The Revenue-per-Dollar Trap: Why Your Easiest Budget Cut May Be the Wrong One

Lauren Turner, CustomerCentrx 8/31/26

You’re a CFO or CMO staring at a budget that needs to come down by $2 million.

You have Demand Gen, Customer Marketing, Customer Success, Customer Advocacy, Community and Lifecycle Marketing all competing for resources. Every leader has a deck explaining why their function is strategically important. Everyone insists their work affects growth or retention. You need to make an actual decision.

So you ask a perfectly reasonable question:

How much revenue does each function generate relative to what we spend on it?

Demand Gen has sourced pipeline and closed-won revenue. Lifecycle Marketing can show emails, clicks, product actions and upgrades. Customer Advocacy talks about relationships, references and champions. Community has engagement data. Customer Success points to retention, although Product, Support and Customer Marketing are also claiming some influence over the same renewals.

The spreadsheet starts making the decision for you.

The functions with a short, observable path to a transaction look productive. The functions whose effects are distributed across customers, departments and time look expensive.

Cut the latter.

There’s just one problem: you aren’t necessarily comparing economic value. You’re comparing ease of attribution.

And those are not the same thing.

The Funnel Is a Useful Model. It Just Isn't a Model of Everything.

Most CMOs learned to measure marketing through some version of a conversion funnel, and for good reason.

Spend money on an activity. Generate a response. Convert the response into an opportunity. Convert the opportunity into revenue. Compare what went in with what came out.

The model has become considerably more sophisticated over the years, but its fundamental appeal hasn't changed: there is a relatively short causal chain between an investment and an economically observable event.

That makes it enormously useful for evaluating transactional activities.

It also creates a problem when the same logic becomes the default standard for evaluating everything Marketing or the broader customer organization does.

Customer relationships don't behave like funnels.

Consider a $500,000 renewal. Over the previous year, the customer's CSM identified an adoption problem and helped resolve it. The customer participated in a community where peers helped them get more from the product. Customer Marketing kept them engaged with relevant education and events. They joined a customer advisory board and raised a product issue. VOC helped surface that issue internally, Product fixed it, and an executive sponsor maintained a relationship with their leadership team.

Which function generated the $500,000?

There isn't a good answer, because that's the wrong question.

Several capabilities may have changed the probability of renewal. None independently generated the revenue.

Force each function to answer the question anyway, however, and something predictable happens. Customer Success claims retained ARR. Customer Marketing claims influenced revenue. Community shows that engaged members retain at higher rates. Advocacy claims revenue associated with participating customers.

Pretty soon $500,000 of actual revenue can appear as $1.5 million or $2 million of “influenced” revenue across internal dashboards, because each function is claiming credit for the same $500,000. Then Finance looks at the numbers and concludes that marketing attribution is bullshit.

Finance isn't wrong.

But demanding that every function produce an even cleaner revenue number doesn't solve the problem. It can make it worse.

When Accountability Rewards Bad Measurement

Suppose Customer Advocacy costs $500,000 annually.

A conservative analysis suggests the program contributed perhaps $900,000 of economic value. A much broader definition of “influenced revenue” allows the team to report $6 million.

Now tell the program leader that anything contributing less than $5 million is likely to be cut.

Which methodology do you think is going to appear in the next QBR?

This doesn't require anyone to be dishonest. It requires them to respond rationally to the incentive you created.

The same thing happens before programs are funded. Leadership decides a proposed program must produce $4 million within twelve months to justify the investment, so—remarkably—the business case forecasts slightly more than $4 million.

The number didn't emerge from the model. The model emerged from the number.

There is a particularly nasty feedback loop here. Finance distrusts inflated attribution, so executives demand harder ROI proof. Programs need increasingly large financial claims to survive, so practitioners adopt increasingly generous attribution methodologies. The resulting numbers become less credible, which makes Finance demand still more proof.

The accountability mechanism begins degrading the very measurement integrity it was supposed to create.

But bad measurement isn't the most important consequence.

Eventually, it starts affecting where you invest.

Your Measurement System Is Also a Capital-Allocation System

Imagine that you have $500,000 to allocate to one of two programs.

The first is an automated post-sale lifecycle program. Customers receive behavior-triggered communications, click through, take actions in the product and eventually upgrade. Within a few months, you have something resembling a funnel.

The second is Customer Advocacy. The team identifies customers with the potential and inclination to advocate, develops those relationships, and creates opportunities for them to participate. Over time, those customers may provide references, refer prospects, speak at events, leave reviews, participate in advisory boards, share expertise, provide product insight and advocate for the company when they move to another employer.

Those behaviors can affect acquisition, conversion, sales velocity, retention, expansion, product decisions and reputation. The effects cross departmental boundaries and may unfold over several years.

At the end of Q1, ask both leaders, “What revenue did you generate?”

One of them has a much better answer.

That doesn't necessarily mean it created more economic value. It means it has a shorter measurement distance from activity to transaction.

When organizations consistently prefer the investments whose economic effects are fastest and easiest to observe, they create what we might call measurability bias: capital allocation begins favoring what can be measured most easily rather than what creates the most value.

That matters enormously as companies automate more customer interaction. AI and lifecycle automation can create real value. They can reduce costs, improve responsiveness, drive adoption and allow humans to focus on higher-value work.

But efficiency and relationship aren't interchangeable.

A chatbot can answer a customer's question. An automated campaign can prompt a customer to use a feature. Neither necessarily creates the relationship in which that customer is willing to spend an hour helping your salesperson close a deal, put their company's logo on your website, candidly tell your product team what it doesn't want to hear, introduce a peer to your company, or bring your product with them when they change employers.

Those behaviors require something more difficult to automate and considerably more difficult to attribute: trust.

And we are starting to get better evidence that those behaviors have substantial economic value.

We Know the Harvest Has Value. What About the Cultivation?

A recent Harvard Business Review article by Fred Reichheld, Jamie Cleghorn and Wojtek Kokoszka, “Don’t Underestimate the Power of Customer Referrals,” examined more than 10 million consumers and found that roughly 20% of new customers came through referrals while accounting for 72% of profits generated by new customers.

That's important research. Executives should pay attention to it.

Customer-facing practitioners may be forgiven for responding with some version of: No !@#$, Sherlock. We've literally been telling you this our entire careers.

For decades, people working in Customer Marketing, Advocacy, Community, CX and Customer Success have argued that customers create economic value beyond what they personally buy. They make referrals, provide references, influence buyers, share expertise, provide product insight, advocate publicly and bring vendors with them as they move through their careers.

The HBR research helps quantify part of that value.

But it also exposes an important disconnect:

A company can recognize the economic value of an outcome while systematically undervaluing the organizational capabilities that produce it.

Referrals don't spontaneously appear in a CRM.

Before the referral, the customer had an experience. They received value. Someone helped them succeed. Someone listened to them. Trust developed. Relationships formed inside the company and perhaps with other customers. Eventually that customer became sufficiently confident in the company to attach something very valuable to a recommendation: their own reputation.

We are getting increasingly good at measuring the harvest.

The harder question is how much we should invest in cultivation.

The Savings From a Cut Are Easier to Measure Than the Cost of the Cut

This becomes particularly important when the conversation moves from investment to cost reduction.

Suppose Advocacy costs $500,000.

The CFO can eliminate it and put almost exactly $500,000 of savings into the operating plan. The amount is known. The timing is immediate. The causal relationship is spectacularly clear.

Cut salaries → save salaries.

Now consider the other side of the transaction.

What happens to reference capacity? What happens to referral generation? Do Sales teams have a harder time finding relevant customers for late-stage opportunities? Does the company need to purchase more acquisition to replace organic customer-driven growth? Does the pipeline of future advocates shrink? Does customer insight become harder to access? Do important relationships become more dependent on individual CSMs or champions?

Those effects are uncertain, distributed across the organization and likely to occur at different times. So the accounting representation of the decision effectively becomes:

Cut Advocacy: +$500,000.

But the economic reality is closer to:

Value of cut = expense eliminated − economic value lost − new costs created elsewhere.

The problem is that the first term is beautifully measurable and the others aren't.

That doesn't make them zero.

It makes them uncertain.

And uncertain value and nonexistent value are not the same thing.

Relationship Functions Have an Additional Measurement Disadvantage

Imagine a company has spent five years building a healthy customer ecosystem. It has hundreds of active advocates, strong executive relationships, an engaged community, deeply connected CSMs and trusted channels for customer feedback.

Then the company cuts much of the infrastructure maintaining those relationships.

The first year may look terrific.

Costs fall immediately, but customers don't forget five years of relationship history on the day the layoffs happen. Existing advocates still take reference calls. Champions continue recommending the product. Community members retain relationships with one another. Customers who trust the company continue behaving like customers who trust the company.

Leadership can cut the cultivation, observe that the harvest still looks healthy, and conclude that apparently the cultivation wasn't necessary.

But what if the company is actually consuming an accumulated asset?

Fewer new advocates may develop. Existing advocates may fatigue. Executive relationships may thin. Customer-to-customer connections may weaken, and the flow of candid information from customers into the company may decline. Over time, a relationship that was once multidimensional can become increasingly transactional.

Eventually that may show up in referrals, competitive win rates, expansion or retention.

But perhaps not for eighteen months. Perhaps not for three years.

By then, the practitioner who warned about the risk may be gone. The executive who approved the cut may be gone too. And whatever deterioration eventually appears may be attributed to pricing, Product, competitors, the economy or the new CMO.

This creates a profound asymmetry:

The savings from cutting relationship capabilities are immediate, certain and attributable. The potential losses are delayed, probabilistic and distributed.

A conventional ROI comparison therefore doesn't just make relationship functions look worse.

It structurally favors cutting them.

Ask a Different Question

None of this means CFOs should stop demanding financial accountability from customer-facing functions.

It means the unit of analysis needs to change.

If you are deciding where to invest another dollar, don't ask only:

“Which function generated the most revenue last year?”

Ask:

“Where is the next dollar of investment expected to create the greatest enterprise value?”

And if you're cutting:

“Where can I remove the next dollar of investment with the least expected destruction of enterprise value?”

That second question is especially important because the economics of adding and removing investment are not necessarily symmetrical.

A mature Advocacy function may not have a compelling case for another $200,000 of incremental spending. But eliminating the $500,000 required to maintain an established advocate network could destroy substantially more than $500,000 of value.

That's not unusual in business.

You don't evaluate maintenance solely by asking what incremental revenue the maintenance department generated.

You evaluate the economic consequences of maintaining versus not maintaining the asset.

Customer relationships deserve the same discipline.

What Should Replace the Revenue Leaderboard?

For every meaningful function or program under consideration, Finance and Marketing should build an inspectable investment case around the same questions: What economically valuable outcomes does this capability create, preserve, accelerate or make possible? Through what mechanisms? How strong is the evidence supporting those mechanisms? When should the economic effects become observable? What value has already accumulated because of past investment? And what is expected to happen if investment increases, remains constant, decreases or disappears?

That produces something very different from an ROI leaderboard.

Imagine Finance needs to remove $1 million and is considering reductions across Demand Gen, Lifecycle Marketing, Advocacy and Community.

Demand Gen may have highly observable short-term revenue effects and high confidence in the estimated consequence of a cut. Advocacy may have effects across references, referrals, conversion and relationship resilience, with a longer time horizon and wider confidence interval. Community may produce some combination of support efficiency, customer knowledge, adoption and retention effects, with varying levels of evidence.

Finance should see those differences.

Uncertainty isn't a reason to assign zero value. It is information that belongs in the decision.

Likewise, time shouldn't be hidden. Future economic effects can be discounted into present value, just as Finance does with other investments whose returns occur over different periods.

The objective isn't to make Community look as measurable as paid search.

It's to make the economic consequences of each decision sufficiently visible that Finance can compare unlike investments without pretending they're identical.

This Is a Higher Standard, Not a Lower One

There is an understandable fear that changing the measurement standard gives relationship-based functions an excuse to avoid accountability.

It shouldn't.

If Advocacy says references improve conversion, eventually there should be evidence that they do. If Community claims to improve retention, the company should test whether comparable customers actually behave differently. If Customer Marketing says its work deepens relationships, we should define what that means and determine whether those relationship conditions predict economically meaningful outcomes.

Some programs will fail those tests.

They should lose funding.

But there is a significant difference between concluding:

“The evidence suggests this capability doesn't create enough economic value to justify its cost,”

and: “This capability cannot show directly attributed revenue quickly enough, therefore it has little economic value.”

The first is capital allocation.

The second is measurability bias.

And for the CMO in particular, this requires resisting the temptation to make the funnel the universal operating model simply because it has served Marketing well.

The funnel isn't wrong.

It's just a model of a particular kind of value creation.

Customer relationships are another kind.

They are longitudinal, multicausal and cumulative. Their economic value can appear in acquisition, conversion, expansion, retention, efficiency, intelligence and risk reduction, often simultaneously and often long after the activity that helped create it.

Trying to force all of that into a linear conversion funnel doesn't make the organization more financially rigorous.

It makes a complex economic system easier to put on a slide.

Those aren't the same thing.

The Question for the Next Budget Meeting

When the next planning cycle arrives and every function is lined up with its revenue number, resist the easiest comparison.

Don't ask which department can attach the largest amount of revenue to its activities.

Ask what you're buying from each capability, what evidence exists that the capability changes economically meaningful outcomes, how long those effects take to appear, and what happens to enterprise value if you stop paying for it.

Then make the cut.

You may still cut Advocacy. You may cut Community. You may cut Customer Success—or Demand Gen.

A rigorous framework shouldn't be designed to save any particular function.

It should be designed to make sure the function that loses is the one with the weakest economic case, not simply the one whose value takes the longest to show up on a dashboard.

Because when the measure is “How much revenue does this function generate, and how quickly?”, relationship-based functions will almost always lose.

The more important question is whether the company loses with them.