Build or buy? The AI battle defining the landscape of media innovation

"The future is already here—it's just not evenly distributed."

William Gibson’s quote perfectly captures the current state of the media industry. We stand at the precipice of a technological revolution, where AI is no longer a futuristic fantasy, but a tangible force reshaping every aspect of our world. The race is on, not just to develop AI solutions, but to develop the right AI solutions.

The media landscape is awash with ambitious promises linked to technology convergence (combination of but not limited to generative and specialised AI, machine learning, deep learning and robotic process automation) but success won't be measured by sheer technological prowess alone. It will hinge on the ability to translate raw innovation into operational reality, bridging the gap between operational insight, a commitment to on-going long-term investment, a clear view on what corporate transformation looks like and smart people to ensure real-world application, delivers real-world value back to companies and their people.

This is where the true battle for dominance will be fought, and the winners will be those who can harness the power of AI, not just to create unparalleled efficiency, but more importantly improved product output and time back to people, elevating people from a job to a rewarding career.  

This presents a crucial question for companies: should they focus on internal innovation, building a team of subject matter experts charged with the development of state-of-the-art proprietary AI tools, or outsource, partnering with established technology-giants like OpenAI, Meta, or Google, in consultation with the globally networked consultancies such as Deloitte, Accenture, Cognizant, and EY?

Before answering this question, we need to break out pros and cons for each.   

The pros of building in-house AI arsenal

• Proprietary advantage: Developing in-house AI solutions grants a company exclusive control over the technology, providing improved development cost efficiency and compressed development timelines. It also prevents companies from ultimately funding market development, as no matter how tightly controlled an agreement is with a third party, outsourcing Research & Development (R&D) is driving future third-party innovation. 
   
• Data security and privacy: Keeping data within the company’s infrastructure ensures greater control over its usage and security, this is particularly important with on-going concerns about data privacy, security, sharing and storage. It also reduces development timelines, with no need to work through third-party contractual and non-disclosure agreements and out-sourced data storage solutions.
   
• Long-term investment: Committing to building out an internal AI lead team, tasked with Technology Convergence¹ solution development, fosters a culture of innovation, allowing for long term big picture strategy implementation.  This increases sustainable growth opportunities via competitive advantages from improved product and resource IP value.

The cons of building in-house AI arsenal

• High initial investment: Developing AI solutions from scratch requires significant upfront financial investment in talent, infrastructure, and ongoing maintenance. Also, as mentioned earlier, due to the global Technology Convergence¹ arms race, securing highly skilled resources in line with a company’s P&L limitations is likely to be challenging. This is particularly true when the length of time before a financial return to the upfront investment is considered.  
   
• Time-consuming: Coding and training AI models can be time-consuming, potentially elongating solution implementation. Also, it is not just the AI team that will be working on the projects. ‘As is² v To Be²’ processes will involve a wider resource base across the company, that will take people away from client business-as-usual (BAU) activities. This needs to be balanced so client stable is uninterrupted.    
   
• Limited scalability: Arguably the most important, internal solutions might not be quickly scalable versus those offered by large technology companies, potentially hindering long term growth and the ability to obtain a competitive advantage ahead of natural market evolution.

The pros of leveraging the power of tech giants

• Faster time-to-market: Partnering with established technology companies in conjunction with a consultancy with vast, historically proven R&D accesses and pre-built off-the-shelf AI solutions, can be quickly adapted for faster, broader application and better results.
   
• Scalability and resources: Globally networked technology companies have vast resources and infrastructure, enabling them to scale AI solutions quickly to meet growing demands. This means as a company moves towards project maintenance phase,  additional opportunities can be identified and developed at speed. 
   
• Cost effective, cutting-edge technology: Partnering with specialised technology companies, that are globally networked, grants access to the latest AI advancements and cutting-edge technologies. Additionally, for large, network projects that require cross-market, standardised adoption and implementation, outsourcing can be more cost-effective than building internal solutions, especially for smaller agencies or companies.

The cons of leveraging the power of tech giants

• Dependence on third parties: Arguably the most important con is that by outsourcing the company’s Technology Convergence strategy there is greater reliance on external solutions which can lead to dependency and potential limitations in control over the technology and future development strategies.
   
• Limited customisation: Pre-built solutions may not perfectly align with a company’s specific needs, requiring costly adjustments or workarounds that may hinder a solutions long-term scalability. Additionally, using widely available AI solutions may not offer a unique competitive advantage. To obtain a bespoke solution, a company needs to enter into a detailed contractual agreement based on a well-defined scope of work. This can add complexity and elongate timelines.
   
• Data security concerns: This is a significant consideration, as sharing data with external platforms raises concerns about data privacy and security. Also, companies may have restrictions in place where data can be stored and where development and run and support resources are located. These considerations need to be worked through in advance.

The optimal approach: A hybrid strategy

"The whole is greater than the sum of its parts."—Aristotle

The ideal approach lies in a balance. Aristotle’s quote highlights the idea that individual parts, whether they are internal teams, external partners, or specific technologies, contribute to a greater whole. By combining these elements in a balanced way, something more powerful and effective than any single component could achieve on its own.

This approach allows a company and its selected partners to come to the table and provide highly specialised expertise. It also provides a company with flexibility. By setting up its own AI solution-based division, a company can develop at speed, smaller agile solutions that provide individual market focus, while broader network-based solutions that require substantial standardisation can be driven through an outsource technology partnership.

The key to success lies in understanding the unique strengths of your agency—and knowing when to let a bigger player help carry the load. In a world where the frenetic pace of innovation is relentless, the future belongs to those who can seamlessly blend both approaches and turn technological advances into a sustainable competitive edge.

Geoff Clarke is the COO of IPG Mediabrands Australia.