A promising inquiry arrives while the owner is with a customer. The details sit in a text message. Someone copies them into a calendar. Tomorrow, another person asks the customer to explain everything again.
That hypothetical scene is where we would start looking for useful AI work. The opportunity is the distance between a person's request and a well-informed response.
Whitmore's view: businesses become strong candidates for AI transformation when they combine repeated administration, scattered information and relationships that need sustained attention. The industry label helps locate the opportunity. The operating habits determine whether a project can work.
Readiness begins with a repeatable job
Look for a task that happens often enough to compare results, uses information the business can provide, and has an owner who can explain what a good outcome looks like. Also look for a clear exception: the point where the system should bring a person in.
A busy inbox alone does not make a business ready. If nobody agrees which calendar is correct or who owns the next customer action, resolve that first. Otherwise, an assistant can produce a quicker version of the same confusion.
Real estate agents: continuity from inquiry to closing
Consider an assistant that turns an incoming inquiry into a contact record, brings together the agent's existing notes, and drafts a response with the next practical step. After a showing, it could prepare a follow-up using the buyer's stated questions rather than starting a fresh conversation.
The agent keeps negotiation, representations about a property and relationship-sensitive decisions. The system's job is to make the relevant history available and keep agreed next steps visible.
Start with inquiry intake and follow-up drafts. Track time to a useful response, missing contact details, overdue follow-ups and how often the agent has to rewrite a draft. Faster replies count only if they move the right conversation forward.
Real estate investors: assemble the file before the decision
An investor reviewing a potential acquisition could ask an assistant to organize seller-provided documents, flag missing information and prepare questions against a fixed review checklist. Each extracted figure should point back to its source. Conflicting figures should remain visible.
The investor sets assumptions, checks the evidence and decides whether to pursue the property. An attractive summary must not become permission to fill a gap with a guess.
Start with document intake for a single kind of property. Measure review preparation time, missing items caught before review, and corrections to extracted information. The useful outcome is a file a person can interrogate, not an automatically confident investment recommendation.
Local services: make the next appointment easier to arrange
For a hypothetical repair or home-service business, an assistant could collect the address, requested work and preferred time, check the approved service area, and prepare scheduling options. It could gather the information a technician needs before the callback.
Keep unusual scope, pricing exceptions and safety-sensitive questions with staff. A customer describing an urgent problem should have a clear route to a person, without being pushed through a routine booking sequence.
Begin with incomplete inquiry follow-up. Track the share of inquiries with usable details, time to a confirmed next step, scheduling corrections and staff minutes spent chasing missing information. Review the conversations that fail, not just the appointments that succeed.
Med spas: improve the front desk, preserve clinical judgment
A med spa could begin with approved answers about opening hours, consultation availability, locations and booking logistics. An assistant could prepare appointment options and route questions outside that administrative scope to the team.
Treatment suitability, symptoms and clinical advice stay with qualified clinicians. For a first trial, use sample inquiries and the approved public service information. Keep the workflow limited to what it needs to arrange the next conversation; adding access to clinical records is a separate decision requiring appropriate review.
Measure administrative response time, booking corrections, unresolved inquiries and how reliably clinical questions reach a person. A smooth booking flow is useful only if it knows when to stop answering.
Choose the first workflow by its evidence
Before a pilot, record the current handling time and a sample of common errors. During the pilot, include human review and correction time in the cost. Compare similar requests, and inspect exceptions alongside averages. Better customer continuity matters as much as minutes saved.
NIST's AI Risk Management Framework provides a voluntary foundation for considering risk through the design, use and evaluation of AI systems. It does not establish that these four sectors will outperform others; the workflow choices here are Whitmore's analysis.
Our starting recommendation is to choose one recurring handoff, give it a reliable source of information and decide in advance what improvement would justify expanding it. That is the kind of assessment described in Whitmore's services.
From this analysis
The strongest candidate is the business that can name the work, show where it breaks, and make someone responsible for improving it.
