The hype around the rise of generative AI technologies makes huge promises about the potential of the technology. Yet it would be fair to say the majority of organisations are only experimenting with the technology or using it in isolated use cases.
The problem with such a rapidly maturing and expanding field of technology is the number of established vendors and enterprise worthy products are low. In addition, proven adoption frameworks and journeys are rare, leaving a lot of uncertainty for new adopters.
This article explores some of the things organisations should do to ensure success when implementing generative AI at scale.
1. Crafting a Generative AI strategy
Generative AI, typified by advanced models like ChatGPT, is rapidly evolving and comprised of many different categories of technologies. These technologies have potential to generate high-quality content across many use cases (emails, blogs, product descriptions…), improve customer service via automation, boost sales leads via CRM and sales integrations, or facilitate market research.
However, to leverage these benefits on more than an individual level, businesses need a generative AI strategy that addresses how to drive adoption and effective use of the selected technologies.
Setting the framework for Generative AI
A robust Generative AI strategy is comprised of:
- Understanding the goals that drive the organisation
- The performance measures of employees within the organisation
- Identifying use cases to enable the organisations goals and improve the performance metrics of employees
- Identifying the right technologies to address the use cases
- Preparing a roll-out program with tech support, change and communications
2. Building a convincing business case
Generative AI involves significant investment of time, resource, and capital. Hence, it's essential to build a compelling business case that highlights the potential benefits and returns on investment. A well-structured business case will secure buy-in from the leadership team to get a program up and running.
Quantifying the benefits of Generative AI
Generative AI is focused on creative tasks most frequently completed by domains such as product design, customer service, sales and marketing. When building your business case, identify specific use cases where generative AI can deliver substantial value. A simple way to do this is to identify the relevant processes suitable for automation, the steps within the process which can be automated and quantify the value of the possible automation.
For example, sales staff may spend a significant portion of their time generating leads via email. Generative AI can be used to generate emails, email campaigns, and can be integrated into CRM and sales tools to distribute the emails.
3. Choosing the right technology
The choice of technology is crucial in the adoption of generative AI. Businesses should consider fundamental factors like ease of integration, scalability, security, and compliance with regulations. However, specific to generative AI there are also extra concerns such as the technology's ability to handle enterprise-specific data securely, or a decision between a prebuilt product such as Jasper, a solution native to a business application or a completely bespoke solution.
There’s a valid assumption that the use of raw ChatGPT is not a good solution, both on the individual level and to scale across an organisation.
Better tools will be tuned to the specific use cases that you’d like to apply generative AI to. However, there are many on the market, to understand a few examples:
- Microsoft is developing copilot tools that integrate with their business applications, Outlook and more to perform common tasks such as writing a sales email or even predicting disruptions in your supply chain. These copilots will be available and integrated into the tools employees use every day and some of them may not require any separate licensing.
- Jasper, Writesonic and other generative AI assistants offer a multitude of templates to use generative AI to write an email, research and write a blog or many other common tasks employees complete. These tools require minimal technical foundation and come at a relatively low cost.
- Bespoke applications of generative AI can be produced for an organisation for purposes such as AI-augmented chatbots, querying massive knowledge bases and more. For example, using Azure OpenAI you can develop a solution built specifically for your business accelerated by OpenAI’s models. These tools can be highly tuned to the requirements of your organisation but are intensive in terms of the investment of time and money required to build and maintain them.
Where possible we lean towards the first two options to begin a generative AI journey, with the third being best suited for large, technologically proficient and AI experienced organisations.
Picking a tool which can address many different use cases with minimal customization allows for large-scale programs to be launched with minimal complexity from the technology.
4. Rolling out the technology incrementally
Implementing generative AI is a complex process that involves significant changes to business processes and workflows. A gradual, phased rollout can help manage this complexity and minimise concurrent support requirements. It allows businesses to test the technology in a controlled environment, gather feedback, and make necessary adjustments before wider deployment.
Common approaches to technology adoption will work here. For example, if you select an off the shelf tool, start with a small group of tech-savvy employees to implement proof of concepts, and to act as champions for the rest of the organisation. When scaling your roll out, you will likely look to segment groups of users into manageable groups. This ensures you have the appropriate support capacity to deliver customised training, responsive tech support and more.
5. Driving adoption through Change Management
Change management plays a crucial role in driving the adoption of generative AI. It involves preparing the organisation for the changes brought about by the new technology, managing resistance, and facilitating a smooth transition. Your roll out should be designed with fundamental change principles in mind.
Rather than provide a change management 101 here are some of the core things I always think about to drive adoption of technology solutions:
- Clearly planned communication and roadmaps
- Differentiated learning and development
- Full documentation and training material
- Ongoing and iterative support, both technical and user focused
- Employee reward
- Visible leadership buy in
Combining these things is a simple means to drive success at scale.
Conclusion
The successful adoption of generative AI across an organisation is more complex than the simplicity of the tools would suggest. Organisations won't succeed relying on ChatGPT alone and require a well-planned strategy and implementation roadmap to be successful.
If you’re interested in how you can deliver this for your organisation, Columbus’s Generative AI strategy framework runs through the tips in this article and more to get you on the right path.
Reach out to Charles Wright today to find out more.