There Are No Five Fucking Ways to Improve Your ROI

Congratulations. You've Found Five Ways to Avoid the Actual Problem.

Improve your marketing.

Optimize your website.

Increase customer retention.

Automate your processes.

Use artificial intelligence.

There. Five ways to improve your ROI.

Five perfectly reasonable ideas that tell me absolutely nothing about whether your business needs any of them.

What investment are we talking about?

How does the business make money?

What is preventing it from making more?

If nobody has answered those questions, what the fuck are we improving?

There are plenty of ways to improve a return. There is no universal list of five interventions that becomes correct simply because somebody puts your logo on the presentation.

Your business might need more customers.

It might need to stop accepting the customers it already attracts.

It might need better technology.

It might need somebody to explain why it bought six systems to manage a process nobody understands.

And it might need the owner to stop treating every disappointing result as evidence that somebody else needs to work harder.

You're the Owner. That Makes This Your Problem.

You approved the agency.

You approved the software.

You approved the pricing, the structure, or the people responsible for making those decisions.

Maybe you inherited the mess. Maybe a supplier failed. Maybe the market changed. Maybe somebody misrepresented what they could deliver.

Not every failure is your fault.

But deciding what happens next is a management responsibility.

You can hold a vendor accountable. You can replace an employee who isn't performing. You can acknowledge circumstances outside your control.

You still need to understand what failed.

Because replacing the agency won't fix a pricing problem. Replacing the sales manager won't create delivery capacity. Buying automation won't decide which work the business should stop doing.

At some point, somebody has to ask:

Who the fuck is running the business?

ROI Is a Result. It Is Not a Diagnosis.

Before arguing about ROI, agree on what the number means.

For a simple project calculation, you might compare the net gain after the investment's costs with the amount invested. For an operating division, an accounting ROI measure can instead compare income with average capital assets. Those are different measurement boundaries. [1]

Name the investment. Define the income or gain. Identify the costs. State the period.

Revenue divided by advertising spend does not tell you what is left after delivering the work.

And a return percentage does not, by itself, tell you when cash arrives. Investment appraisal also needs to consider timing, risk, and the resources committed; discounted cash-flow methods explicitly account for the time value of money. [2]

Payroll does not wait for your spreadsheet to become correct.

But even a correctly calculated return leaves another question:

Why is it that number?

Is the price wrong? Is acquisition too expensive? Is delivery inefficient? Is the business carrying unused capacity? Is the owner expecting a return the market cannot support?

The number starts the investigation.

It doesn't finish it.

Your Business Might Not Need Another Customer.

Imagine a hypothetical service company. These are invented numbers, not a client's results.

It completes 100 jobs a month at $1,000 each.

At that workload, variable delivery costs are $600 per job. For this example, those costs genuinely vary with the work; regular salaried payroll is included in overhead instead.

That leaves $400 per job, or $40,000 a month, before fixed overhead.

This is contribution: revenue minus variable costs. It helps pay the costs that remain before the company reaches operating profit. It is not the same as bottom-line profit. [3]

Now the owner wants growth.

The agency brings in 20 additional jobs for $4,000 in additional advertising spend. Assume those jobs would not otherwise have arrived.

Great. Another $20,000 in revenue.

Except the company is already at its current operating capacity.

To complete those additional jobs on time, it needs more expensive subcontracting, overtime, and extra materials. The additional delivery cost is $900 per job.

Here is the incremental result:

That assumes existing overhead and the original 100 jobs are unchanged, with no later repeat business or other benefits included. Financing, taxes, and longer-term effects would require separate analysis.

The revenue target was achieved.

The agency delivered actual customers.

The company completed more work.

And the business made $2,000 less than it would have without that expansion.

Everybody did the thing management asked them to do.

Management asked for the wrong fucking thing.

Now change one assumption. Suppose the company has spare capacity and can deliver those extra jobs at the original $600 variable cost.

Same revenue. Same advertising spend.

$20,000 minus $12,000 minus $4,000 leaves a $4,000 improvement instead.

Same recommendation. Different operating conditions. Opposite financial result.

That is why I need to understand the business before telling it to buy more customers.

Your Problems Don't Respect Departmental Boundaries.

A law firm might receive relevant inquiries and still sign very few suitable clients.

Maybe intake takes too long to respond. Maybe the consultation doesn't explain the value. Maybe the offer is wrong for those prospects. Maybe the supposedly qualified inquiries aren't qualified at all.

Calling it a marketing problem doesn't establish the cause.

Neither does calling the leads shit.

Follow what happens after the inquiry arrives.

Who responds? What does the prospect need? Why do they leave? Which accepted matters can the firm actually handle well?

Now imagine sales improves by promising faster delivery than operations can provide.

Contracts increase. Overtime increases. Mistakes increase. Refunds increase.

The sales dashboard looks better while the business absorbs the consequences.

A process bottleneck limits what the wider operation can get through. Improving activity elsewhere can simply create more work waiting at that constraint. ACCA's discussion of constraints describes exactly this problem. [4]

My application is straightforward: before paying to accelerate one department, establish what happens to the work when it reaches the next one.

Your employees cannot fix that relationship by being told to become more productive.

And automating the handoff doesn't help if the handoff sends the wrong work to the wrong place faster.

You Might Be Selling the Wrong Fucking Thing.

Customers can love a service that the business cannot deliver sustainably at its current price.

They can leave excellent reviews.

They can refer their friends.

They can ask you to do more of it.

None of those things pays the difference between what you charge and what the work costs.

The answer might involve changing the price, the scope, the delivery method, or the customer segment.

It might involve retiring the offer.

But don't jump from “this service looks unprofitable” to “shut it down.”

First establish what actually disappears when you stop selling it.

An allocated share of rent is not automatically a saving. If the company owes the same rent after the service closes, the expense remains. The decision needs the revenues and costs that genuinely change, including any better use of the released capacity. [5]

“Unprofitable after overhead allocation” and “the business would make more money without it” are not interchangeable conclusions.

The correct question is what changes economically when the decision is made.

That applies to expanding, cutting, outsourcing, automating, and doing nothing.

Your Market Doesn't Give a Fuck About Your Growth Plan.

Suppose delivery works. Prices make sense. Customers are satisfied. Management understands the numbers.

Now you want to double revenue.

Fine.

Where are the customers?

The SBA's market-research guidance asks about demand, market size, location, saturation, and what customers pay for alternatives. Those questions help establish whether the opportunity exists. [6]

Your growth target doesn't answer them.

A specialized business can operate in a city of millions and still serve a narrow market. Reaching the next group of buyers might require a different offer, a wider service area, or more expensive acquisition.

You have a forecast. What evidence supports it?

And please stop pointing at your competitor's advertising budget as if that settles the question.

Do you know its margins? Its cash reserves? Its customer mix? Its operating capacity? Its actual results?

You can see the billboard.

You cannot see the economics from the fucking freeway.

The competitor might have an advantage you don't have. It might be funding an experiment you cannot afford. It might be making a terrible decision.

Copying the activity doesn't tell you which one.

This Is Why I Examine the Business Before Recommending the Service.

If I sell advertising, I have an incentive to recommend advertising.

If I sell SEO, I have an incentive to recommend SEO.

If I sell automation, I have an incentive to recommend automation.

That doesn't make the recommendation wrong. It makes the diagnosis necessary.

I want to understand how the business works beyond the presentation management gives about it.

Follow a real transaction.

How did the customer arrive? What was promised? What was sold? Who delivered it? What did delivery require? When was the invoice paid? What happened when something went wrong?

Then compare the transactions that appear healthy with the ones that don't.

What changes by service, customer type, acquisition source, salesperson, or delivery team?

Where does work wait? Where does the company redo it? Where are employees compensating for a decision management refuses to revisit?

Reconcile that operating picture with the financial records. A reassuring average can hide a part of the business consuming what another part earns.

The investigation determines the recommendation.

Sometimes the recommendation is a better acquisition system. Sometimes it is a narrower offer, better intake, different scheduling, clearer authority, or less unnecessary software.

Sometimes it is to stop buying the service I could have sold you.

If every investigation ends with the same package, I want to know whether anybody investigated anything.

You Might Be the Bottleneck.

Let's return to the owner.

What if the business has competent employees, a viable offer, and enough demand—but every meaningful decision waits for you?

You approve every proposal.

You change priorities halfway through the work.

You override the specialist you hired because you wanted somebody who knew more than you.

You demand accountability while withholding authority.

Then you wonder why the team cannot move faster.

Before calling that a staffing problem, look at how long work sits in your inbox. Look at what happens after you've already approved it. Look at how often employees have to start again because the direction changed.

That is something you can investigate.

“My employees lack initiative” is something you can keep repeating.

The opposite failure is possible too. You hand advertising to the agency, sales to the sales manager, delivery to operations, and the numbers to the accountant.

Then assume somebody is connecting the decisions because everybody has a job title.

Who is responsible for deciding whether the whole arrangement works?

You don't need to configure every campaign or personally prepare the accounts.

You do need clear expectations, decision authority, and somebody responsible for reconciling the result.

Delegation is necessary.

It does not eliminate management.

Yes, Sometimes the Obvious Fix Is the Right Fix.

A broken contact form may need fixing.

A duplicated subscription may need canceling.

A business with profitable spare capacity and reachable demand may need more advertising.

You do not need a six-month consulting engagement to discover that nobody answers the fucking phone.

And a useful list of common mistakes can help you decide where to look.

My objection is turning those starting points into prescriptions without checking whether they fit.

Diagnosis should be proportionate to the decision. Make a clear, low-risk correction when the evidence supports it. Investigate more carefully when the commitment is expensive, difficult to reverse, or capable of disrupting the business.

You will never eliminate uncertainty.

You can stop pretending it disappears when somebody puts five recommendations on a slide.

What Would Prove My Diagnosis Wrong?

Take the hypothetical service company.

I suspect the next group of jobs loses money because delivery capacity is constrained.

What would change my recommendation?

Evidence that those jobs can be delivered at a lower cost, at the required quality, without displacing more valuable work. Or evidence that a tested price change or credible later contribution covers the extra cost.

Then the case for acquiring them changes.

If I suspect slow intake is losing suitable clients, define suitable before testing faster intake. Look at completed consultations and accepted work, not just how quickly somebody clicks “contacted.” Account for changes in the inquiries arriving during the test.

If I suspect owner approvals are delaying work, delegate a defined class of decisions within agreed limits. Compare waiting time, errors, and financial consequences. If delays remain elsewhere, stop blaming the owner for that particular bottleneck.

State the proposed cause. Identify the evidence that would contradict it. Choose a test small enough to survive being wrong and long enough to observe the outcome that matters.

Agree on the spending limit, the person responsible, and the point at which the business continues, changes course, or stops.

An improvement after a change is useful evidence. It isn't automatic proof that the change caused it. Demand, staffing, pricing, and customer mix may have moved too.

The investigation is allowed to reject my recommendation.

It is allowed to reject yours.

That is what makes it an investigation.

The Five Fucking Ways

All right.

You want five ways to improve your ROI?

Here you go.

1. Understand your business.
2. Understand your business.
3. Understand your business.
4. Understand your business.
5. Understand your fucking business.

That's the list.

Understanding alone doesn't produce profit. You still need decisions, execution, resources, and results.

But if you haven't established what is wrong, why it matters, and how the proposed change could improve it, what exactly are you approving?

Sometimes the guess works.

Sometimes you spend a fortune discovering that the problem was sitting in plain sight.

Sometimes it was sitting at the head of the table.

The Problem Might Be You. Fix That First.

You don't need every answer.

You do need to know what the business must answer before spending its money.

If marketing is the problem, fix the marketing.

If sales is the problem, fix the sales process.

If delivery is the problem, fix the operation.

If the economics don't work, confront the economics.

And if your decisions prevent the organization from doing what it needs to do, another consultant cannot compensate indefinitely for your refusal to change them.

You are allowed to be wrong. You are not entitled to remain wrong at the company's expense.

Ask what the recommendation addresses. Ask what supports it. Ask what happens if it is wrong.

Then be willing to listen to the answer.

Because there are no five fucking ways to improve your ROI.

There are problems to identify.

Decisions to make.

Consequences to measure.

And responsibilities you don't get to escape.

Understand the fucking business.

Then follow the money.