All posts by BIAnalytix Team

AI, Great Whites and Powerball

The human brain simply doesn’t have the capacity for making quadrillions of campaign decisions.

In our journey to develop the first cross-media planner that optimizes inventory and audiences across both proposals and in-flight campaigns, one of our learnings was that you simply cannot do this without machine intelligence. And without optimization supported by machine intelligence, it is impossible to maximize the revenue potential of your media business.

Why?

Consider optimizing five campaigns, each spanning four weeks and making use of four inventory pools. That’s 80 possibilities, right? Well, in the current way of thinking, that number may be in the right ballpark. Unfortunately, the current way of working pushes more effort onto the folks who steward the delivery of those campaigns. And, that’s where things start to get tricky.

You see, current work practices dictate that stewardship is just a fact of life. The emphasis is on taking business at any cost. The problem gets pushed down to traffic and delivery systems, et voila! We waste our inventory to make up deficiencies, and we churn our inventory and customer expectations. We wrack our brains to make stuff fit, and we usually end up doing this at the last, most stressful minute.

You see, the math actually is 80 choose 16 — i.e., how many combinations of 16 can I choose from those 80 possibilities to derive an optimum answer? Do the math. You’ll see 550 quadrillion or 550,000,000,000,000,000 possibilities that need to be reduced to a set of possible optimum answers. Certainly well beyond the capabilities of the human mind!

Now you see why we’ve used our own artificial intelligence stack to crack this problem. There really is no alternative if you want to do it right.

Interestingly, one of the keys to this solution lies in understanding the fundamental flaw in all of the traffic and sales systems and agency buying systems in the marketplace today. They simply focus on workflow, a workflow based upon priority placements. Their plotting/placement engines focus on getting the best overall delivery per order. Not a single one of them considers the destruction they wreak to other campaigns in the process, or the revenue potential they sacrifice to the business as a whole in doing so. So, they stuff placements in the schedule based upon some priority, push out others, and kick the debt down the road.

How did we ever allow our industry to accept this as best practice? Instead, we should optimize all orders, evaluate all proposals, and optimize all inventory pools for your entire sales ecosystem in real time. A total understanding of media economics is at the heart of this process, with a machine learning about your business in order to help ensure contracts are fulfilled.

Oh, and by the way, we did some math for fun; it’s just what we do! Did you know that your chances of being bitten by a great white shark and winning Powerball in the same day are — you almost guessed it — it’s actually about 110 quadrillion to one. That means your odds of finding one of those optimum answers are five times worse than of hitting it big in the lottery!

A Solution for Cross-Media Sales Challenges

The most serious question facing media organizations in the cross-platform world is how to value inventory, optimize the use of inventory, and negotiate advertising campaigns utilizing that inventory.

The solution? Use intelligent agents and data integration to surround existing infrastructure — aggregated operational systems — and elevate the decision process above that infrastructure.

Rather than rely on supplemental manual workflows to achieve sufficient integration to execute business processes and analysis, add an intelligent layer around operational systems. This approach enables an elegant and immediate solution for unifying people and processes, for overcoming fractured workflows, and for maximizing revenue opportunities.

BIAnalytix delivers all of these capabilities and more. The Decentrix media analytics platform allows the sales organization to organize itself to sell cross-platform or to specialize in a delivery vertical. Instead of trying to build pricing bundles across incompatible delivery vehicles, the organization’s pricing and planning team can develop ad products through research and policy-making. Clients can negotiate cross-platform campaigns more quickly with higher execution fidelity. Inventory is leveraged more fully and at a higher value.

How does BIAnalytix do it? The platform accurately tracks all available inventory for sale across all audience targets and distribution channels, and it prices that inventory based upon demand. Using data from the many operational systems above which it sits, the platform forecasts capacities and sellouts, and then reconciles that information with delivery across all delivery vehicles and inventory pools. As a result, the media organization can easily construct cross-platform campaigns with optimum inventory usage and audience fidelity. Intuitive visual interfaces simplify evaluation of proposed campaigns for maximum revenue gain with inventory usage efficiency.

To ensure accurate billing and revenue projection, the platform not only forecasts alignment of audience measurement but also tracks the campaigns in flight and makes adjustments so that contractual obligations are met automatically. All of this occurs in real time.

By removing sales friction, preventing revenue leakage, enhancing the value of ad inventory, and maximizing the yield of campaigns without manual stewardship overhead, BIAnalytix allows media companies to unlock the hidden potential of their media inventories.

Using AI to Maximize Ad Sales Revenues

Big Data, the Internet of Things (IoT), and artificial intelligence (AI) technologies — these are the new drivers of the modern media business. As the industry adapts to new and emerging content delivery technologies, data and analytics are the tools essential to extracting actionable intelligence and presenting it to the right people at the right time. AI and machine learning are taking analytics to a new level, improving and enriching the quality of data and, in turn, offering even more specific forecasts and more accurate insights into potential revenue opportunities.

National and local broadcast groups traditionally have considered monitoring of key performance indicators (KPIs) to be sufficient for performance analysis. They have relied heavily on one such KPI, revenue performance, which compares the networks or stations actual and projected revenue trend against their budgets. While executives can use this KPI to understand if they are on track to meet their financial commitments, they need more and better information if they are to understand how to improve revenues and address the factors contributing to budget shortfalls.

Only AI can effectively examine the tens of millions of data points that reflect sales performance, market trends, and business cycles across many different categories, such as lines of business (local, national, digital, etc.), products, sales offices, and even individual advertisers. In addition to monitoring sales operations and surfacing new alerts, opportunities, and risks to relevant stakeholders, AI learns as market conditions change and as the business grows and evolves. Using machine learning (ML), AI-driven analytics systems use algorithms to parse and learn from data, and then use that understanding to make forecasts or predictions.Product Data Point Analysis

AI learns what “normal” business cycles look like, then continuously monitors for new risks and opportunities.

While many industries are applying AI and ML to the massive collections of data associated with their operational systems, the media industry is just beginning to leverage these powerful tools to support business strategy and success. In fact, Decentrix is the first media technology company to leverage sophisticated media-centric AI and ML to enhance the revenue opportunities of media, entertainment, telecommunications, and advertising companies in the cross-media marketplace. Our BIAnalytix platform uses AI and ML to expose critical data within operational systems and to deliver insights that yield maximized inventory pricing, enhanced audience values, and optimized campaigns across all properties and platforms, across linear and digital business models, and across OTT and ATSC 3.0.

Both national and local sales enterprises may be able to produce some of these insights without the benefit of AI or ML, but the sheer size and scope of data prevents them from doing so quickly or comprehensively. Implementation of media-centric, AI-driven analytics allows a media group to automate and accelerate analysis of all data and to extract the insights essential to capitalizing on opportunity in a complex and competitive marketplace.

Top 5 Reasons You Need to Know About BIAnalytix

REASON 1: BIAnalytix is a comprehensive suite of analytics tools for media enterprises — and the industry’s only solution to fully address today’s cross-media marketplace.

REASON 2: BIAnalytix is at work for the world’s biggest media, telecom, and satellite companies to maximize revenue across their biggest lines, but its modular toolset can help enterprises of any size realize a revenue boost.

REASON 3: BIAnalytix is designed and built by industry engineers and data scientists specializing in the media advertising sectors.

REASON 4: BIAnalytix is uniquely capable in aggregating data from across current and legacy transactional systems, in using AI and machine learning to quickly surface critical business insights, and in reporting that in-depth information in a manner that simplifies strategic decision-making.

REASON 5: BIAnalytix is a proven tool for maximizing the value of inventory, driving growth and ensuring a rapid ROI for any level of deployment.

Five Predictions for Media Technology in 2018

1. AI emerges from the hype
From CES in 2017 and throughout the year, we have been bombarded with everything from cars, beds, toothbrushes, and toasters endowed with artificial intelligence, or AI. Some have risen to the occasion, others have struggled to find relevance, but all have benefited from the associated marketing cachet. In 2018 we will see the direct application of this class of technologies to media-technology. AI will assist media corporations in developing better content and allow consumers to have that content targeted to them more judiciously.

2. Data hoarding is no longer cool
It costs money to store data. While many organizations still talk about the need to process big data, those media companies that haven’t worked it out that “bigger is not better — better is better” will be significantly disadvantaged. What is better? Is data better because it is relevant, better because it is fresh, or better because one has lots of it? Or none of the above? Data is only as good as its utility. It takes real alchemists to turn lead into gold. Ask yourself this: If your business cannot make money out of your data, then why collect it?

3. Machine learning (ML) is about the right teacher
Knowing what data creates a viable economic outcome is central to the inevitable value of AI in media-technology. Although demeaning, as George Bernard Shaw wrote, “Those who can, do; those who can’t, teach.” And, so it is with ML. The ML promise has implied that plenty of data with the right algorithm will deliver riches to your media organization. In 2018 it will. But, not because of ML alone; it will be a result of those who can teach the ML because they know the business, and can do. In this respect, Decentrix leads the pack in media analytics and ML.

4. Trust becomes a force
We all intuitively know that success is built upon a foundation of trust. The events of the past 12 months have eroded that foundation through the proliferation of “facts” with dubious veracity mixed with opinion, memes, and emotion. The year ahead will be especially trying for media companies as product and service quality must rise above the increased noise of confusion and negative publicity. The only way to break through that noise is to focus on delivery commitments and customer promises. Be the company that always says what it does, and does what it says — to shareholders and stakeholders, to customers and employees alike. Competitive marketing will simply become noise.

5. Subscriptions become ubiquitous
Some may not realize that the subscription model was pioneered by print media. Now it is used by many businesses and websites. When electricity was first implemented in New York City, each building had its own generator and was dependent upon the Edison Company for everything from light bulbs to wiring. Now power is a utility, and consumers subscribe to the service on a consumption basis. It is becoming widely understood that companies adopting subscription models are growing their revenue significantly faster than those with traditional business models. Subscriptions drive media options and other consumer services, and so it will increasingly become critical for enterprise solutions to implement subscription-based services as part of their business models.