AI in Landscaping: Why Better Data and Processes Must Come First

What landscaping leaders can do now to build the foundation for useful, responsible AI

What landscaping leaders can do now to build the foundation for useful, responsible AI

At the Lawn & Landscape Technology Conference, it was nearly impossible to have a conversation about the future without talking about artificial intelligence. There were demonstrations from automated client communications to robotics The possibilities are exciting, but one of the most useful lessons from keynote speaker Theresa Payton was much less futuristic:

AI cannot fix unclear processes, unreliable data, or a team that does not know how to use the information it already has.

Before landscaping companies automate more work, they need to understand the work they are trying to improve.

That does not mean waiting until every process is perfect. It means starting with a real operational problem, testing carefully, and making sure the people closest to the work are involved.

Here is a practical way to begin.

1. Don’t begin with AI. Begin with the bottleneck.

A common starting question is: Where could we use AI in our company?

A better question is: What repeatedly takes our team away from higher-value work?

Payton encouraged conference attendees to look for high-friction tasks: work that has to be done but consumes more time, attention, or manual effort than it should.

For a commercial landscaping company, that might include:

• Pulling information from several Aspire reports to prepare for renewals
• Reviewing work tickets to find scheduling problems
• Comparing estimated, scheduled, and actual labor hours
• Building weekly performance reports by hand
• Searching for the properties that need attention
• Turning an internal process into a training document
• Drafting repetitive customer updates
• Finding exceptions hidden inside a large amount of operational data

These are better starting points than trying to “add AI” across the entire company. They are specific. They can be observed. And they give the team something concrete to improve.

Try this with your team

Ask each manager to identify one recurring task that:

1. Takes at least an hour each week
2. Follows a repeatable process
3. Requires gathering or reorganizing information
4. Pulls them away from customers, crews, or strategic work

Do not evaluate tools yet. First, make a list of the work that creates the most friction.

2. Make sure your data deserves to be trusted

AI can process information quickly. It cannot make unreliable information accurate. That distinction matters in landscaping operations.

Consider a few common questions leaders may eventually want AI to help answer:

• Which properties are most likely to need a price increase?
• Which crews need coaching?
• Where are we losing labor margin?
• Which jobs are consistently taking longer than estimated?
• How many hours should we schedule for this property next season?
• Which operational problem should a branch manager address first?

The answers are only useful when the underlying data reflects what actually happened.

If crews enter hours days later, managers regularly correct time on their behalf, visits are completed without being scheduled, or tickets are dragged across several days, the analysis may be precise but still wrong.

The same is true when property information lives in multiple systems or when managers use different definitions for the same metric. Before introducing more advanced automation, review whether the inputs are dependable.

A simple data-readiness check

Choose one important operational metric and ask:

• Where does this data come from?
• Who enters or updates it?
• How quickly is it entered?
• How often is it corrected later?
• Does everyone interpret it the same way?
• Can a manager trace the number back to the underlying job or work ticket?
• Would we feel comfortable using it to make a pricing, staffing, or performance decision?

You do not need perfect data. You do need to understand where it is reliable, where it is not, and what your team must improve before making bigger decisions from it.

3. Choose one contained use case

The conference made one thing clear: landscaping companies have no shortage of possible AI applications. The challenge is choosing a first project that is small enough to manage and useful enough to matter.

A good first use case is:

• Repetitive
• Easy for a person to review
• Based on information you already understand
• Low-risk if the first result is imperfect
• Connected to a measurable business outcome

Good starting projects might include:

Turn an SOP into a training outline

Provide an approved operating procedure and ask the tool to create:

• A short training agenda
• A crew-level checklist
• A supervisor discussion guide
• A Spanish-language draft for review by a fluent speaker

The company still owns the process. AI simply helps reorganize it for different audiences.

Draft a renewal preparation checklist

Give the tool your existing renewal process and ask it to organize the steps by:

• Information required
• Owner
• Timing
• Decision point
• Customer communication

A manager should still review the property data and make the final pricing decision.

Summarize an approved internal report

Use a report that does not contain sensitive information and ask for:

• The three most important changes
• Exceptions that require review
• Questions a manager should investigate
• A concise summary for the leadership meeting

Every observation should be checked against the source.

Draft a customer communication

Use AI to prepare a first draft of a routine message, such as:

• A weather-delay update
• A site-visit recap
• A renewal meeting request
• A summary of recommended property improvements

A person should confirm the facts, tone, and final recommendation before anything is sent. The goal of the first pilot is not to prove that AI can do everything. It is to learn where it genuinely helps.

4. Put people and guardrails around the experiment

Payton repeatedly emphasized that AI should serve people, not the other way around. That requires clear boundaries.

Employees are already experimenting with public AI tools, sometimes with or without formal company approval. A blanket rule telling everyone not to use AI may simply push that activity out of sight.

A more practical approach is to define:

• Which tools are approved
• What information may be entered
• What information is prohibited
• Which outputs require human review
• Who is responsible for the final decision
• Where employees should bring new ideas or questions

Customer lists, pricing information, payroll records, employee information, proprietary processes, and identifiable customer data should not be placed into unapproved public tools.

Teams should also assume that AI-generated work may contain mistakes.

A polished answer is not necessarily a correct answer.

A basic review standard

Before using an AI-generated output, confirm:

1. Accuracy: Are the facts correct?
2. Source: Can the result be traced to approved information?
3. Context: Does it understand how your company actually operates?
4. Risk: Could an error affect a customer, employee, contract, price, or safety decision?
5. Ownership: Is a person clearly accountable for approving the result?

The more serious the decision, the more important human review becomes.

5. Create a repeatable learning rhythm

One of the most practical ideas from Payton’s keynote was her company’s “AI Power Hour.”

Her team regularly brings real business problems into a structured session, experiments with approved tools, documents what happened, and shares both failures and successes. Landscaping companies can adapt that model without turning it into a major technology initiative.

How to run a landscaping AI Power Hour

Before the meeting

Choose one operational problem from the team’s friction list.

Good examples:

• Preparing a weekly operations summary
• Reformatting a training procedure
• Organizing renewal preparation
• Drafting a customer update
• Reviewing a process for unnecessary steps

During the hour

10 minutes: Define the problem

Explain the current workflow:

• What triggers the task?
• Who performs it?
• How long does it take?
• What makes it frustrating?
• What does a successful result look like?

20 minutes: Test one approved tool

Ask the tool to assist with one clearly defined part of the process.

Do not ask it to redesign the whole company.

15 minutes: Review the result

Compare the output with the source information.

Identify:

• What it handled well
• What it misunderstood
• What it invented or missed
• What still required human judgment

10 minutes: Decide what happens next

Choose one:

• Stop the experiment
• Revise the prompt or process
• Test it again with another example
• Run a small pilot with one person or team

5 minutes: Document the lesson

Record the prompt, result, errors, time saved, and next step. The documentation matters. Without it, companies repeat the same experiments and mistakes.

Make failure safe

Payton’s keynote included an important leadership lesson: employees must feel comfortable reporting when technology does not work. A failed test is useful when it is small, documented, and caught before it affects customers or production.

If employees believe they will be blamed for an imperfect experiment, they are more likely to hide problems or avoid participating. Leaders should reward clear thinking and honest reporting, not just successful outputs.

6. Measure operational value, not AI activity

The number of prompts written is not a business outcome. Neither is the number of employees with an AI account. The real question is whether the experiment improves the operation.

Depending on the use case, track:

• Administrative hours eliminated
• Time required to prepare a report
• Number of corrections or rework cycles
• Time between identifying and addressing an exception
• Manager time returned to field coaching or customer work
• Adoption of a standard process
• Data-entry accuracy
• Speed of decision-making
• Consistency across branches or managers

For example, if a tool helps prepare a weekly report in 15 minutes instead of two hours, that is measurable. But the next question should be:

What did the manager do with the time that was returned?

The best result may not be a smaller payroll. It may be more time visiting properties, coaching a struggling crew, preparing for a customer conversation, or correcting a problem before it affects the margin. That is where AI can create practical value without trying to replace the people who understand the work.

A 30-day AI readiness challenge

Landscaping leaders do not need a multi-year transformation plan to get started.

Over the next 30 days:

Week 1: Find the friction

Ask managers and employees to identify repetitive work that takes them away from customers, crews, or higher-value decisions.

Choose one task.

Week 2: Check the foundation

Document the current process and review the quality of the information it depends on.

Fix obvious data or process problems first.

Week 3: Run one controlled experiment

Use an approved tool, remove sensitive information, and keep a person responsible for reviewing every result.

Week 4: Measure and decide

Compare the old and new process.

Ask:

• Did it save meaningful time?
• Was the output accurate?
• Did it reduce or create rework?
• Would the team use it consistently?
• Is the risk manageable?
• What did we learn?

Then decide whether to stop, revise, or expand the pilot.

The foundation is the competitive advantage

AI tools will continue to improve. They will also become widely available. Most landscaping companies will eventually have access to similar assistants, models, and automations. The lasting advantage will not come from being the first company to open an account.

It will come from having:

• Operational data the team trusts
• Clear processes worth improving
• Employees who understand the work
• Leaders who encourage responsible experimentation
• A regular rhythm for turning information into action

The same principle already applies to dashboards and business software. A tool can surface an issue. It cannot coach the crew, correct the process, talk with the customer, or build accountability across the team.

AI may help landscaping companies move faster. But first, the operation has to know where it is going.