How to Use AI to Decide Which Products or Services to Keep, Change, or Drop

A product can sell well and still be a bad product for your business.

A service can bring in steady revenue while consuming so much labor, owner attention, travel time, customer support, or overhead that it contributes very little profit.

And sometimes an offering that appears small on a sales report is actually one of the most valuable things you sell because it has strong margins, generates repeat customers, or leads to larger purchases later.

That is why deciding what to keep selling should involve more than looking at revenue.

AI can help small-business owners bring together sales, costs, margins, customer demand, time requirements, operational complexity, and strategic value so they can make better decisions about where to focus.

Direct Answer: How Can AI Help You Decide What to Keep or Drop?

AI can help you compare products or services using multiple factors at the same time, including revenue, direct costs, labor, margins, customer demand, repeat purchases, and the amount of effort required to deliver each offering. Once that information is organized, AI can help identify which offerings should receive more attention, which may need pricing or operational changes, and which may no longer justify the resources they consume.

AI should support the decision, not make it blindly. The quality of the recommendation depends on the quality of the business information you provide.

Why Revenue Alone Can Give You the Wrong Answer

Most businesses naturally pay attention to what sells the most.

That makes sense—but sales volume is only one part of the picture.

Imagine a contractor offers three types of projects:

• Small repair jobs generate $60,000 annually.

• Bathroom renovations generate $150,000.

• Kitchen renovations generate $220,000.

Looking only at revenue, kitchen renovations appear to be the obvious priority.

But then the owner looks deeper.

Kitchen projects require extensive quoting, subcontractor coordination, materials management, permit issues, customer changes, and frequent owner involvement. Bathroom projects are more standardized, finish faster, and produce substantially better margins.

The $150,000 service may ultimately contribute more usable profit than the $220,000 service.

This is the distinction between asking:

“What sells the most?”

and asking:

“What contributes the most value to the business?”

What Should You Measure Before Asking AI for Help?

AI becomes much more useful when you give it structured information.

You do not necessarily need perfect accounting data. Many small businesses do not track every minute or allocate every overhead expense precisely.

Start with the best information you have.

For each major product or service, gather the following.

1. Revenue

How much does the offering generate over a reasonable period?

Depending on the business, you might analyze:

• The last three months

• The last six months

• The last 12 months

• A full seasonal cycle

Avoid comparing one unusually strong month with a full year of another service.

2. Direct Costs

These are costs directly associated with providing the product or service.

Examples include:

• Materials

• Inventory

• Subcontractors

• Shipping

• Payment processing

• Sales commissions

• Supplies

• Job-specific equipment rentals

If a landscaping company charges $4,000 for a project but spends $1,900 on plants, materials, and subcontracted labor, those expenses materially change the value of the sale.

3. Labor and Owner Time

This is frequently underestimated.

How many employee hours does the offering require?

More importantly for many small businesses:

How much owner time does it consume?

Consider time spent on:

• Selling

• Quoting

• Scheduling

• Preparing

• Delivering the work

• Managing employees

• Correcting problems

• Answering customer questions

• Following up

• Collecting payment

An offering that requires six hours of owner involvement may deserve to be evaluated differently from one that employees can deliver independently.

4. Gross Margin or Contribution Margin

You do not need to turn the exercise into an accounting project.

At minimum, estimate how much remains after the direct costs required to deliver the offering.

A simple starting calculation is:

Revenue – Direct Costs = Contribution Before Overhead

If a service sells for $1,000 and requires $300 in direct costs, it contributes $700 before overhead and other business expenses.

Then consider labor and owner time.

That often changes the picture significantly.

5. Customer Demand

Profitability matters, but demand matters too.

Ask:

• Are customers regularly asking for this?

• Is demand increasing or declining?

• Do you have to discount heavily to sell it?

• Is it easy to explain and sell?

• Does it generate inbound inquiries?

• Do customers buy it repeatedly?

A highly profitable service that almost nobody wants may not deserve significant expansion.

6. Repeat Business

Some offerings have value beyond the first transaction.

For example, an HVAC tune-up might generate less immediate revenue than a large repair but create an ongoing relationship that leads to maintenance plans, repairs, and eventual system replacements.

A salon's basic haircut might not have the highest ticket price, but a client returning every six weeks could be worth far more over time than a one-time premium service.

7. Operational Complexity

Some products and services create disproportionate headaches.

Consider:

• Scheduling difficulty

• Special equipment

• Inventory requirements

• Vendor dependence

• Training requirements

• Customer complaints

• Returns

• Rework

• Customization

• Travel

• Licensing or permitting

• Quality-control issues

Complexity does not automatically make an offering bad.

It needs to produce enough value to justify the complexity.

8. Strategic Value

Not every offering should be judged solely on immediate profit.

Some offerings:

• Bring new customers through the door

• Lead to larger purchases

• Differentiate the business

• Create recurring relationships

• Strengthen a premium brand position

• Fill unused employee capacity

• Provide valuable referrals

This is why simply sorting a spreadsheet by margin can sometimes lead to the wrong decision.

The Keep, Change, or Drop Framework

A useful way to analyze your offerings is to place each one into one of three categories.

KEEP

An offering generally belongs in the Keep category when it combines several positive characteristics:

• Healthy margin

• Consistent demand

• Reasonable delivery effort

• Strong customer satisfaction

• Repeat or referral potential

• Strategic importance

Some Keep offerings may deserve additional marketing, capacity, or sales attention.

CHANGE

The Change category is often the most interesting.

These are offerings that have potential but contain a weakness that may be fixable.

Examples include:

• Strong demand but weak margins

• Good margins but excessive owner time

• High sales but too many custom variations

• Good customer retention but prices that have not increased in years

• Profitable jobs that take too long because the process is inefficient

Possible changes include:

• Raising prices

• Packaging services differently

• Setting minimum order sizes

• Removing customization

• Standardizing the process

• Reducing delivery areas

• Outsourcing part of the work

• Bundling services

• Improving purchasing

• Automating administrative work

Before dropping an offering, ask whether the underlying problem can be corrected.

DROP

An offering may be a candidate for elimination when several problems occur together.

For example:

• Weak margin

• Low demand

• Significant owner involvement

• Frequent problems

• Limited repeat business

• Little strategic benefit

Dropping an offering can also free capacity for more profitable work.

That opportunity cost matters.

If a contractor spends two days completing a low-margin project, those are two days the company cannot use for higher-margin projects.

A Simple Scorecard AI Can Help You Build

You can make the comparison more objective by scoring each offering.

Use a scale of 1 to 5 for:

Do not treat the resulting total as an automatic answer.

Use it as a way to identify what deserves further investigation.

For example, a service might score very well overall except for owner time. That suggests the service itself may not be the problem.

The delivery process may be the problem.

Step by Step: How to Conduct an AI Profitability Analysis

Here is a practical process a small-business owner can follow.

Step 1: List Your Main Offerings

Do not begin with every tiny variation you sell.

Start with the major categories that matter financially.

A plumber might use:

• Drain clearing

• Water heater replacement

• Fixture installation

• Emergency plumbing

• Sewer line repair

• Maintenance agreements

A retailer might analyze its main product categories rather than hundreds of individual SKUs initially.

Step 2: Create a Basic Data Table

For each offering, record what you know.

Include:

• Annual or monthly revenue

• Average selling price

• Estimated direct cost

• Estimated labor hours

• Owner hours

• Number of sales

• Repeat-purchase potential

• Complaint or rework frequency

• Demand trend

• Strategic importance

Estimates are acceptable if exact numbers are unavailable.

Just label them as estimates.

Step 3: Ask AI to Find Missing Information

This is an overlooked use of AI.

Instead of immediately asking:

“Which service should I stop offering?”

ask:

“What information is missing that would make this comparison more reliable?”

AI might point out that you know revenue but not labor hours.

Or you know gross margin but have not considered how often customers return.

That is useful because it prevents false confidence.

Step 4: Ask AI to Compare the Offerings

Once you provide the information, ask AI to analyze each offering across multiple dimensions.

You want patterns, not just rankings.

For example:

• Which services have strong demand but poor margins?

• Which produce good profit with little owner involvement?

• Which create repeat customers?

• Which consume disproportionate operational capacity?

• Which should be investigated further before making a decision?

Step 5: Separate Problems From Opportunities

Suppose one service has:

• Strong demand

• Good customer reviews

• High revenue

• Poor margins

That does not necessarily mean you should eliminate it.

AI might help you investigate whether the issue comes from:

• Underpricing

• Material costs

• Excessive employee hours

• Too much customization

• Travel time

• Poor scheduling

That offering could potentially move from mediocre to excellent with one operational change.

Step 6: Run Scenarios

AI is particularly helpful for exploring “what if” questions.

For example:

What happens if I raise the price 10%?

What happens if I stop accepting jobs under $500?

What happens if I reduce service territory from 40 miles to 20?

What happens if I standardize the service into three packages?

You should still validate the assumptions yourself, but scenario analysis helps you see alternatives before making a permanent decision.

Step 7: Decide: Keep, Change, Drop, or Investigate

Do not force every offering into an immediate decision.

You can use four categories:

Keep: Performing well.

Change: Valuable but needs adjustment.

Drop: Likely consuming resources without enough return.

Investigate: Not enough information yet.

That fourth category prevents premature decisions based on incomplete data.

Copyable AI Profitability Analysis Template

Use the following prompt with an AI assistant after replacing the brackets with your own information.

Small Business Offering Analysis Prompt

I want to evaluate which of my products or services I should keep, change, grow, or possibly stop offering.

My business: [describe business]

My primary business goal for the next 12 months: [goal]

Here are my main offerings:

Offering 1: [name]
Revenue: [amount]
Average price: [amount]
Estimated direct costs: [amount or percentage]
Employee labor required: [hours]
Owner time required: [hours]
Approximate number of customers/sales: [number]
Repeat business potential: [low/medium/high]
Customer demand trend: [declining/stable/growing]
Operational difficulty: [low/medium/high]
Common problems or complaints: [details]
Strategic value: [details]

[Repeat for other offerings.]

Analyze each offering across:

1. Profitability

2. Customer demand

3. Labor requirements

4. Owner time

5. Operational complexity

6. Repeat business

7. Growth potential

8. Strategic value

Then classify each as:

• KEEP AND GROW

• KEEP

• CHANGE

• POSSIBLY DROP

• NEED MORE INFORMATION

For each recommendation:

• Explain why.

• Identify the biggest risk.

• Identify one change that could improve the offering.

• Tell me what additional data I should collect before making a final decision.

Finally, rank the three actions that could have the greatest positive impact on my business.

Example 1: A Plumber Finds That Small Jobs Are Creating a Scheduling Problem

Suppose a plumbing company offers:

• Drain cleaning

• Fixture installation

• Water heater replacement

• Emergency repair

• Sewer replacement

Fixture installations generate a respectable amount of revenue, so the owner assumes they are valuable.

After comparing the services, however, the owner notices something.

Fixture installations frequently require:

• Multiple customer calls

• Driving across town

• Small material purchases

• Scheduling around customer availability

• One to two technician hours

The jobs are profitable individually, but they occupy schedule slots that could be used for water heater replacements or larger repair work.

AI might suggest testing several options before eliminating the service:

• Increase minimum service charges.

• Restrict installations to existing customers.

• Bundle installation with supplied fixtures.

• Set geographic limits.

• Offer certain appointment windows only.

The right conclusion may not be Drop fixture installation.

It might be Change how fixture installation is sold.

Example 2: A Salon Discovers a Lower-Priced Service Has High Strategic Value

Consider a salon comparing:

• Haircuts

• Color services

• Blowouts

• Treatments

• Extensions

Extensions have the highest transaction value.

Haircuts appear less attractive when comparing individual tickets.

But the salon then considers customer behavior.

Haircut clients may:

• Return every six to eight weeks

• Purchase products

• Add color services

• Refer friends

• Develop long-term stylist relationships

AI analysis can help reveal that the lower-ticket service plays an important role in customer retention and lifetime value.

Eliminating it simply because another service has a higher ticket price could damage the broader business.

Example 3: A Consultant Realizes Custom Projects Are Consuming the Business

A marketing consultant offers:

• Strategy sessions

• Monthly retainers

• Custom research projects

• Workshops

Custom research projects generate substantial invoices.

But every project requires:

• A custom proposal

• New research

• Different deliverables

• Significant owner involvement

• Multiple revisions

Monthly retainers generate slightly less revenue per client but have standardized work, predictable schedules, and recurring income.

The analysis may show that custom projects are not necessarily unprofitable, but they are difficult to scale.

Possible changes could include:

• Raising custom-project pricing

• Requiring larger minimum engagements

• Turning common requests into packaged services

• Limiting revisions

• Prioritizing retainer clients

Again, the useful question is not simply:

“Which service makes the most revenue?”

It is:

“Which business model gives me the strongest combination of profit, demand, repeatability, and manageable workload?”

Where AI Helps—and Where It Can Mislead You

AI is excellent at organizing information and comparing multiple variables.

It can help you:

• Structure messy business data

• Find patterns

• Identify missing information

• Compare scenarios

• Generate questions you had not considered

• Suggest operational improvements

• Turn analysis into an action plan

But AI does not automatically know your true costs.

If you tell it that a job takes three hours when it actually consumes six hours across quoting, preparation, service, travel, and follow-up, the recommendation will be distorted.

The same problem occurs with inaccurate margins or assumptions about customer demand.

A useful rule is:

Use AI to improve your reasoning, not replace your business records.

For financial calculations, verify figures against bookkeeping records, accounting reports, payment systems, inventory data, or other reliable sources.

Common Mistakes When Using AI to Analyze Profitability

Mistake 1: Providing Revenue but No Cost Information

AI cannot determine profitability from sales alone.

At minimum, provide reasonable estimates of direct costs and labor.

Mistake 2: Ignoring Owner Time

For many owner-operated businesses, this is one of the biggest hidden costs.

A service may look profitable because the owner's labor effectively appears as free.

It is not free.

That time could be spent selling, managing employees, improving the business, or delivering more profitable work.

Mistake 3: Treating AI's Recommendation as a Final Decision

If an AI assistant recommends dropping a service, ask why.

Then challenge the conclusion.

Ask:

• What assumptions led to this?

• What information could change the recommendation?

• Could a price increase fix the problem?

• Could operational changes improve it?

• Does the service have strategic value that the numbers do not capture?

Good analysis should survive follow-up questions.

Mistake 4: Looking at Margins Without Capacity

A high-margin offering may still have limited business value if it can only be delivered occasionally or requires specialized employees you cannot easily hire.

Capacity matters.

Mistake 5: Ignoring Customer Relationships

Some services lead to future work.

Others create referrals.

Others keep customers connected to your business.

Those benefits may not appear in a simple transaction-level profit calculation.

Mistake 6: Making Too Many Changes at Once

If you simultaneously raise prices, drop services, reduce territory, restructure packages, and change marketing, you may never know what actually improved the results.

Prioritize one or two meaningful changes and monitor what happens.

A Better Question Than “What Should I Stop Selling?”

When analyzing your business, start with a broader question:

Which offerings deserve more resources, which can become stronger with changes, and which are using resources that would create more value somewhere else?

That wording encourages a better decision.

Dropping products or services is only one possible outcome.

Often the biggest opportunity is improving something you already sell.

How Often Should a Small Business Review Its Offerings?

For many small businesses, a structured review once or twice a year is reasonable.

However, you may want to review sooner if:

• Costs have increased significantly

• Employee wages changed

• Customer demand shifted

• You introduced several new services

• Margins are falling

• The owner is becoming overloaded

• Certain services are generating frequent complaints

• Revenue is growing but profit is not

A restaurant facing rapidly changing ingredient costs may need to review menu profitability much more frequently than a consulting business with stable service costs.

The review frequency should match how quickly your economics change.

What Should You Do After Identifying Your Best Offerings?

Once you identify the offerings that create the most value, decide how to support them.

That could mean:

• Giving them more marketing attention

• Training employees around them

• Improving sales scripts

• Increasing available capacity

• Creating packages

• Building recurring plans

• Improving follow-up

• Targeting customers most likely to buy them

The point of profitability analysis is not simply to eliminate bad offerings.

It is also to identify where the business should concentrate its limited time, money, people, and marketing attention.

A Practical 30-Day Action Plan

After completing the analysis, avoid turning it into a report that sits unused.

Choose three actions.

For example:

Week 1: Confirm the costs and labor requirements of the two questionable services.

Week 2: Test new pricing or packaging for one underperforming service.

Week 3: Increase promotion of one high-margin, high-demand offering.

Week 4: Compare the early results and determine whether further changes are justified.

Small improvements repeated consistently are usually more manageable than restructuring the entire business at once.

Using BizClearAI to Turn the Analysis Into a Plan

Once you understand which products or services deserve attention, the next challenge is deciding what to do about them.

BizClearAI can help small-business owners take their own business information and work through practical next steps, such as creating a customized profitability review, pricing checklist, service package, operating procedure, customer script, marketing plan, or 30-day improvement strategy.

The goal is not simply to tell you that one service looks better than another.

It is to help turn the analysis into specific actions you can test in your actual business.

Frequently Asked Questions

Can AI determine which products are most profitable?

AI can help compare profitability when you provide reliable sales and cost information. It can calculate or organize margins, labor requirements, customer demand, repeat business, and other factors, but it cannot know costs or business conditions you have not provided.

What information should I give AI for a profitability analysis?

Provide revenue, selling price, direct costs, labor requirements, owner time, number of sales, customer demand, repeat business, operational complexity, complaints or rework, and any strategic value associated with each offering.

You do not need perfect data to begin, but identify which numbers are estimates.

Should I drop a service if it has a low profit margin?

Not automatically.

First determine why the margin is low. The problem might be pricing, labor efficiency, material costs, excessive customization, travel, or other issues that can be corrected.

Also consider whether the service attracts new customers or leads to more profitable work.

How can AI help a service business analyze profitability?

AI can compare different service types based on job revenue, labor, materials, travel time, owner involvement, customer demand, repeat business, and operational complexity. It can then help identify which services appear strongest and which deserve further investigation or changes.

Can AI help me decide whether to raise prices or discontinue a product?

Yes. AI can help model scenarios such as increasing prices, changing package sizes, adjusting minimum orders, reducing delivery costs, or eliminating an offering.

The decision should still be validated against actual customer behavior and financial results.

How do I know whether a popular service is actually worth keeping?

Look beyond how often it sells.

Evaluate how much profit it contributes, how much time it consumes, whether customers return, how difficult it is to deliver, whether problems or rework are common, and whether it supports other profitable services.

How often should a small business analyze its products or services?

A full review once or twice a year may work for many businesses, with additional reviews when costs, customer demand, staffing, pricing, or business conditions change significantly.

Final Takeaway

The products or services generating the most revenue are not automatically the ones creating the strongest business.

A better analysis considers:

• Profit

• Direct costs

• Labor

• Owner time

• Customer demand

• Repeat business

• Operational complexity

• Strategic importance

AI can make that analysis easier by organizing the information, highlighting patterns, identifying gaps, and comparing possible changes.

The most useful outcome is rarely a simple list of winners and losers.

It is a clearer understanding of what to grow, what to fix, what to investigate, and what may no longer deserve your limited business resources.