Navigating Gen AI's middle miles

How software and platforms can break through the gen AI middle miles

Marathon runners spend months training. When race day arrives, it is filled with excitement, cheering crowds and nerves. Runners are often warned to hold that excitement at the beginning of the race, so they do not come out too strong. Inevitably, after the start of the race, runners hit the “middle mile” – legs are heavy, and the finish line is not close enough for motivation. It’s during this time runners need to dig deep, rely on their training, stay mentally focused and simply put one foot in front of the other.

The software and platforms industry is facing the equivalent of a “middle mile” moment in the adoption of generative AI technologies. In fact, our “Reinventing with a digital core” research revealed that only 34% of software and platform companies have embedded AI into all business processes.

The media hype cycle has positioned gen AI as a driver of the next wave of industry growth and a key lever to increase profitability. However, similar to running a marathon, the next stage of gen AI will be characterized by the significant effort required to instantiate the rapidly evolving technology into the business and product.

Even for the most seasoned runners, there comes a time during a marathon where fatigue overwhelms excitement. Similarly, even the most sophisticated software and platform leaders, with top engineering talent, and the best large language models will face challenges integrating and driving adoption of gen AI. These challenges will manifest in several ways, including rapid model evolution, messy underlying data, internal silos, and stakeholder misalignment. These represent some but certainly not all potential barriers. Industry leaders that remain agile, place bets that provide future optionality and address internal blockers head on, will exit the middle miles with a greater trajectory.

Shifting the mindset from how to now

The theoretical understanding of gen AI is now widespread, but practical applications and integration of the technology is where the challenge lies.

Our generative AI for customer growth research found that companies that apply gen AI to customer-related initiatives can increase revenue by up to 25% in five years. Additionally, increasing user expectations and winning the battle for attention accentuates the opportunity to use gen AI across marketing, sales and service functions.

Software companies have been ahead of the curve when it comes to integrating gen AI it into customer-facing products. However, unlocking the full promise of gen AI requires more than technology investments. It needs a holistic approach where AI is treated as the foundation of the reinvention required to be an industry leader.

We see three imperatives for generative AI to go beyond incremental optimizations and drive growth at scale for software and platform leaders.

First and foremost, have the courage to take a fresh look at your business architecture. Years of hypergrowth, rapid innovation, regulatory changes and mergers and acquisitions have resulted in many disparate systems and silos. Tech companies that commit to proper aggregation of data as well as a secure and have an AI-led cloud strategy will have the building blocks needed to realize the true potential in this next chapter of gen AI. While it’s never possible to get everything right the first time, protecting your origination is free of key barriers that inhibit gen AI-driven growth should be prioritized this year.

Second, be flexible. There are currently many different large language models (LLMs) built for different applications. As we saw with adoption of hybrid or multi-cloud environments, it seems increasingly likely that GenAI will be a multi-model story for most companies. This will enable future optionality without compromising the pace of innovation. We are seeing the emergence of a host of GenAI use cases that are uniquely done by some models and not others. In parallel, this multi-model environment is creating new, breaking down industry competitive barriers, enabling first time partnerships among leaders.

Third is about people: GenAI is no doubt going to disrupt the workforce. Our modeling shows that nearly 44% of working hours in the US are in scope for automation or augmentation. This is not bad news: moderate performers will be able to become super-stars. Super stars will be unlocked to drive growth in new ways. Ensuring this positive outcome will require investing in the skills, capabilities and AI literacy required to evolve as quickly as GenAI.

We are dog-fooding this at Accenture and initial results our promising. Our Marketing and Communications organization has put gen AI at the heart of its reinvention strategy. We are using AI Navigator, our proprietary AI assessment tool to help with planning and showing where generative AI can drive the most value.

Our investment in Writer, a gen AI assistant tool, plays a key role in powering our content supply chain, improving audience engagement through personalization while speeding production and operational efficiency. We also built a gen AI—proposal developer that helps focus more deeply on solutioning our assets for clients rather than focusing on mundane tasks such as formatting. We’ve also built and killed a number of tools that did not pan out well. The efforts are amplifying our people’s capabilities and enabling them to focus on the work that matters most.

There’s no doubt that generative AI is transformative. But pushing through the middle miles to the top will take focus and commitment. Success with gen AI is only possible with sound architecture, flexibility and a willingness to place bets that provide future optionality. Now is the time to dig deep. Doing so will enable the software and platforms industry to unlock productivity gains, drive profitability and stay ahead of the competition.