The digital revolution has begun to show us its results, and the winner is…
It’s incredible how everything had changed since I started writing this blog more than a decade ago and is not because the human is bringing some new behavior to the table, as a matter of fact there’s nothing new (really new) under the sun. It’s just basically that everything that we use to do in analogical way is getting digital, and it’s not just “a different way”.
It’s actually way more than that, probably like a revolution and like in any other revolution there are winners and losers. We all probably thought that this revolution would be different, specifically in its decentralization capacity, but is there where we are really going?.
People make money pretty much in the same way that they did in the past:
1- Trade based: This is basically companies that generate revenues based on a commission per trading product and services. Could be own product and services (e-commerce) or from other companies (market place). We are talking about a global market for Retail commerce through internet of more than $1.9 Trillion dollars, actually eMarketer was reporting that almost that back in 2016, so probably the number is even higher than that.
2- Media/content based: This was a simple business. Publishers use to generate content that generates interest from people, this interest was monetized by selling advertising to advertisers that pay an amount of money based on the business it can be generated by displaying their messages to the people (readers, audience, users, etc) part of the publishers audience. This is a Global Market of $1.7 Trillion.
The cost of advertising is basically the amount we can pay for converting a person in a client. In other words the amount of money advertisers are willing to pay per unite of ad (seconds, cpm, cpc, etc) is basically an amount that allow them to generate business that pays for the campaigns plus generates a margin for their company. Technology allowed the publisher to:
- Increase control: Having more information, making it possible to connect the advertising with it’s impact, its costs with the business it generated. Knowing what works and what doesn’t work and make ex-ante decisions based on that information.
- Accelerate decision making: Due to real time information, whatever that means (1 second, minute, hour or sometime today) companies can make decision on the go, allowing also constant optimization process through the use solutions like A/B or multivariate testing.
So basically we are talking about of a Global Market between media and commerce of almost 4 trillion dollars.
The digitalization of media and commerce is, as mention at the beginning of this post, not a secondary thing. The possibility of having everything measured is key, but the companies that will take over the market are the ones that can really take advantage of that information. Is not just about having it, but using it.
So why companies are struggling on using data? Because of the lack of data integration. “Non integrated data” is what’s making the biggest companies in world slowing down their growth (in the best of the cases).
But why is so complex integrating data? Let’s take a look at the technology used for business today:
- PMP: Private Marketplaces are invitation-only real-time bidding auctions where one or several publishers invite a select number of advertisers to buy their inventory.
- DSP: A demand-side platform (DSP) is a system that allows buyers of digital advertising inventory to manage multiple ad exchange and data exchange accounts through one interface. Real-time bidding for displaying online advertising takes place within the ad exchanges, and by utilizing a DSP, marketers can manage their bids for the banners and the pricing for the data that they are layering on to target their audiences (wikipedia).
- DMP: A data management platform (DMP) is a centralized computing system for collecting, integrating and managing large sets of structured and unstructured data from disparate sources. Personalized marketing enabled by DMPs, is sold to advertisers with the goal of having consumers receive relevant, timely, engaging, and personalized messaging and advertisements that resonate with their unique needs and wants (Wikipedia).
- Data Warehouse: The Data warehouse is a system that stores and preprocess data so it can be accessed more rapidly for reporting and data analysis. Due to the huge quantity of data most companies are facing and the need of real time reporting, Data Warehouses are considered a core component of business intelligence.
- ETL: The term comes from extract, transform and load and refers to processes in database usage and especially in data warehousing.
- Data extraction is where data is extracted from homogeneous or heterogeneous data sources.
- Data transformation where the data is transformed for storing in the proper format or structure for the purposes of querying and analysis
- Data loading where the data is loaded into the final target database, more specifically, an operational data store, data mart, or data warehouse.
- SSP: A supply-side platform is a technology platform to enable web publishers to manage their advertising space inventory, fill it with ads, and receive revenue (Wikipedia). Many of the larger web publishers of the world use a supply-side platform to automate and optimize the selling of their online media space.
- Statistical Computing solutions: After 2002, where the digital era began, the information increase at an outstanding rate, at a point that more than 90% of todays data was created just during the last two years.
That revolution of data can’t be assimilated without pain. You can’t analyze information just reading reports anymore, actually generating reports that drives you to make decisions with that volume of data is really complex. In order to have a proper understanding of the data now we have to use techniques that allow us finding the needle in the haystack.
With that in mind we need two things:
- Statistical techniques: The most used techniques are related to machine learning, both supervised, that it is mainly used for prediction and unsupervised that is used for clustering or classification proposes. There’s nothing new on this, most of the techniques are being used for decades, so we can’t say that the is something revolutionary on this.
- Technology: In order to run the above mentioned algorithms you need hardware and software. Whether you are hosting your own infrastructure or you are using cloud services, it’s key that you have software that allow generating the algorithms in the simpler way, at the lower cost in the required time. Kinda challenge right? The main used technology is R, which is an open source programming language and software environment for statistical computing and graphics used for developing statistical software and data analysis, mainly algorithms.
So basically the full environment would looks like this
At the beginning of the post we said that in every revolution there are winners and losers. We also talked about the importance of data integration. Actually we said that only the companies with all their data integrated in an homogenous way will rule the market.
If we talk about data integrated in an homogenous way there are two companies in that podium. Google and Facebook. Let’s see the degree of dependance we have with them:
Google’s control over the digital environment:
Let’s try with Facebook:
And even when I’m not considering it Facebook profiles and company’s fan pages as part of an AdNetwork we both know that that’s what it really is. Because what’s Facebook if not the dreamed Internet, I mean an Organized Internet.
So let’s finish this post with some interesting questions:
- Do you know who have all that information, integrated homogeneously and ready to use?
- Do you know who is today generating the algorithms that will drive you to make decisions?
- Do you know who is telling you where and how much do you have to invest (attribution model) your marketing budget?
Remember when we thought that internet will help decentralize the power and democratize the information?
So you are basically giving your vendor (and competitor) your money and your data (which can became a better asset than money if you use it properly) besides you are allowing them to see what works and not so they can use it to improve their business. Isn’t it grisly and bloodcurdling?