The traditional annual strategy plan is already obsolete, according to some industry experts. In the AI era, it is argued, where new capabilities and competitors pop up daily, companies that plan once a year risk locking themselves into strategies that are outdated within months.
As quarter one ends, many executives are already reassessing plans that looked solid just weeks ago. For example, web traffic data suggests Google searches for ‘Update business strategy’ soared +7,583% while searches for ‘Agile strategic planning’ jumped +415%, according to the Google Trends analysis by Gradient Labs.
Neal Lathia, Co-Founder and CTO of Gradient Labs tells Digital Journal about the optimal strategy principles for business strategy planning in this era of artificial intelligence.
Build what’s needed, not trending
“Does the world actually need your product, and does it solve the problem it claims to, or is it an opportunistic attempt to capitalise on hype? If so, what’s your plan when innovation inevitably moves on, and your customers do too?” advises Lathia, who adds:
“Rather than focusing on a trend, we concentrated on building domain expertise and offering genuine value. AI powers our agents, but the real intelligence comes from our team of human experts, who possess the know-how to adapt to the rapid pace of development in our space.”
Don’t miss your launch window
According to Lathia: “When innovation moves fast, execution is everything. Making unrealistic plans will result in constantly missing deadlines and delaying projects because they’re not quite ready. That doesn’t work in a landscape where technology can become obsolete overnight. With new and improved models launching every month, a failed launch isn’t a short delay. It means you’ve missed a window of opportunity that will likely never come around again”.
Make space for experimentation and innovation
“Every employee has their own public space – a “brain channel” – where they can post interesting articles, knowledge they’ve picked up, or experiments they’ve conducted. Colleagues can follow each other’s channels and leave comments, encouraging everyone to share, discuss, and develop ideas together. Whether it’s industry shifts or better ways of working, posts don’t have to relate directly to someone’s role or project. Not every post will lead to a breakthrough or have a huge impact, but every discussion helps to sharpen critical thinking,” Lathia shares.
“We also have a ‘meeting-free calendar’ culture. We found that standing meetings quickly become formulaic and achieve very little, which wastes time and reduces focus. Our culture, instead, gives us all more time to focus on more important tasks. With our schedules clear, meetings can happen on demand when threads become too long, a system isn’t working, or multiple people want to get on the same page. In other words, when they’re actually necessary,” Lathia adds.
Start each project with a new team
Lathia next words of advice are: “Every project is unique, requiring different knowledge and skills to build and ship fast. That’s why we don’t run the company with a set of permanent ‘function’ teams. Each project starts by assembling the perfect mix of people and ends by disbanding the team, regardless of success. It keeps us nimble, avoids the collective team slowdown that often follows a completed project, and ensures the best person for the job is almost the person actually doing it”.
Lathia continues: “High-priority projects are handled by what we call ‘strike teams’. These are cross-functional groups that bring together the best talent to achieve a particular goal. As each member completes their task, they drop out to focus their attention elsewhere. Teams of two handle other tasks – usually discipline-specific, such as an AI issue – while some are single-person quests. These are usually big ideas or customer requests that must fit around more pressing priorities, with one person handling the entire process from planning to building and delivery”.
Don’t hide behind the data
Data is only good if it is meaningful and acted upon. Lathia recommends: “The truth is, it’s becoming easier to build an AI agent. Building a reliable one, however, isn’t. Numerous AI agents might reach the same resolution. What separates them is the difficulty it, and the user, faced. Some experiences might be okay, some frustrating, and some amazing, even though the same resolution was reached every time”.
