The Present IS the Future: Real-Time Marketing In the Era of the Stream – Part Two

June 7th, 2013

In Part I of this article series, we looked at how the real-time Web has precipitated Nowism as a fundamental shift in how we understand and engage with information. Nowism is a cultural shift to a focus on the present, instead of the past or future.

One example of Nowism in action is Nowcasting, which attempts to make sense of the present in real-time, before all the data has been analyzed, in order to project trends sooner or even continuously. Nowcasting is quickly becoming a necessary and powerful function in our media, culture and society.

These ideas are being harnessed by savvy brands and companies not only in how they operate, but also in how they conceive of themselves.

Next we will look how they impact social marketing, and why brands must learn to act like media companies in this new environment.

The Three Stages of Real-Time Marketing Evolution

The Stream is more real-time than the Web. And it’s even more real-time than blogging and the early days of social networking. But it’s not only faster, it’s also orders of magnitude bigger. Instead of millions Web pages every month, we’re dealing with billions of messages every day.

There’s vastly more activity, more change, more noise, and when trends happen they are more contagious and spread more quickly. It’s therefore even more important to sense and respond to change in the present, right when it happens.

Unlike the Web, the Stream is constantly changing, everywhere, on the second timescale: It is a massively parallel real-time medium. And instead of a few channels there are billions of channels — at least one (if not many) for each person, brand, organization and media outlet on the Net — and they are all flowing with messages and data.

Keeping up with the deluge of real-time conversation across so many channels at once is a huge challenge, but making sense of so much change in real-time is even harder. Yet, even harder still is intelligently engaging with the Stream in real-time.

These three objectives represent three levels of maturity and mastery for real-time marketers.

Stage One: Week Marketing.

Today, most brands and agencies are still stuck trying to accomplish Stage One, if they are even that far along.

Stage One social marketers are focused on simply monitoring the Stream and trying to keep up with the conversations about their brand.

Some Stage One marketers are also actively trying to drive perception and optimize engagement through social media. But their ability to measure the effects of their actions and optimize their engagement are primitive at best.

The timescale of their measurement and engagement with the stream can range from hours to days, or even to the weekly timescale.

Stage Two: Day Marketing.

A smaller set of organizations have learned how to make sense of the Stream in real-time and are operating on the hour to daily timescale. They have graduated to Stage Two.

Stage Two marketers don’t merely monitor and respond, they digest and interpret. They measure and engage in sense-making and trend discovery. They generate live insights from millions of messages and incorporate these insights into their thinking and behavior on an hourly to daily basis.

Stage Two social marketers have evolved past the stage of simple reflexive response to the stage where they can interpret and reason about the Stream intelligently in near-real-time. They leverage social analytics, data mining, and visualization tools to facilitate insight, and this leads to smarter behavior, more optimal engagement and better results.

But Stage Two organization response times are still not real-time. Instead of seconds or minutes their responses often take hours or even days and that’s not fast enough anymore.

Stage Three: Now Marketing.

Stage Three marketing organizations are incredibly rare today. But there will be more of them soon.

Stage Three’s are able to monitor, make sense of, and then engage intelligently with the Stream, all in real-time, not after-the-fact.

In other words, they are consistently able to detect, measure, analyze, reason, and respond to signals in the Stream within seconds or minutes, or at most within the hour.

Stage Three organizations continuously run a real-time marketing feedback loop that works like this:

  1. Sense. First a signal is sensed: it might be a breaking story rumor, a complaint by an influencer, a change in customer perception or audience sentiment, a shift in engagement levels, a crisis, or a sudden new trend or opportunity. Sensing signals, and differentiating important signals from noise, in real-time, requires new approaches to determining relevance, timeliness, importance that don’t rely on analyzing historical data. There is no time for that in the present. Sensing has to intelligently filter signal from noise by recognizing the signs of potentially interesting trends, regardless of their content. Organizations that can do this well are able to detect emerging trends early in their life cycles, giving them powerful time advantages.

  2. Analyze. Next, the signal is analyzed in real-time to understand what drives it, and what it drives. Its underlying causes, influencers, effects, demographics, time dynamics and relationships to other entities and signals are mined, measured, visualized and interpreted. This requires new live social discovery and analytics capabilities – what’s new here is that this isn’t merely classical social analytics (measuring follower count or message volumes, or charting historical volume), it’s massive “big data” mining and discovery in real-time. It’s live prospecting around all the potentially relevant signals that are connected to each signal to get the context.

  3. Respond. Then an intelligent response (itself a signal) is generated across one or more channels within seconds to minutes. For example a reply or an offer may be sent to a customer, an alert may be sent to a team, a new piece of content may be synthesized and published, the targeting for an ad campaign may be adjusted, or the tone of social messaging may be modified, prices may be adjusted, and even policies may be changed – all in real-time. Organizations that get good at this process are able to respond so quickly they cross the threshold from being reactive to being proactive. They are able to drive the direction of trends by being the first to detect, analyze and respond to them.

This feedback loop is exemplified in real-time social advertising targeting and campaign optimization, for example. But as organizations mature to Stage Three they must learn to apply this same methodology to ALL their interactions with the Stream, not just advertising. They must apply this feedback loop across ALL their engagement with customers, the media and the marketplace.

Within a decade, all leading brands will be stage three marketing organizations.

Mapping The Ripple Effect

“Stage Three” agencies and brands need to master ripple effects to thrive in the next generation social environment.

Ripple effects are the key forces in the emerging real-time social Web. Information propagates through ripple effects along social relationships, across channels, communities, and media. Ripple effects are how trends emerge and rise, how rumors spread, and how ads and content are distributed. But we’re currently almost completely blind to ripple effects, we have almost no way to detect, measure or predict them.

The average Facebook user has 190 friends. The average Twitter user has 208 followers. Each group contains a number of influencers. Within each influencer’s social graph there are another set of even more powerful influencers. And so on and so on. When you seed a branded message on Facebook, for example, it’s not a straight trajectory. A ripple starts with the above numbers, but each new impression creates a new set of ripples.

Suddenly, your branded message is being seeded across a number of platforms — even spreading to platforms you never intended — and 99% of brands and agencies have no way of mapping this ripple effect, let alone controlling it.

But what if you could track your ripple effects? What if you could guide them? What if you could measure their effectiveness, or even predict where they are going? Suddenly there would be a wealth of new insights to pull and learn from. And this is precisely what is now possible, using emerging tools.

Problems with Existing Tools

Stage Three marketing organizations have to keep up with ripple effects in real-time, and they have to anticipate where those ripple effects are headed, in order to react immediately.

But most social analytics and engagement tools fail to show ripple effects. They provide loads of raw data — lists of messages from various social accounts and searches. But they expect humans to do most of the work of actually figuring out what’s important in those lists of messages. That is no longer realistic. Humans can’t cope with the data — it’s overwhelming.

Furthermore, most existing social analytics tools focus on simply measuring engagement via follower counts, mentions, likes, Retweets, favorites, impressions, click-throughs, and basic sentiment. But those metrics are no longer sufficient: They aren’t the trends, they are just signals that may or may not be relevant to actual trends. Not all signals are trends. The art is in finding a way to pull out the actual trends from the rest of the signals that are not in fact trends of any value.

What existing tools fail to do is actually make sense of what’s going on for you — they show you either too little or too much information, but they fail to show you what’s actually important; they’re not smart enough to figure that out for you.

Existing tools are good at finding known topics and trends (“known unknowns”) things you explicitly ask to know about in advance — but what we need in the era of the Stream are tools that show you what you don’t even know to ask for (“unknown unknowns.”) They have to detect novelty, outliers, anomalies, the unexpected — and they have to do this automatically, without being instructed on how to find these nuggets.

Existing social media analytics tools are too retrospective in nature – they show how a brand performed on social channels from the past up to the moment a question is asked. But these reports are static. They don’t show change happening, they don’t say anything about what’s next. The minute they are generated they become obsolete. It’s interesting to look at past performance, but what is really needed is more predictive analytics.


We need a new generation of tools that are designed for identifying real-time ripple effects and filtering them to figure out which ones are noise and which are actual trends we should pay attention to. Better yet, we need tools that can not only identify the trends, but that can project where they are headed in real-time. Think of this as the next evolution of Nowcasting.

We might call this Trendcasting. Where Nowcasting figures out what’s happening now in real-time, Trendcasting figures out what’s happening next in real-time. 

No human can Trendcast in real-time anymore without help from software, the Stream has too much volume and velocity for the human mind to comprehend or process on its own. This is a problem that can only be solved by cloud computing against big data in real-time.

In the today’s real-time Stream, marketers cannot afford to be hours or days behind the curve. They need to know and understand the present in the present, while it is still unfolding. They need tools for Trendcasting – for finding and predicting trends. Trends are not merely raw data, they are particularly meaningful and noteworthy trajectories in the data.

Trendcasting is going to be come absolutely key. The next-generation real-time marketing platforms will provide automated trendcasting as a key feature. They will sift through the noise, find the signal, and then measure it to see if it actually matters. Trendcasting is about filtering for the trends that actually matter, because not all signal is important and not all trends are equal.

Traditionally, finding and forecasting trends has always been thought of as an exclusively human skill — but today we’re starting to automate this function. I believe Trendcasting can be fully automated, or at least dramatically improved, using massively parallel big data analytics approaches. This is the where the cutting-edge of innovation for real-time marketing will focus for the next decade. (Disclosure: My own company, Bottlenose, is focused on exactly this goal, for Fortune 500 brands).

Trendcasting tools are the next leap in a long process of measurement tools innovation that has included inventions like telescopes, microscopes, X-rays, weather satellites, MRIs, and search engines. In a sense trendcasting engines could be thought of as automated cultural measurement tools — the social equivalent of a weather satellite —  social satellites. They help us to visualize, understand and project the weather of markets, cultures, industries, communities, brands and their audiences, just like satellites have helped us understand and map the weather patterns of our planet.

Every Brand is a Media Company

The shift to real-time and the advent of the Stream changes how brands must think of themselves.

Whether they are ready or not, all brands have to learn to function more like media companies – and in particular like news networks – in order to remain competitive in the era of live social media.

For the first decade of social media the emphasis was clearly on social, but now it is shifting to media. Leading brands have learned how to be social for the most part. Now they have to learn how to act like media companies.

Consider a network like CNN: They have reporters all over the world, constantly giving them text, images, video, opinion, insights and leads. They have viewers, some of whom are also contributing news tips and stories, and opinions, all over the world across many platforms and channels.

CNN’s bread and butter is finding breaking stories first, getting the best information about them, and covering them most comprehensively and creating original news content and analysis for their audience.

CNN is a good model for what every Stage 3 brand has to learn to do in order to master Now Marketing.

Brands that want to lead in the Stream era have to gather intelligence constantly, using social media. They have to create content, share it, and engage. They have to keep their fingers on the pulse of their markets and culture in general in order to remain relevant and timely. They have to respond to a huge influx of questions, opinions, complaints, suggestions, leads. And they have to do this across many platforms and channels at once, in real-time.

The distinction between content provider and audience is dissolving. It’s now a two-way live conversation with the market, a conversation among equals. Brands have to learn to share, interact, make friends, and socialize just like people do. They have to not only create content for their audiences, they have to use their audiences as the content. And they have to do it on a massive scale.

Some brands – like Nike and Red Bull – have gone very far down this path and even think of themselves as media companies to some degree. But for most brands thinking like a media company is still a completely new orientation and set of skills.

Brands need new tools in order to think and operate like media companies. They can’t work on the weekly or monthly timescale anymore. Even daily timescales are too slow: they have to go live.

They can’t just market to their customers, they have to engage them in marketing the brand and creating media, together. They can’t just analyze key metrics anymore, they have to understand the trends that are emerging, and what’s driving change.

The Stream is here, and it’s happening in real-time. Marketers who can adapt to this shift early will be the leaders of tomorrow; Brands that are late in adopting these practices risk become nothing but historical data points.