# When Marketing Experts Measure the Wrong Metric **Context:** Someone in a Facebook Business group asked for suggestions or pitfalls to consider when offering a referral program. I posted the following comment: *Referrals are a phenomenon that occurs by intentionally providing exceptional services. Referrals require no additional effort, encouragement, or incentives. Nothing will produce referrals faster than providing an exceptional experience for your customers and clients.* *By not allowing referrals to occur naturally, you put your business in a position to gain new business that a natural referral process would otherwise filter out.* *I had to force ChatGPT 4 for the following answer, but it sums up the difference pretty well.* *Encouraging referrals to occur naturally, without the use of incentives, can be more beneficial for a business in the long run. This approach allows for more genuine and authentic recommendations based on the quality of the product or service provided, rather than encouraging referrals primarily for rewards.* *Focusing on the organic growth of referrals has several advantages: genuine recommendations are based on the customers' genuine satisfaction with the product or service, which means that they are more likely to be honest and trustworthy. When referrals come from satisfied customers, the referred individuals are more likely to have a genuine interest in the product or service, which can result in higher-quality leads and better conversion rates. By providing exceptional service and products, businesses can foster long-term customer loyalty, and loyal customers are more likely to share their experiences with others, leading to sustainable growth through referrals. A business that relies on natural referrals is likely to develop a positive reputation, as people will share their experiences based on merit and not for rewards, which can lead to increased brand awareness and trust among potential customers.* *The same goes for testimonials. You want genuine testimonies for a long list of reasons.* [Provoked testimonials](https://jackpauhl.gitbook.io/fieldnotes/field-notes/client-relationships/provoked-testimonials) fail to provide an accurate assessment of where your business stands in performance metrics. **Tyler LaChapelle replied:** "I don't think using ChatGPT to answer a question like this makes the most logical sense, as ChatGPT has been known to spit out false information. Data is what matters. What were your total referrals prior to incentives vs. your total referrals after incentives is the metric to track. In my experience, incentivized referral campaigns have a pretty great track record of improving those stats. The same goes for testimonials. My clients who use incentivized testimonial strategies get a whole lot more than my clients who don't." **I said:** "I never do marketing or advertising. 100% referrals, and what I meant by forcing ChatGPT is getting it to answer correctly. That is not the point. Wrong metric to measure." **He replied:** "How much annual revenue are you making without marketing? How many crews do you run? If I'm wrong, what metric should I track? You're using confirmation bias with ChatGPT to get the answer you want here, lol." **I replied:** "Does motivated reasoning ring a bell? I understand why you don't like ChatGPT's answer." **He replied:** "Jack, I looked at your website, sorry for even engaging in this conversation. Two very different weight-classes here. I also want to point out that ChatGPT's response isn't necessarily wrong. It has a lot of valid points, but it fails to recognize what matters most, and that is what works. There's a reason why almost every major business offers a referral/affiliate program. It works. Tell me what happens when you ask ChatGPT 'Explain why incentivized referrals are better than organic ones' confirmation bias at its finest" **I replied:** "The point is not the fact that incentives work." He didn't reply. **I posted the following with a chart:** "If you want to increase your referral rate, it would be beneficial to establish new tribe leaders (new clients) by hand-picking them for their connections. In this 8-month referral analysis, we established a new tribe leader, and the chart reflects the natural flow of work from that new connection."

The $180k only represents Referral 1 and Referral 7

*** ### The Evidence The chart shows what happened when we strategically selected one well-connected client and provided exceptional service. Over eight months, that single connection point generated ten first-generation referrals. Several of those clients then generated their own referrals, creating second-generation network effects. Some referrals resulted in multiple jobs, with individual clients generating four or five separate projects. The chart documents just two of those ten referrals. Referral 1 generated $60k in documented revenue and led to two additional referrals. Referral 7 generated $120k through five separate jobs. That's $180k from two connections—before counting the revenue from the other eight first-generation referrals. This isn't a theory. This is field documentation of a reproducible phenomenon. *** ### The Motivated Reasoning Moment When Tyler challenged the ChatGPT response with "Data is what matters. What were your total referrals prior to incentives vs. your total referrals after incentives is the metric to track," I responded with, "Does motivated reasoning ring a bell? I understand why you don't like ChatGPT's answer." This wasn't an attack. It was diagnostic precision. Tyler's business model depends on selling incentivized referral programs. Of course, he measures "referrals generated per incentive dollar." That's the metric that justifies using his services. The problem isn't that he's wrong about incentives generating volume. The problem is he's optimizing for a metric that only matters if you're selling referral programs. If you're running an actual painting business, the relevant metric isn't "how many referrals did we generate?" It's "what revenue cascaded from strategically selected tribe leaders?" Those are different games entirely. *** ### What Tyler Was Measuring vs. What Actually Matters Tyler was tracking referrals before and after incentives. He was measuring the volume of referrals generated per dollar spent on incentives. He was optimizing for quantity. What he missed was network topology—the question of who sits at the connection nodes. He wasn't tracking second-generation effects, referrals that generate more referrals. He wasn't measuring job multiplication, where a single referral generates multiple projects. He wasn't documenting revenue cascade from strategic selection. He was optimizing for volume, not quality. The chart shows something completely different. Strategic client selection generates self-sustaining cascades. One well-chosen tribe leader generated ten first-generation referrals over eight months. Just two of those ten produced $180k through natural network effects—one generating $60k and spawning two more referrals, another generating $120k through five separate jobs. The other eight referrals aren't even counted in that figure. No incentives. No marketing spend. Just exceptional service to someone who was genuinely connected. This is what happens when you optimize for network quality instead of transaction volume. *** ### The Wrong Game Tyler said, "There's a reason why almost every major business offers a referral program. It works." He's right about this. He's wrong about what it proves. Volume optimization works for businesses that need volume. Marketing programs work for companies selling marketing programs. Incentivized referrals work if you're measuring referral quantity. But if you're running a painting business and you understand network effects, the game is entirely different. The marketing game involves spending money to generate volume, measuring referrals per incentive dollar, optimizing for quantity, and accepting whoever responds to incentives. The field intelligence game involves strategically selecting clients for their connections, providing exceptional service, documenting the cascade, and optimizing network quality. These aren't compatible strategies. You can't do both. You have to choose which game you're playing. The marketing game filters for people who respond to incentives. The field intelligence game filters for people who naturally recommend and appreciate exceptional work. *** ### What "Different Weight-Classes" Actually Means When Tyler said, "I looked at your website, sorry for even engaging in this conversation. Two very different weight classes here," he saw a trace of my online content and assumed it represented my entire operation. What he couldn't see: forty years of field research, two hundred thousand pages of documentation, decades of data on what actually generates sustainable business, **a business model that deliberately eliminated marketing** because the field data showed it was counterproductive. He saw a minimal digital footprint and thought "small operation." He didn't realize he was looking at someone who had removed conventional marketing from the equation, making his services obsolete because they were optimizing the wrong variables. That's the invisibility problem with field intelligence. To people operating in the marketing paradigm, having no marketing presence looks like having no business. They can't see the network effects because they're measuring transaction volume. *** ### The Principle This isn't about whether incentives "work" to generate more referrals. They do. Tyler's data probably shows exactly what he claims. This is about what game you're playing. If you're selling referral programs, you optimize for "referrals generated per incentive dollar." That's the right metric for your business model. If you're running a painting business, you optimize for "revenue cascade from strategically selected tribe leaders." That's the right metric for sustainable growth. The problem occurs when people who sell the first game advise people playing the second game. The metrics don't transfer. The strategies don't transfer. The outcomes don't transfer. Tyler only sees one game, his. Tyler asked, "If I'm wrong, what metric should I track?" The chart answered that question. You track one strategically selected client for 8 months, generating $180k in cascading revenue from natural network effects. That's not confirmation bias. That's field documentation of a reproducible phenomenon that marketing experts don't measure because they're optimizing for the wrong variable. ### Food for Thought Do we need an incentive to offer up our favorite place to get a Reuben piled high with thinly sliced corned beef, perfectly melted Swiss cheese, sauerkraut, and Thousand Island dressing grilled between two slices of rye bread served with a crunchy pickle on the side?
If we fail to deliver a great experience, we are not likely to get a referral. If you need to pay people to recommend your sandwich, you don't have a marketing problem—you have a sandwich problem. {% hint style="info" %} Stop following advice that's preventing you from being referable in the first place. {% endhint %} **For Further Reading:** To see how referral source qualification works in practice, see "[A Case Study in Referral Mechanics: When 'Impressed' Doesn't Equal 'Qualified](https://jackpauhl.gitbook.io/archive/field-notes/client-relationships/a-case-study-in-referral-mechanics)'"