Understand your workflows. Before they transform, media companies need to identify workflows and sales channels that would benefit the most from increased automation, optimization, and content creation capabilities. For broadcasters and cable programmers, for example, automating direct sold traditional TV campaigns—with their manual processes, ad-hoc analyses, and significant revenue contribution—is ripe for AI and automation. (See Exhibit 1.)

Generally, most media companies will want to focus on teams supporting the direct sale and delivery of campaigns rather than support functions such as HR, legal, and customer. The teams supporting sales, planning, and operations generally are the largest in ad sales and so are promising places to seek efficiencies through generative AI.

Identify pain points. AI excels at solving frictions and process bottlenecks. Exhibit 2 shows four of the main common pain points:

• Proposal Development. In most media organizations, responses to RFPs are manually generated by analyzing proprietary data and market research to address agency questions. Instead, generative AI models, customized with proprietary and licensed third-party data, can analyze historical data, and identify client KPIs to incorporate into materials.
• Plan Development. Most plans are manually created based on forecasts, commitments, and other inputs. It is difficult to ensure yield is optimized across plans and the broader portfolio. Instead, AI and machine learning tools can optimize plans by analyzing buying based on past business cycles and running scenarios that take into account proposal requirements, available inventory, and previous commitments.
• Insights Development. As with proposal development, insights are often manually generated by aggregating and analyzing a common set of data. It can take days to produce a single campaign for one client. Instead, customized generative AI models can quickly create reports, allowing executives to spend more time thinking about the strategic implications.
• Creative Review and Testing. Media organizations need to review, test, and approve many forms of creative outputs delivered to many distribution outlets, a process often managed by third parties. Instead, AI tools can create assets and execute creative reviews to decrease human interaction.

Overcome resistance. Teams can sometimes resist the adoption of AI on the grounds that it cannot execute a sales team’s activities. That belief is only partly true. AI may not yet help with strategic tasks such as developing the pricing strategy but is fully capable of taking on more tactical tasks such as demand forecasting and calculating premiums and discounts.

Media companies can overcome employee resistance by decomposing work into tactical, operational, and strategic activities. This approach also helps diffuse potential pushback by recognizing some use cases and activities are not good candidates for AI.

Establish priorities. All use cases are not created equal. Media companies should start with those that are feasible in the near-term and provide high benefit in the long term.

When thinking about feasibility consider:

• Does the organization have access to the necessary AI capabilities?
• Can the use case be launched in a reasonable period?
• Is the necessary data available?
• Is the organization ready to adopt the new approach?

Likewise, there are many ways to evaluate the benefit, or potential, of a use case:

• What is the potential revenue and contribution margin?
• How will the business process become more efficient?
• How will direct costs be affected? How quickly?
• Will the initiative improve the brand or customer experience or both?

In our experience, plan automation and creative review and testing are likely to yield the most benefit across the four dimensions of potential. But media companies should analyze their own operations to be sure. (See Exhibit 3.)