More Data, More Problems. Simplify by Adding AI

The appeal of data collaboration has always been its potential to bring simplicity to the traditionally fraught process of compiling a comprehensive picture of customer behavior and preferences from various sources. This, in turn, was supposed to make it easier for media buyers and sellers to derive meaningful insights and drive informed decision-making for specific outcomes. Yet the road from potential to reality has, thus far, been encumbered with endless back-and-forth between data intermediaries, a tangle of antiquated systems and successive compliance hoops for partners to jump through. It’s all because data collaboration is typically sold as a technology solution—but what gets delivered is consulting, with all its human complications. The hurdles associated with data collaboration are particularly high in ad tech. About two decades ago, ad networks were supposed to revolutionize and simplify the buying and selling of advertising. But as programmatic campaigns became more commonplace, the costs involved in running them grew exponentially. Why? In a nutshell, the programmatic revolution devolved into bureaucracy; manual processes associated with campaigns only proliferated with time, which led to middlemen interrupting the exchange of data between buyers and sellers. As a result, these new “solutions” brought huge margins to ad-tech platforms—but not a great deal of efficiency to anyone else. How Amazon Is Making Ad Tech the MVP of Thursday Night Football The standard procedure for data collaboration follows this path: First, you recognize that your company needs data. Then you sift through data suppliers. Once the field is narrowed, you have to acquire the data. After that, forecasting determines the scale of need, followed by an evaluation of the data quality. By this point, a month or two has likely passed. After the data quality assessment phase, security and privacy issues must be tackled. Interested parties will fill out a detailed questionnaire to ensure the proper governance rules are in place for GDPR, CCPA, Rule 10b5-1 and more. Then come the lawyers preparing legal papers to be signed. Add another month before moving on to pricing agreements and further quality assurance integrations. After six months, the process still isn’t in full swing: The management process lasts about half a year, involving reports on transactions and data usage, chasing down payment (and reconciliation), then reviewing and updating the entire strategy. Given how much is at stake in managing and deploying vast data sets, these steps are not arbitrary. But in 2023, a process that can sometimes stretch out to eight months should take no longer than eight hours to complete. AI’s answer: a single system of record Expanded access to artificial intelligence-driven software has facilitated the ideal of frictionless connections between data partners, but these players often find themselves grappling with a paradox: The very approach that promises effectiveness and insight often leads to complexity and frustration. Right now, the mainstreaming of generative AI, particularly large language models like OpenAI’s ChatGPT, is paving a smoother path toward efficiency in data collection. By eliminating the need for multiple intermediaries and providing a unified platform, the entire process is simplified right from the start. Just as real-time bidding eliminated the need for phone calls and faxes between ad buyers and sellers, a similar set of software tools can simplify data collaboration. To envision this possibility in action, it helps to look beyond the narrow lens of ad sales and analytics. Consider the way legacy enterprise software sales gave way to the software as a service (SaaS) model. Much like programmatic, procurement was a convoluted, lengthy process for buyers and sellers. SaaS methods did away with the multimonth rollouts of software and million-dollar contracts because their “try before you buy” feature allowed for contracts to be canceled with a month's notice without the need for prolonged unraveling and winding down. Using AI, centralized data collaboration platforms should be able to mirror the SaaS model, revolutionizing the way businesses handle data and offering flexibility, ease and adaptability. Unlocking new possibilities One corollary to the benefits of time and cost savings afforded by a unified data collaboration system is enhanced creativity. When a process takes several months, experimentation becomes impossible. But if partners can streamline and resolve data sharing in several hours, the time saved can be spent exploring new strategies and responding to challenges swiftly. When data collaboration becomes seamless, your team is no longer limited by the inefficiencies of the past. The process is no longer a challenge, but rather a competitive advantage—one that’s available right now. You just have to demand it of your organization and your partners. This story is part of the Power of Tech Marketing special feature.