Before and After: A Demo of a Better Enquiry Management Workflow

This is a fictional workflow demonstration, not a customer case study. The example illustrates how a service business could improve the way it handles enquiries. It does not represent measured results from a real client.

A new enquiry arrives through WhatsApp. Someone reads it, copies the details into a spreadsheet, and plans to follow up later.

Meanwhile, another team member checks the inbox, asks for the same information, and tries to work out whether anyone has responded.

Nothing appears seriously wrong. But the process involves repeated work, unclear ownership, and several opportunities for follow-ups to be missed.

This is where reviewing an enquiry management workflow can help.

Instead of immediately introducing AI, start by understanding how an enquiry moves through the business, where delays occur, and which steps could be simplified.

1. Before: The Manual Enquiry Process

Imagine a fictional service business receiving enquiries through WhatsApp, email, and its website.

The current process looks like this:

1. A customer sends an enquiry.
2. A team member reads the message.
3. The customer’s details are copied into a spreadsheet.
4. Another team member checks the conversation for context.
5. Someone tries to determine who should respond.
6. A follow-up is planned, but its timing is not always recorded.

Each step may seem manageable on its own. The difficulty comes from moving between tools and relying on people to remember what happens next.

The workflow problem: There is no consistent view of the enquiry, its owner, its current status, and its next action.

2. The Hidden Cost of Repeated Work

Manual processes can create work that is easy to overlook.

For example, a team might:

• Enter the same customer details into multiple places.
• Search old messages to understand a conversation.
• Ask colleagues whether an enquiry has been answered.
• Reconstruct missing CRM information.
• Remember follow-up dates without a reliable reminder.
• Recheck whether a meeting has been scheduled.

These activities consume time and make it harder to see the current state of each enquiry.

However, not every manual step is unnecessary. Some checks protect data quality, provide context, or require human judgment.

The goal is to identify repetitive work that can be simplified without removing useful checks.

3. After: A Proposed Enquiry Workflow

Now consider a proposed process for the same fictional business.

The objective is to create one clear path from the first enquiry to the next action.

Step 1: Capture the enquiry

Record the enquiry in an agreed system, including its source and the information needed to respond.

Step 2: Assign an owner

Make responsibility visible so the team knows who is expected to take the next action.

Step 3: Record the status

Use clear stages such as New, In Progress, Waiting for Customer, and Resolved.

Step 4: Define the next action

Record whether the next step is to reply, request more information, arrange a meeting, or wait for a response.

Step 5: Set a follow-up reminder

Where appropriate, create a reminder based on the agreed process instead of relying entirely on memory.

Step 6: Update the record

Keep the enquiry status and relevant notes current so the next team member can understand what has happened.

Depending on the business's existing tools, this process might use a spreadsheet, a CRM, simple rules, or an automated workflow.

AI may be useful for selected tasks, but it is not automatically necessary for every step.

4. Where Should the Human Checkpoint Be?

Automation should not mean that every enquiry receives an automatic response regardless of context.

The proposed workflow should define when a person needs to review or take over.

For example, a human checkpoint may be appropriate when:

• The customer's request is unclear.
• Essential information is missing.
• A pricing exception requires approval.
• The request falls outside approved response guidelines.
• A complaint or sensitive issue needs personal attention.
• The system cannot confidently determine the next action.

In these situations, the workflow should preserve the available context and route the enquiry to the appropriate person.

The team should also be able to correct records, pause an automated action, and handle exceptions.

A useful workflow defines both what can happen automatically and when the process must return to a person.

5. How Would We Measure Whether It Works?

A proposed workflow should be evaluated using defined measures rather than assumptions about its impact.

For a small pilot, a business could track the following metrics before and after the change.

Metric	What to measure
First-response time	Time between receiving an enquiry and the first response
Follow-up completion	Percentage of due follow-ups completed within the agreed period
Unassigned enquiries	Number of enquiries without a clear owner
Record completeness	Percentage of enquiry records containing the required fields
Duplicate data entry	Frequency of the same information being entered more than once
Scheduling delays	Time between a meeting request and a confirmed booking
Human escalations	Number and type of cases requiring manual review

Establish a baseline before changing the workflow. Then compare results over a defined period, using similar enquiry types and volumes where possible.

For example, the team could review a sample of enquiries before the change and another sample after implementation.

Document any changes in staffing, enquiry volume, tools, or operating procedures that could affect the comparison.

Do not assume that introducing automation will improve every metric. The results need to be measured.

6. Limitations: What This Demo Does Not Prove

This example illustrates a possible workflow design. It does not establish that a particular business has experienced these problems or achieved specific improvements.

Several factors could affect the outcome:

• Data quality: Missing or incorrect details can still create problems.
• Team adoption: A process works only when people follow it consistently.
• Integration limits: Existing tools may not support every desired connection.
• Exception handling: Unusual enquiries may require manual review.
• Privacy and access: Customer information must be handled appropriately.
• Measurement: Results depend on the baseline, observation period, and quality of the data.

A better-defined workflow can make responsibilities and next steps clearer, but it cannot guarantee more sales, faster conversions, or a particular financial return.

The right solution depends on the actual process, the tools already in use, and the needs of the team.

How Gravity For AI Approaches Workflow Improvement

At Gravity For AI, the starting point is the business process—not the assumption that every problem needs an AI agent.

A practical review begins with a few questions:

1. Where do enquiries arrive?
2. Which steps are repeated manually?
3. Who owns each next action?
4. Where do delays or handoff problems occur?
5. Which steps need human judgment?
6. How will the business measure whether the new process works?

The answers help determine whether the appropriate next step is a clearer procedure, a simple rule, a CRM improvement, workflow automation, or an AI-supported process.

The aim is to make the workflow easier to understand, maintain, and evaluate.

Want to Review Your Enquiry Workflow?

Start with one enquiry path. Map what happens from the first message to the final follow-up.

Identify one repeated task, one unclear handoff, and one measure that would help you evaluate the process.

You do not need to automate everything at once.

Before choosing the technology, make the workflow visible.

If your business manages enquiries across multiple channels, consider which step creates the most uncertainty for your team.

That is a useful place to begin.

Note: All business details and workflow examples in this article are fictional and illustrative. They are not a customer case study or a claim of measured results.