.png)
Everyone who has ever had a job knows that it can be intimidating to get started in a new role. You come into the workplace on day one knowing very little about the way things operate, who key stakeholders are, and what the company culture is like. Companies that prioritize employee development put a lot of stock into getting their team members ramped up, providing context and structure to ensure new workers are put in a position to succeed. So why should things be any different for an AI Agent?
While an AI Agent might not feel first day nerves the same way other employees do, they’re still a blank slate when it comes to context and relevant information about the company. An AI Agent can’t make answers up out of the blue—it’s only as good as the training and information you provide it. With that in mind, here’s how you can make sure you’re getting the most out of your AI Agent as you bring them onboard.
Train Your AI Agent Like a New Hire
An AI Agent has the potential to be as fast, knowledgeable, and efficient as any high-performing member of your team, but it won’t happen overnight. Just like any new employee, your AI Agent isn’t going to know any of the specifics of your property management operations on their first day. That’s why it’s crucial to train your AI Agent exactly like you would a new hire. Here’s what you should expect and consider when training your AI Agent.
Think of Your AI Agent as a New Employee
AI can be thought of as a bright new team member, with immense potential but still in need of a ramp-up period. Here’s a few questions you should ask yourself when training your AI Agent just like you would a new team member:
- What company-level knowledge is traditionally relayed to new hires?
- What property-level knowledge is traditionally relayed to new hires?
- What guidelines do existing training materials provide for answering questions related to fair housing, rental requirements, screening requirements, or other sensitive topics that require consistent answers?
Make sure to take the time to input this information into your AI Agent’s Knowledge Bank, just as you would sit down and review information with a new hire.
Encourage Onsite Team Members to Input Knowledge
There’s a common misconception that AI “makes up answers” to questions, but the reality is that it pulls information from a system of record to provide sufficiently helpful information. For example, EliseAI’s conversational AI tools utilize information available from the property management software (PMS) and supplemental information that has been added to its Knowledge Bank. One major positive of this feature is that you have full visibility into what your AI Agent does and doesn’t know about the community.
With that in mind, you should encourage onsite team members to get involved in adding supplementary knowledge for your AI Agent to reference. Do you train new team members on common FAQs from prospects and residents? Do the same with your AI Agent! Doing so will allow them to see in practice how AI can help scale their communications using the insights they provide to create helpful responses to both prospects and residents.
Optimize Your AI Agent’s Performance
Ramping up your AI Agent to perform at its full potential requires some initial feedback and knowledge sharing, just like with a new hire. The more time and effort you put into optimizing your AI Agent from the start, the better it will perform and the quicker your ramp up period will be to maximize the machine-learning aspect of AI support. Here’s what to keep in mind when you’re looking to improve the performance of your AI Agent.
Review Pending Knowledge to Optimize Automation
Think of any outstanding tasks in “Pending Knowledge” as if it were a new team member asking for clarification on how best to answer a customer’s questions. Just like a new hire who doesn’t want to say the wrong thing, your AI Agent may want to double-check before it answers a question, or if it needs to surface information it doesn’t yet have the answer to address.
With that in mind, it’s imperative that you review “Pending Knowledge” on a regular basis (we recommend doing it on a daily basis to start) to ensure you’re giving your new AI Agent all of the insights they need to effectively communicate with customers.
Input Knowledge Updates When Things Change
When things change - be it a new leasing special, a new SOP or a change in policy - those changes have to be communicated to the onsite team. Similarly, your AI Agent also needs to be kept in the loop to ensure it is communicating correct information to prospects and residents. If there is training being provided to introduce or reinforce an update, consider if the information being provided also needs to be input into your AI Agent’s Knowledge Bank. You’ll notice that your Knowledge Bank includes a section specifically for Specials, which we’ve seen tend to change most frequently.
If you see conversations where there is a misunderstanding or miscommunication with what your AI Agent conveyed, start with seeing if a Knowledge update will address the issue. In many cases, providing the AI Agent with new information will help resolve any subsequent instances of the same question. If you stay on top of adding information into the Knowledge Bank, your AI Agent won’t make the same mistake twice.
Identify Who Is Responsible for Knowledge Updates
Choosing a point person for everything AI is key to keep the project rolling forward. In order to ensure timely approval of Pending Knowledge updates or opportunities for further Knowledge enhancements, you should make one member of each of the onsite & corporate teams directly responsible for managing Knowledge updates.
Onsite team members can focus on inputting the answers to common community-based questions they receive on topics ranging from parking, to packages, to rent payments. Corporate team members can focus on topics and answers that are relevant across the portfolio or subsects of the portfolio, on subjects like:
- Application screening requirements
- Income requirements for application approvals
- Acceptable forms of income
- Acceptance and management of housing vouchers
- Corporate-approved responses to common fair housing-related questions
- Clarifying answers to corporate-managed services and processes, like utility billing
If there are subjects where - in an ideal world - agents should (or are legally required) to say the same thing to every person, every time, those topics are perfect additions to your AI Agent’s Knowledge Bank. That way, you can ensure your AI Agent is able to handle even the most delicate questions in a standardized, company-approved manner.
Metrics to Watch
Data gathering and reporting is an important part of measuring the efficacy of your AI Agent. Two important metrics to keep an eye on when optimizing your AI Agent are handoff percentage and early takeover rate. Handoff percentage refers to the % of conversations that your AI Agent is forced to hand off to an onsite team member because it lacks the knowledge necessary to answer the questions. In general, high-performing teams have a handoff percentage of around 5%, meaning the AI Agent is able to handle 95% of conversations autonomously and has to hand off 5% of conversations to a team member to take over.
If your handoff percentage is over 10%, your AI Agent is likely missing some key information in its Knowledge Bank that could help it more efficiently support your operational efforts and provide instant responses to customers. This is akin to having a team member who doesn’t have all of the information they need and has to ask for help too often. Teach the AI Agent what it’s missing by approving Pending Knowledge, or adding in additional knowledge to fill in the operational gaps.
Early takeover rate is the percentage of conversations where a team member took over a conversation that your AI Agent was currently managing, without needing help. We see this occur when onsite leasing teams want to conduct supplemental follow-ups and take over entire conversations. There are pros and cons to this approach. It does allow agents to quickly provide value by resolving prospect and resident inquiries, but it also limits the ability of your AI Agent to get better over time. This is akin to telling your new team member, “That’s okay, I’ll do it myself,” which prevents that new team member from learning how to do things themselves in the long-term. In general, high-performing teams have an early takeover rate under 5%, ensuring that they are maximizing the use of AI for customer conversations, so they can focus on customer experiences.
What’s important to note is that these two metrics go hand-in-hand. As more information is added to the Knowledge Bank, you should see your handoff percentage decrease. As team members observe that the AI is able to answer more questions, it builds trust in the technology and subsequently reduces early takeovers.
Plan for Seamless Handoffs
Just like a new employee getting ramped up, your AI Agent will occasionally need to loop in a more experienced team member to help them resolve an issue they haven’t faced before. These handoffs are both an opportunity to improve your AI Agent’s performance as well as an opportunity for a conversation with a customer to be missed. Here’s how you can ensure you’re set up for success when it comes to handoffs from AI Agents to team members.
Get Ahead of the Hard Part
If the handoff from your AI Agent to an onsite team member isn’t handled promptly, there’s a chance the delayed response could result in either a negative interaction for a resident or a missed opportunity to convert a new lead. With that in mind, it’s important to be proactive rather than reactive when laying out standard operating procedures for handoffs. Update your leasing policies around handoffs so that onsite teams have clear instructions as to whether handoffs should prompt an outbound phone call, text message, or email to continue to foster relationship-building with the customer. If you leave it up to the discretion of the agent, you open yourself up to the risk of inconsistency in communications that could negatively impact customer experiences.
It’s also pivotal that you identify who onsite is responsible for handoffs for accountability purposes. Setting a designated point of contact for handoffs and making it a key part of their responsibilities both promotes adoption of the AI Agent as well as ensures handoffs don’t slip through the cracks. If the team member responsible for handoffs is experienced enough, you could also consider having them be responsible for reviewing “Pending Knowledge” on a daily basis. Making one experienced team member the AI Agent’s de facto “onboarding buddy” could help get the AI up to speed quickly, while also ensuring that none of its handoffs slip through the cracks as your team becomes familiar with working alongside an AI. This has the added benefit of utilizing AI to scale your highest performing team members.
Looking to Hire the Best AI Agent on the Market?
Now that you have the framework you need to get your AI Agent up to speed fast, it’s time to consider which one to bring onboard. If you’re looking for an AI Agent to manage renewal offer issuances, speed up the lead-to-lease timeline, or take over collections conversations with current and past residents, look no further than EliseAI.