AI for Business: Adoption challenges - people

This is post #7 in our AI for Business series.

This week we will be focusing on the largest hurdle to a successful AI-powered business transformation. No, it’s not broken code, missing data or limited budget. It’s not even data quality or computing power. It’s Brian from finance who still prints out his emails.

That’s right, it is your people. At the end of the day, people have the ability to make or break the adoption of a new technology in any organisation.

Even if you have carefully identified your problems, goals, processes and where you can add customer value, you are not set up for success without first considering the people on your team. Folks generally don’t respond well to top down change, so it is important that they feel they are part of the solution and not part of the problem.

Below are some challenges and recommendations that your business will need to address to ensure you can reap the benefits of a successful adoption of AI across the organisation.

---

Vision-led

Any move to incorporate AI into the company should see AI aligned with the company vision, not the other way around. Doing so ensures that AI serves a strategic purpose and is not just being adopted for its own sake. Your team is likely already on board with your company vision, so it is less of a leap for them to get onboard with your AI strategy if the two are intertwined.

While internal or external pressure may make us feel like we need to invest in a new technology, if it doesn’t fit in with the company mission, people will ultimately be resistant to its adoption.

---

Organisational Structure

A flat, flexible and silo-free organisational structure will allow people and AI to work effectively together, maximising the benefits AI has to offer. If it’s possible, try to shift your organisation to a structure with flexible, cross functional teams set up to tackle specific problems, as opposed to individual roles and responsibilities.

This kind of structural shift is a major challenge, particularly for large and established organisations. Implementation will require engaging with employees, tolerating risk, and incentivising cross-functional coordination.

---

Communication, Expectations and Transparency

According to this article in Harvard Business Review (a former thoughtbot client, #humblebrag 😎), top performing companies spend significant time communicating and educating their teams, so that talent understands how machines make their jobs easier, not obsolete.

Let’s take an example from thoughtburgers (our delectable, AI-powered and entirely fictitious side-business, home to the tastiest and most thought-provoking burgers!). In blog #3 of this series, we looked at how Cognitive RPA could be used to automate processes. If we applied this to our delivery driver scheduling, the system could automatically, flexibly and predictively schedule drivers based on changing historic patterns in the data.

This move affects several steps in our delivery process. Therefore, it will be really important for us to communicate with our team of delivery managers that the automated scheduling of drivers does not make their role redundant; instead it will give them more time to liaise with their team of drivers and to tackle the long list of additional tasks in their remit.

To build trust in AI, it is imperative for leaders to communicate the vision transparently, explaining the goal, the changes needed, how it will be rolled out, and over what timeline. This alignment of expectations should leave employees feeling excited about the coming changes rather than resistant to them. This is a challenge for every leadership team.

---

Communication… again

It is important to note that transparency and communication must not stop after the initial plan is outlined to employees; there must be an ongoing process to ensure continued employee engagement.

For example, in blog #4 of this series, we explored how cognitive insights could be used to better triage thoughtburger delivery orders between branches during rush hours. But if we make this change, thoughtburger branch managers will need to know why an order is being sent to an alternative branch that is not theirs. What was the logic behind this decision? What could they do to increase the number of orders that do get sent to their branch?

Having visibility into these decisions and into other restaurants’ performances would give branch managers better insight into why their branch is not being chosen to fulfil orders as often as they would like. If this information is not clear or accessible, we risk alienating a branch. They will blame the new triaging system for an issue that they actually might have the power to resolve.

The ability to question the decision engages people in the process, it educates them and it could even lead to improvements in our order triaging algorithms if mistakes are spotted.

---

Training

Throughout your rollout, training should be provided to your team to help them manage the new systems. Incentives for learning and contributing, proper documentation, and the provision of ethical training could all help to build camaraderie and foster a growth mindset.

Ultimately, the changes you’re making should benefit staff by giving them time to focus on other tasks and by creating less stressful work environments. Providing training that demonstrate these benefits to them clearly should make employees more excited about the changes.

In some cases however, job displacement will inevitably occur. For example, if thoughtburgers were to pursue the drone delivery strategy discussed in our previous blog, we would need fewer regular delivery drivers on the ground. However, there could be an opportunity to retrain staff.

Drone handlers will be required to complete pre-flight checks, attach deliveries to drones and to ensure the fleet is running smoothly. New managerial roles like “Director of Drones” could also open. Furthermore, we will still need delivery drivers for larger orders which are too heavy for most commercial delivery drones or for orders going near a no-fly zone (like an airport, stadium or army base).

Training like this can be complex to organise at scale. A phased rollout works well as it accommodates phased training (we will discuss this topic in more detail in the final blog of the series).

---

Ownership

How the changes are rolled out is also crucial for employee adoption. Phased rollouts and pilot projects will make your team feel like they have some ownership over the adoption process. Giving workers this opportunity to tinker with the technology before a final adoption decision is made, eases the transition. Furthermore, your team will probably spot flaws and ways the system could be improved, giving them an even bigger stake in the new technology.

Again, we will discuss the topic of pilot projects and phased rollouts in more detail in the final blog of the series.

---

Measuring success

Defining and accurately measuring the success of an organisational change, rather than a specific project where AI is being applied, is another long term challenge you might face.

A unique set of metrics may need to be created. These metrics should address changes in behaviors and processes deep within your organisation. A key metric could be the speed with which new ideas are transformed into frontline tools, or a digital-maturity index to track how many processes have been automated.

---

Bottom line

People can make or break your organisation’s adoption of AI.

To get your team on board, you need to align your use of AI with the company vision. Be transparent with your team and keep two-way lines of communication open continuously. Let your team have ownership of the technology and give them an opportunity to shape it by starting out with pilot projects. Provide the necessary training on an ongoing basis and adjust your company structure to capitalise on the powers of AI. Finally, create metrics to measure the success of this organisational change and share those with your team.

---

💡 If you’re ready to start using AI to digitally transform your business, thoughtbot would love to work with you. Let’s talk about making your AI initiative a success!

This blog post is part of a series based on the course Artificial Intelligence (AI) for Business run by University College Dublin (UCD). I took this course and found it so helpful that I’m recapping my top insights. thoughbot has no affiliation with UCD.