over multiple players.

virtual reality competition


Once there are more competitors in the industry, its MAP often changes and follows a more steady climb. Of course, some tech trends just progress steadily from the outset as detailed next.


MAP #2 – The Steady Climb

Sometimes, a technology trend just gains steam after the Discovery phase. The Internet of Things and Big Data as shown below are examples.

big data versus internet of things


The important variables to understand with a technology trend on a steady climb are (1) how much news volume is increasing over time, i.e. how much the slope of the MAP is increasing (or decreasing) and (2) the volume relative to other industries as this validates whether something is really a trend.

Point 1 above is obvious. A trend getting more media attention to the point its trajectory (slope) is changing highlights heightened interest in it (or decreased interest as the case may be).

The 2nd point is a bit more nuanced.

We’ll illustrate as follows using the term “gamification.” The first graph below for gamification highlights a MAP that although volatile shows an increasing trend (as indicated by the gray dotted line).

gamification trend


Based on just this graph, you might draw the conclusion that gamification is a technology trend worth watching. However, when we compare against a more well-known trend, you can see the relative strength of gamification doesn’t suggest it is a trend at all. In this case, we compared to artificial intelligence in the graph below.

The MAPs of gamification vs artificial intelligence highlight that the gamification trend may not really exist in a meaningful way right now. This could mean that the trend remains in the Discovery phase or is, in fact, not really a trend. AI vs gamification

Finally, it should be obvious that technology trends going through a steady climb never continue forever. The Trend may become mature and mainstream (internet or web 2.0), fragment into smaller more specialized sub-trends in which the larger trend is implicit (conversations about chatbots eventually assume the use of AI) or in some cases get a bit of a rebrand (driverless car shifting to autonomous vehicle).


MAP #3 – The Rolling Hill

Sometimes, a Trend comes to light and then fades away. The MAP in this case resembles a hill hence the name. These are less trends than fads in reality.

Quantified self as shown earlier is a great example of this.

quantified self


What’s next for Trends at CB Insights

A few final notes on the significance of Trends for CB Insights. At its core, Trends gives our clients another weapon to use to predict and discuss technology trends. It’s our most significant product update since the Fall of 2015 when we announced our $10 million Series A and rolled out several new products.

In terms of its importance and applicability to CB Insights, some thoughts below:

  • Mosaic – Trends and technology MAPs will figure into the Mosaic scores of companies. Mosaic is our algorithmic view / score on private company performance. One of the critical factors in understanding a company’s health is the health of the underlying space or industry the company competes in. Trends provides a data-driven and objective view into technology health which will figure into Company and Market Mosaic scores.
  • Competitive Intelligence – To date, many of our clients have used the M&A and investment history of competitors to track their rivals. Trends offers them another view into the strategy of competitors.
  • Predictions on what’s next – Clients have used our data on financings, whether they be government grants, angel investments or VC deals, as well as M&A to understand where technology trends are going. Trends represents another piece of that puzzle, and our vision is to go deeper with more datasets and continued algorithmic analyses of the existing datasets we have.

CB Insights Trends provides an objective, real-time, and data-driven perspective on the technologies and trends that business strategists, chief innovation officers, R&D leaders, product marketers, emerging technology teams, investors, and governments should consider in developing their strategies and responses to emerging technologies and competitor actions.

Trends also delivers on our vision that the best way to understand and predict emerging technology trends is through the use of technology.

We look forward to seeing how you use CB Insights.

  • Jeff Stoltman

    As a huge fan of The Hype Cycle and CB I am delighted to see the level-up thinking, and the data power behind it. Kudos!

  • Juan Tatay – silta

    Insightful. Thanks.

  • Meenu Sarin

    Great insights into trends watching. Thanks.