Many of the defining global tech companies are built on proprietary graph data, and some of these have in turn used that strong core position to go from advertising technology to “all kinds of”-technology and industries. I will use Unacast’s proprietary graph to show how.
There are many ways of dissecting the modern industry-defining software companies and products. The usual suspects that have for quite some time now basked in the bright sunny-side of tech: Facebook, Google, LinkedIn, and also newer entrants like Spotify, Tinder, and up and coming Slack. And many, many more.
These companies can be looked at through a wide range of optics, from how they connect people in an increasingly globalized world, and how they have used the aggressive VC scene of the last decade to build impenetrable fortresses, to how they have attracted the best of the best talents by building work environments people actually want to work in, and how they never stop innovating – wanting to forever be the startups they once were.
When I look at these companies, I first and foremost try to identify the graphs they are built upon. And by graph in this context, I mean the data sets derived from the interconnection of people through a specific denominator – giving a unique and valuable view on those relationships.
Case in point. Facebook is built on the social graph, Google on the search graph, LinkedIn on the job graph, Spotify on the music graph, Tinder on the dating graph, and Slack on the business communication graph. These companies, by largely owning the data well, the drilling rights to that well, and being the best at drilling it, makes it hard for other companies to venture into their domain and overtake them. Because they know more and continue to know more as they amass more data = knowledge.
Secondly, I try to identify the business model from where the company originated. Of the six mentioned, the first three companies are built on advertising revenue-centric models, being ad-tech companies at core, and the latter three on user payments – although some with growing advertising supported revenues.
In this post, I am going to focus on the intersection of graphs and advertising, and how that base can be used as a vantage point to venture into new industries, like especially Google has done with unprecedented success, and where Facebook is following quickly and aggressively. I’m going to use Unacast’s proprietary graph to demonstrate my point.
We believe there are room and a need for (at least) one more industry-defining graph, and that is a graph that understands the real world out there, where we move about and with who, and not only what happens when interacting with a screen - underscored by the fact that we humans use about 70% of our time interacting with the physical world around us. Unacast is building that graph.
It has been difficult in the past to get closer to this understanding as there has been limited data to extract from the real world, but that is changing quickly with the rise of various sensor technologies.
Read more about the rise of sensors at Proximity.Directory
These, combined with more traditional GPS data, that is often imprecise and lacking context, and other more static data sources like demographics, says something about how people and places are connected. This data is aggregated, matched, and improved in our Unafy engine and all data sets given a Unascore which will be transparent to the user of the graph.
We believe that graph is going to form a powerful basis for a multitude of new products in many industries, as well as supporting existing ones. Not surprisingly we have named it the Real World Graph™.
To repeat, all graphs are inherently valuable in a myriad of use cases, as the data sets themselves are by definition agnostic to the industries they can provide value to. Think of it as the foundation on where products are built, and it is the products that define the use. The below are some of the many applications we see of the Real World Graph.
The first one is obvious, and where we, as Google and Facebook too, have initially focused. How can the added physical dimension make sure the right message is delivered at the right time? We don’t build advertising products for the end-user ourselves but support others by improving their products through retargeting, attribution and data modeling.
Many of the more popular products of today are based on location, with Tinder and Uber being two recent examples. With the added precision from the Real World Graph, Tinder can greatly improve the precisions of their products and use actual physical locations as “virtual chat rooms”. Tinder could effectively turn any bar or restaurant to the confines of their now famous left swipe/right swipe system so that the visitors could engage easier with the ones at their specific location that moment in time. This is also a relevant example of how a new graph can create new monetization possibilities.
Another recent example is Pokemon. With the added data from the Real World Graph, one could place Pokemons behind a specific clothes rack at H&M, driving in-store traffic to H&M, new revenue to the Pokemon franchise, and new gaming mechanics to the end-user.
Finance and hedge funds are all about the microseconds and being able to take more decisions quicker than the competition. Now that the trading desks are placed physically next to the stock exchange to shave off as much time as possible, their focus has turned to using data to understand the world – before quarterly results are published. A typical use case is understanding foot traffic at stores (and soon foot traffic at the product level) so that bets can be placed before the market has full information.
No hedge fund in the future will be successful if they don’t have the Real World Graph to support their billion dollar bets.
I have written two extensive posts on this topic lately, on the back of Amazon’s acquisition of Whole Foods. By using the Real World Graph a retailer can understand where their customers are before and after a store visit, completing the customer journey so that products and communication can be better tailored to that specific customer – bringing her or him back to store (either digital and physical).
“After getting the data you have on a platform where you can utilize it, and you have started to implement technology to gather data you don’t have yet, you inevitably come to this conclusion: You can only see and understand my customers when they interact with your entities. To put it another way, your customers seize to exist before and after a visit to your location(s)”
The same logic can also be applied to vastly improve the intelligence of pure e-commerce platforms so that the products shown fit previous physical history. When I visit HM.com, I should first and foremost see products related to products I have been close to in-store.
We all believe Virtual Reality is going to be big, sometime in the future, but AR will probably rise to stardom first. With Apple building AR into their core operating system developers will soon have plenty of tools and distribution to get AR products in front of end-users.
AR really shines when it fully considers the world around it, and although the camera can place a 3D object at the correct location, there will be little understanding of the context the phone is in, the products it is near and how user patterns correlate to the specific location and context. The Real World Graph can add that layer of understanding to AR products, making sure they can be utilized both for the benefit of the end-user and brands.
There are two main drivers behind the rise of AI: Processing power (the ability to take many decisions fast) and the amount of data available. The latter is why the big graph companies are leading the AI revolution.
Any leading AI platform in the future will need to consider the physical dimension to better understand the world around it and to provide the end-user with the desired advice and context. GPS won’t cut it alone, as it lacks the precision and context. We’re probably going to see the use of this graph first in the AI assistants, like Google, Alexa, and Siri.
I hope I have managed to showcase some of the possibilities a strong graph data platform gives to a wide variety of industries. There are many use cases I have not mentioned, partly because of keeping this post digestible, but mainly because I don’t know about them yet. That’s why we have been so adamant of basing the Real World Graph on the Swiss Army Knife-principle, ready to tackle any known and unknown challenge.
Building the Real World Graph is a continuous project and the map in hand is changing as we learn more about what data sources provide the highest level of confidence over time by opening up our Unascores to our partners for increased transparency and quality. In addition, new real world data sources are still being introduced to the market place, and it’s our job to slot these new signals into the graph to keep enlightening the dark spots on the map.
The big global graphs, that built big global companies, have one defining trait in common: They are walled-in gardens, only really usable from inside the garden, behind the moat. And that is exactly what separates these graphs from The Real World Graph. This graph is built to be accessible for all companies in all industries.
Will this mean the new graph in class will reach even farther than the ones that came before?
That’s too early to say, but we are confident it will help us conquer the world beyond advertising.