The advertising business has always been supported by data. In the past, this was done in the form of reports you could buy from folks like Nielsen (TV) or Arbitron (Radio). Looking at last month's data, or last year's data, you could make more informed media purchases to achieve your desired goal. Today there is a blossoming world of data that is for sale and used in real time, in auction based media transactions. This cookie is an “auto intender” or this device is a female 25-34. However, much of how this data is utilized needs to be looked at differently than how data is conceived today to truly take advantage of what is possible.
If you take a look at the history of the internet, the Yahoo directory was once the biggest destination on the web as it categorized websites. If you had a golf website you could submit to be categorized under “Recreation and Sports”, then under “Sports”, and then “Golf”, and you would be listed alongside many other golf websites. This was very basic but provided some organization to the early web. What became clear was that there were too many websites and pages that were being created, and traditional categorization no longer worked. Also, as a user, if you were searching for golf clubs vs. the rules of golf vs. making a golf reservation, the categorization didn’t solve for your intentions. Google’s search algorithm broke the internet down into not only websites, but web pages, and individual keywords. This allowed for massive scaling of internet search via an empty box where you placed your intentions. You can search for any combination of keywords and you will get search results in order of relevance from Google within milliseconds. As we all know, this was a game-changer.
The advertising data business today looks similar to the Yahoo days of website categorization. If you are an automotive advertiser, you can buy 3rd party data of “auto intenders” or, if you are a CPG company, you can buy users that fit broad demographic segments such as Females 25-34. In some cases, the data is sold separately from the media you are buying. Understanding its value and how it can impact a given campaign objective is hard; borderline academic. If you utilized data sold in aggregate buckets you will struggle to distinguish the individual value of each data point. There are also thousands of data segments for sale, so understanding the true value of data segment A vs. B vs. C is nearly impossible. In the same way that Yahoo had a hard time categorizing certain sites that may fit into multiple categories (or no category), many advertisers will never find a 2nd or 3rd party data segment that is a 1:1 match to the users they really want to reach. Therein lies the challenge.
If you think of how the data business will evolve, my guess is that it will look similar to how Google search evolved. This means that 3rd party data, as it is described above, will have far less value. Modeling of a marketer's first party data will have a lot more value, and “data” will not exist in a warehouse or platform solution you can buy in a static form. Instead, it will become the underlying nature of how you view your business and being able to apply this at a granular level to advertising efforts will pay large dividends. In Google’s case, they are able to rank and understand the web as it relates to keywords (your intentions), Netflix for movies, Uber for transportation, Amazon for e-commerce, and Facebook for relationship connections. There is something fundamental and unique to these companies, and it makes them some of the most powerful companies in the world.
Real-time, customized, and measurable are the traits these companies have mastered and are using to drive their businesses forward. The most important data to them is how you interact with their assets, as opposed to matching your behavior to someone else’s definition of it (3rd party data). Netflix cares about what content you consume on their platform as it helps them recommend other relevant content. Amazon cares about what you shop for and buy as they can scour their entire portfolio of products and recommend additional products to you. Does Netflix know your gender? Does Amazon know your credit score? Who cares? Would this data be interesting to them? Sure, and it may help in a very small way in making their own recommendations better, however, the most valuable data to them relates directly to the interactions you have with their given product/service offerings.
The real-time economy is fundamentally built on different business principles. If you were to ask someone what and Uber ride costs they wouldn’t have an answer. There are many factors that play into what you will pay, but when you request an Uber you are a buyer in the market affecting every other buyer and seller in that market. The price of your Uber will depend on how far your ride is, whether there is surge pricing in that given area (meaning more demand than supply), what type of car you request, how many drivers are available, and eventually things like your Uber rating, your frequency of using the service, and other factors. The same thought process will be applied to media in the future, especially as it relates to programmatic, where many different factors will play into the cost of media. The price will be determined in millisecond auctions and the bulk-buying mentality that exists in media buying today will continue to fade away.
Business models of the past looked for duplication as it helped them scale. People are buying the red sweater vs. the blue sweater, so let’s make more red sweaters. The business model of the future makes value out of every consumer interaction being differentiated. Yes, you can still sell red sweaters and blue sweaters but understanding the intricacies behind those sales matters. What are all of the unique details about those sales? These intricate understandings of a customer is what separates the new wave of technology powered businesses vs. the old wave businesses of quality product makers. Every Google search is unique, ever Uber ride is unique, and these companies use this uniqueness to make their next interaction with you, and everybody else using their systems, better. If we both search Google for “car insurance” the keywords are the same but if you combine that with the geo-location, the device we are searching from, and user search histories, we will both get slightly different results which are more relevant to each one of us.
As the title of this post suggests, the data business is similar to the idea of holding mercury because the minute you try to touch mercury to define it, it will split into smaller pieces. If you touch those pieces they will split again, and again, and you can’t control how it will split apart. The business models of the past can’t make sense of this as it doesn’t fit the mold of capturing similarities, defining them, and then scaling. The business model of the future captures the differences in consumer interactions and leverages this to help them scale and provide a unique offering for each given customer. Today, the computing power exists to effectively understand differentiation and uniqueness of consumer behavior and the biggest businesses in the world are using this to catapult their growth.
Part of how we at Wholetone work with advertisers is to provide a deeper level of granularity on their customers and the media that is being purchased. If you can set up your business in this way you have a far better understanding of who your customers and prospects are, which gives you the opportunity to engage with them in a more relevant way. We believe that all successful businesses of the future will be built with a fundamental understanding of this.