Should they deploy a AI chatbot now or wait?
I believe that PulsePoint should deploy an AI chatbot but in a careful manner by launching a constrained, human-in-the-loop pilot now rather than full automation. PulsePoint should take advantage of GenAI’s upsides, such as efficiency and more personalized responses. As John said in the case, the pilot should also “offer opt-outs and hybrid models during the transition,” along with imposing proper guardrails and explicit success metrics, to mitigate risks, such as hallucinations, privacy, customer backlash, and missed upsell and cross-sell intuition. Additionally, as Dharmesh Shah says in response to the case, customers care most about competence, consistency, and convenience which GenAI can increasingly meet, but customers still want the option to reach a human, further giving support to launching the hybrid model without delay.
What should they consider as they make that decision?
As they make the decision to green-light the GenAI salesbot, PulsePoint should preserve “personal client relationships,” which its customers like Tyrell Durant value, by giving customers the opportunity to speak with a human rep. PulsePoint should also keep humans in the loop where sales depends on “intuiting needs and recognizing hidden opportunities,” a concern that Mark raised. Additionally, PulsePoint should take into account risks like “hallucinations, data privacy, [and] outages” when designing the bot and set rules and keep records, as it “will never fully understand or control the model,” as John initially said. Lastly, PulsePoint needs to be aware of “the implications of head-count reduction,” which Linda mentioned, with a transparent people plan. Ultimately, as I have seen through past experience, implementing new AI solutions is most smooth and effective when one takes a human-centered approach that aligns the technology with the needs of all of the people involved.
