Incorporating generative AI into your business can be a massive unlock when it comes to increasing productivity and ultimately your bottom line. At the same time, as companies rush to embrace this technology, the gap between hype and real business value also grows wider. While many organizations dive straight into implementing their best ideas, experience shows that a structured discovery process is crucial for identifying truly transformative opportunities. But what happens after discovery, and how do we turn these insights into measurable business impact?
What we mean when we say discovery
Traditional ways of implementing generative AI solutions often start with individual stakeholders championing specific use cases based on their personal perspectives. This tends to lead to siloed solutions that may improve one team's workflow while potentially creating bottlenecks elsewhere. Our discovery process here at Reaktor takes a different approach.
Discovery is the critical phase before building new solutions that allows organizations to gain important insight into their internal and external operations, and how certain workflows might be hindering progress.
When we engage with and learn from multiple stakeholders across the organization, the discovery process can reveal interconnected challenges and opportunities that might otherwise stay hidden. This holistic point of view makes sure that any improvement made in one area doesn’t create unintended consequences in another. As an added bonus, early stakeholder involvement creates necessary buy-in for eventual implementation or changes in ways of working.
Generative AI brings an extra source of power into discovery. Unlike with many other technologies, generative AI enables the building of quick demos to demonstrate value, often in hours instead of days or weeks. In one recent project, our team was able to create and test potential solutions in real time while the brainstorming session was happening. This rapid feedback loop allows teams to quickly learn what works and what doesn't.
Common issues the discovery process unveils
One of the most revealing aspects of discovery work is how it shines a light on certain organizational dynamics and workflows that have evolved over time. Through our client engagements, we often observe how different teams and departments operate in ways that aren’t optimally aligned to serve customer needs. For example, let’s say a customer requests support or service. Their request often travels through multiple teams, each working in relative isolation before an answer is found and delivered. This might result in delays or even the request falling through the cracks altogether.
Another common example is how technology systems, originally implemented to improve efficiency, can actually constrain how people work. In many organizations, we find that tools like CRM systems have turned from enablers to rigid frameworks that dictate processes and workflows. Rather than starting with the ideal way to solve a problem and then finding technology to support that approach, teams find themselves adapting their work methods to fit the limitations of their tools. It's a classic case of the tail wagging the dog: humans serving machines rather than machines serving humans.
Discovery work can also reveal how long-standing organizational habits can turn into routines that no longer serve their original purpose. We often encounter teams who have been performing the same processes for so long that they've lost sight of the actual value they're meant to create. A common example is regular reporting that continues simply because "it's always been done that way," even though the reports no longer inform any meaningful decisions or actions. These legacy practices consume valuable time and resources without contributing to better customer outcomes.
The siloed nature of large organizations can further compound these challenges. Teams often develop a kind of tunnel vision. They become highly proficient within their own domain but lose sight of how their actions impact other parts of the organization. They may be unaware of how their processes could help (or inadvertently hinder) another teams' success. For instance, one department might optimize their workflow in a way that creates downstream complications for others, simply because they lack visibility into the broader organizational ecosystem. This disconnection between teams not only affects internal efficiency but ultimately impacts the quality and speed of customer service delivery.
Examples from the real world
Concretely speaking: what kind of initiatives can you expect to emerge out of discovery? Well, basically anything under the sun, as every organization is unique. But the key is to keep an open mind: expecting specific results or findings tends to lead to confirmation bias, while the point of discovery is to find what you might not know you are even looking for. Here are a few real-world examples of what we’ve seen out in the field
1. AI-assisted tool to help data transfers
One manufacturing company's discovery process revealed that their specialists were spending a lot of effort in manually transferring data between their various systems (CRMs, ERPs, and other key tools). It takes significant resources to build out integrations, and therefore, integrations have not been a priority in the past. But understanding just how much inefficiency these workflows were creating allowed the company to realize its impact and, therefore make an educated decision to act. With AI, solving problems like these has become much easier than in the past, and this discovery process allowed us to define just what kind of AI-assisted tool could be built to help facilitate smooth transfers of data between various systems.
2. AI-powered case summarization tool
In another case, a client with a large and complex organization found that a request coming from a customer was often being passed through multiple teams and people, leaving too much room for error and resulting in a slow and fragmented process. As a response, we were able to identify just the right technical solution to support a faster and more efficient workflow. In this case with an AI-powered case summarization tool that enabled faster handoffs between teams while making sure that critical details were not being lost in transit. This not only reduced response times but also improved the quality of service being delivered.
3. Genereative AI solution to streamline pricing
A company within the industrial domain discovered a critical inefficiency in their product catalog management during our discovery process. With large numbers of suppliers using different naming conventions for identical products, their team struggled to maintain accurate pricing. Every supplier price update required manual cross-referencing of extensive spreadsheets to identify affected products in their own catalog and calculate new pricing – a process that took days and risked costly errors.
The discovery process revealed an opportunity to build a generative AI solution that could intelligently match products across different naming conventions and automate pricing updates. By understanding product descriptions and technical specifications, the system could identify identical products listed under different names. This transformed an incredibly manual process into an automated workflow, significantly reducing errors and ensuring more timely pricing updates across their product portfolio.
Looking to the future
Going from discovery to implementation in a world of generative AI (and beyond) isn’t just about technology. We’re talking about strategic transformations in how organizations identify and solve their most pressing challenges. Through discovery, and regardless of which technology we’re talking about, companies can move beyond surface-level automation to find opportunities that are uniquely mission-critical for them.
And as generative AI capabilities continue to evolve, the insights gained through this process become even more valuable. Organizations that maintain a clear understanding of their challenges and opportunities will be best positioned to take advantage of these advancements and solve more problems faster. The key to success isn’t implementing the most advanced tech – it’s finding the right problems to solve and measuring impact holistically.
