With an estimated $1 trillion in savings on the table, there’s good reason banks are scrambling to adopt artificially intelligent solutions that promise to reduce operating costs by 22% and increase revenue by 34%. Most of the buzz surrounds technologies such as chatbots, integrated receivables, fraud detection and robo-advisors; however, AI is more than just tech. If you want to implement solutions that can scale, you need to invest in people and processes, too.
Why banks struggle to scale AI projects
When a bank commits to AI, it can unwittingly set up roadblocks to success. Commonly, banks fail to understand that artificial intelligence permeates the entire organization, and therefore changes the organization and how it must conduct business. AI alters workflows and processes, so traditional management hierarchies can impede progress.
Banks can also overlook the role of humans in AI. It’s not an independent solution, but one that requires human input to succeed and scale. It becomes part of everyday work life for all leadership and staff. Without proper training for staff members, it’s impossible to harness AI’s full potential. Moreover, employees can become fearful of new technologies they view as threats to their livelihoods, which discourages collaboration and motivation to help improve artificially intelligent solutions.
These are the reasons that banks with the most successful, scalable AI solutions invest heavily in people and processes; indeed, they often invest just as much (or more) in people and processes as they do in the technology itself. They understand that for AI to realize its full potential, organizations must embrace a cultural shift that fuses the power of AI with smart workflows executed by skilled people.
Here’s what you can do to ensure your artificial intelligence initiatives include people and processes so you can position your bank for scalable success.
Invest in people
The better your staff members understand AI and how they can leverage it to their benefits, the more successful you’ll be. And it starts at the top: leadership must become well-versed in the technology so they can identify ways it can improve performance and so they can secure buy-in from banking staff.
Develop training programs that teach employees how to work with, not against, AI and how it can help them achieve their own desired outcomes. Ongoing employee development programs are fundamental to successful business; you’re just adding an AI track to the mix.
Training isn’t limited to IT staff, either. All bank personnel should understand how artificial intelligence impacts their jobs, and new roles should be created to address new demands. For example:
A member of leadership should be trained as a “translator” to serve as a liaison between tech and business staff. The ideal candidate should already have a keen understanding of business operations. By learning how AI works, the translator can connect business and technical stakeholders
Another leadership member should adopt the role of “evangelist” to champion projects across the bank. The evangelist needs to understand different use cases for AI and how they can positively impact the bank’s bottom line
“Product owners” spearhead individual projects, and they’re ultimately accountable for the outcomes. It is their job to ensure a given project is successful, to report progress to leadership and other stakeholders, and to track results
Human Resources personnel should learn to properly evaluate candidates for new roles such as data scientists, engineers and management specialists. If they’ve only vetted candidates based on business and finance acumen, how else will they know if the perfect data engineer is sitting in front of them?
Some organizations have added an “ethicist” role to study the impact of AI on consumers and evaluate potential bias in machine learning. This is another position that requires ample training
Other staff members must be trained in new workflows and technologies. Loan officers, for example, should learn how to use the predictive power of AI to make data-based judgments about loan applicants. You’re showing staff how to use new tools, but also assuring them the new tools aren’t there to replace them, rather to help them optimize performance
You’ll need to budget for employee training, of course, as well as continual development; but the investment will likely be well worth it as it will pave the way to scalable AI solutions that could pay massive dividends.
Invest in processes
Artificial intelligence changes the way your bank does business, which means it’s time to rethink traditional hierarchies and workflows. AI should be cross-functional, which means it doesn’t reside within traditional IT departments or solely within executive leadership.
Organizations that have successfully implemented AI in various industries recommend creating an AI Center of Excellence (CoE). The CoE exists independent of other departments yet enables teams to work side by side toward common goals.
This contrasts with a common approach in which individual teams work independently to implement agile AI solutions. The problem with that approach is you end up with different splinter solutions: each team has its own application built on its own architecture. Teams do not consider the need to mesh with other departments, and since nothing works together, the “system” is impossible to scale.
This isn’t to say you should centralize every component of AI, or that various departments should not have input. Rather, your CoE should manage data governance, set architectural standards and develop training programs; while individual teams or departments should focus efforts on things like workflow design. That way, staff have a vested interest in helping you achieve success.
It’s all about building a digital culture. AI isn’t just a tool, it’s a way of life and a way of doing business. The business doesn’t change, but the way you conduct it does.
Understand that artificial intelligence is as much about people and how they work together as it is about gaining new insights, creating efficiencies and generating revenue. When you set your sights on a new AI project, make sure your budget includes room for adoption and training. In this manner, you can streamline processes that mesh organization wide and build AI solutions that can easily scale.
Perfection doesn’t just impede AI projects, it stops them in their tracks. One example: even though data drives AI projects, if you try to harmonize every aspect of your data before moving forward, you’ll never get anywhere.
Successful banks develop a process in which they prioritize business and end-user needs first, then plan for quick wins. They develop small pilot projects comprised of minimum viable products to illustrate the full potential of AI, which makes it much easier to earn buy-in from bank leadership. Then, they build out full-scale projects based on those successes.
For example, let’s say you want to implement a chatbot capable of delivering personalized offers based on customer behavior and transaction history. Since it will require deep machine learning and intelligent automation, such an initiative might require several budget cycles to complete; indeed, it might be years before your chatbot is up and running.
A better approach might be to quickly implement a chatbot that delivers generalized offers. While you work toward automated personalization, the chatbot can collect data: which offers yield the greatest conversions? Which customers are most likely to act? Which products do they respond to? What time of day is best to deliver the offer? This type of data can later be used to deliver the personalized offers you originally set out to achieve.
Successful banks aren’t afraid to fail, either, so long as they learn from failure and apply that knowledge the next project. An oft-repeated AI adage is “fail fast, learn fast.” It’s one your bank should adopt if you hope to build artificially intelligent solutions that can be quickly scaled upon success.
That leads us to another key point: while most organizations find it easy to list desired outcomes, many fail to properly track results – so they don’t even know if their “solutions” achieve the stated project goals.
You need to develop a process to measure and track outcomes, which will make it easy to determine whether a given project is a success or not. If it’s not successful, you can then identify what went wrong instead of scrapping the project wholesale. Often, a single change can turn failure into success.
If your project is a success, the ability to demonstrate that with measurable outcomes makes it far easier to get buy-in not only from bank leadership, but also the people responsible for working with AI products every day.
Remember how we said to forget “perfect?” That’s a true statement, but that doesn’t mean you shouldn’t strive to develop AI applications that achieve desired outcomes. Tracking is part of the equation, but you should also have processes in place to continually develop your solutions. Not only that, but you must hold staff accountable for applying what they’ve learned about using your AI tools. Plan to improve in increments so both your data and AI tools improve over time.
No one’s expecting you to become an expert in AI overnight. Artificial intelligence can represent a major upheaval in your bank, which is why it’s important to reach out to experts who can help.
Consulting services and technology firms that have experience working with banks understand your unique challenges, are well-versed in various use cases for AI in banking and know how to develop solutions and implement processes that achieve desired outcomes. They can train your leadership team on AI and even help set up staff training programs – or at least point you in the right direction.
Perhaps most importantly, experts understand the role people and processes play in AI. They know AI tools must be able to scale, and that people and processes are paramount to scalability. Thus, they can help you develop your people and your processes to put you on a path to success.
When you work with outside resources, you can rapidly jumpstart projects with a resource push, then reduce the workload as your AI project matures. It eliminates the ebb and flow on your bank’s internal staff, while allowing them to learn and grow alongside your application. In short, outside resources make it easy to scale.
Be sure to properly evaluate technology partners before you commit to working with one. For example:
Create a scorecard
Determine the most important criteria your partner must meet, then rank potential partners according to that criteria. Common elements include industry knowledge, past experience, available functionality and customer service. You might also include training options and whether they can help you develop ongoing processes that foster success.
Vet their expertise
Use your scorecard to create a shortlist of potential partners, then screen each candidate with phone or in-person interviews. Follow up on references and carefully consider online reviews to drill down to the best candidates. The more research you conduct, the clearer your choice will be.
Assess cultural fit
Establishing a digital culture is critical to scalability, and since your technology partner will be ingrained in your bank it’s important they’re a good fit for that culture. Make sure any partner shares your core values and has good communication skills. You want their team’s personalities and work styles to mesh with your team’s so you can create a trusting working relationship.
Every technology project carries risk, so you need to understand those risks and plan accordingly. External partnership risks can be skills-based, reputational, financial or regulatory.
Scalability should be a fundamental component of any AI project. However, many organizations struggle to scale AI tools because they overlook two of the three most crucial factors: people and processes. Recognize that artificial intelligence will change your organization forever, and that it will permeate every aspect of your bank. Develop training programs to bring your people into the fold and establish processes that foster intelligent, efficient workflows. Invest as much – or more – into people and processes as you do technology, and you’ll be able to build AI tools that can easily be scaled when you’re ready.