The AI world, and specifically the Generative AI world has been discussing the transition from AI Tools to AI Agents. Since 2020 multifamily has done an exceptional job at embracing new technology and achieving an impressive “catch up” on implementation of cutting edge tech in our space. While one could argue that multifamily has traditionally lagged in the uptake of modern software and tech options, we are seeing that change at a more rapid pace each year.
AI bursting on to the seen has caused both great progress and great chaos. As marketers and operators how do we know what “good” AI looks like and what “bad” AI looks like? How do we make the right choice of what platforms to implement at our properties? How do we monitor and govern AI solutions we implement?
These are all the right questions to ask, and these question can and will be the topics of many panels, webinars, blog posts, and podcasts to come.
But the specific topic I want to dive into here is how do you know if the AI platform you are implementing is actually going to move the needle on the metrics you measure as marketers and operators at your properties?
AI Tools < AI Agents
When ChatGPT first went viral in late 2022/early 2023 the worlds imagination exploded with the possibilities of what Generative AI could do for us and how we could begin using Generative AI in our everyday lives. Despite the grand possibilities the general public pretty quickly gravitated toward leveraging Generative AI to accomplish tasks in our existing workflows. This idea of accomplishing a task is what we would call an AI Tool and lacks the impact of what we would call an AI Agent.
While this was an incredible advancement in the average everyday person leveraging AI I would challenge us to ask how impactful it actually was. Lets use a specific example of replying to an email.
If you asked someone “how can you use AI to help you with this job of replying to an email” most people would pretty quickly show you how ChatGPT can write a brilliantly worded email on their behalf. If we zoom out, we notice that this is actually a very small aspect of the actual job at hand.
Example of an AI Tool
Think of all the steps you really take when replying to an email:
- First you have to notice the email in your inbox
- Then you have to take the the entire conversation in the email chain prior to that message into account to understand the context of what is being said
- Then you need to understand your goal of how you want to reply to that email
- You might need to cc or loop in additional contacts into the email chain
- Next you type, proof read, and send your email
- You send off your email
- You reach out to a few people on Slack or Teams to update them on the progress of this potentially business critical situation
- You update the notes in your internal CRM or systems so everyone is up to speed on this account
In this example, there are 8 steps, some of them quite complicated. In our example of AI Tools, an AI Tool can help you with one third of task number 5. The rest is up to you. This is where AI Agents shine and start to provide real value to our business.
AI Agents > AI Tools
AI Agents offer us real impact on our businesses by carrying out complex tasks from start to finish. In our example above, an AI agent can handle steps 1 through 8. This allows us to start thinking of the benefits our AI platforms provide in terms of outcomes not tasks completed. Outcomes have real impact of the KPIs and metrics we already measure.
In the case of Nurture Boss, an AI Agent can be measured on leads nurtured, leads converted to tours, tours converted to applications, new move ins, on time rent payments, positive online reviews, delinquency reductions, increased renewals, reduced vacancy loss, and so on. As you can imagine, that Tool you use to write your email can’t be measured against those metrics.
Example of an AI Agent
Lets look at our email example above to see how an AI Agent outperforms and AI Tool:
- First you have to notice the email in your inbox. An integrated AI Agent is immediately aware of a new email.
- Then you have to take the the entire conversation in the email chain prior to that message into account to understand the context of what is being said. Your AI Agent can consume not only the entire conversation to date but also any previous communication with that contact giving the AI Agent accurate context to form a reply
- Then you need to understand your goal of how you want to reply to that email. AI Agents understand their mission, whether it be collecting rent or scheduling a tour, and are capable of driving conversations toward that goal
- You might need to cc or loop in additional contacts into the email chain. AI Agents can understand when additional help is needed, these guide rails are critical to success
- Next you type, proof read, and send your email. AI Agents handle this with ease
- You send off your email. Integrated AI Agents accomplish this
- You reach out to a few people on Slack or Teams to update them on the progress of this potentially business critical situation. AI Agents can be instructed to follow any set of workflows, including the example listed here
- You update the notes in your internal CRM or systems so everyone is up to speed on this account. Robust integrations allow AI Agents to work across multiple systems to accomplish complex goals
In Conclusion
As we all continue to boost our education on AI and specifically Generative AI, we must always remember to “think bigger”. AI Tools are exciting and often easy to implement. However, AI Tools are not going to launch multifamily into the next era of efficiency and progress. AI Agents are robust solutions that can automation entire workflows and drive measurable outcomes. Remember, AI can take action and we should be implementing platforms like this at our properties.
