HENRY FORD'S DILEMMA
Every two dollar spent in Marketing is waste but which one?
8/12/20265 min read
Henry Ford, the celebrated pioneer of mass production, coined phrases that have become legendary over the years. My favorite is this one: “For every two dollars I spend on advertising, I throw one away; I just don’t know which one.”
He was referring to a long-standing marketing problem: the difficulty of measuring the return on brand investment. When you run a campaign and sales follow immediately, you can tell if the investment was sound. However, when you invest in branding, the payoff isn't immediate; at best, it is deferred—meaning part of the result is immediate, while another part unfolds over time (if it materializes at all). And it is that latter part where it’s hard to determine whether the investment was truly worth it.
When digital marketing emerged, I was among the first to claim that Henry Ford’s dilemma was coming to an end. I believed digital measurement would be so precise, complete, and comprehensive that we would know the exact return on every penny invested. I was sorely mistaken.
It’s not that this advancement in measurement isn't real; it is a fact. But it is also a superficial one. If you invest in your brand, you can simply conduct a recall survey before and after the investment. If perception improves or deepens following the campaign, it’s a success; otherwise, it’s a failure.
But that was back in the 1990s. Since then, the communications landscape has shifted from a handful of channels to hundreds of thousands. The undeniable impact TV once had on perception became diluted, messages faced increasing levels of noise, and the consumer's reality began to speak louder. A brand cannot boast about quality and satisfied customers while facing a list of complaints on social media and consumer advocacy sites like Reclame Aqui. You can no longer make promises, fail to deliver, and get away with it.
Yet, the challenge doesn't stop there. We have increasingly grown accustomed to more precise, detailed data regarding what happens in both sales and branding. A brand recall survey today won't objectively reflect a campaign's results—partly because most companies lack the budget to invest in a campaign with that kind of impact. We are forced to view results incrementally.
Today, gaining a complete view of the customer is a complex and often convoluted process. We’ve made great strides, but not enough. The process should be straightforward: I run an ad, track the interested party, and see if they turn into a lead, a prospect, or a customer. Then, I follow their journey to observe their behavior as a customer and identify the best moment to sell a new product or service—right?
But that’s not quite how it works. Google, through its free tools, has done more for this type of measurement than all software companies combined. Even so, integrating your Google and Meta Ads campaigns with WhatsApp, telephony, legacy systems, your website, and Analytics requires navigating an ordeal of connectors, access permissions, "fake" AI, passwords, code snippets, two-factor authentication, spreadsheets, webhooks, and more.
In short, what should be quick and easy turns into a tangle of processes and systems that complicates everything—and, worse, tends to result in delayed data. There is no shortage of promises regarding accurate, real-time data, yet the process of actually obtaining it is complicated.
This is where AI should step in, but it seems the market hasn't quite grasped that yet. I’m going to compare two AI-powered tools, and you’ll understand the dilemma.
. Temso.ai – This is one of the most fascinating tools to emerge recently. Since I started using it nearly four months ago, it has only improved—though it was brilliant right from the very first version. Why? Because it clearly shows what is happening with your brand within AI systems: who is talking about it (and who isn't), the volume of mentions across different platforms, the specific citations, and more. It generates prompts, analyzes the competitive landscape, shows where you are visible (and where you aren't), and highlights what needs work versus what is already performing well. Everything is handled with speed, clarity, objectivity, and precision. The tool keeps getting better with every update. It uses AI exactly as it should: to understand the situation and propose the right actions, while boosting speed and simplicity. The AI performs its job masterfully.
. OneClick.ai – the promise is tempting: integrating your sales funnel with AI to achieve peak campaign performance. However, the reality falls short. Native integration is limited to only a few e-commerce platforms; other integrations require Tag Manager, and the setup process is complex and time-consuming. Results are questionable (such as the WhatsApp integration, which lacks clear utility), and reports are basic and uninspiring. Campaign optimization is very rigid, focusing solely on value-based alerts (pausing or restarting the campaign once a certain value is reached). What did I expect? To integrate my entire funnel as simply as possible, link campaigns using IDs, and have an AI optimization process that showed me exactly what to do to improve results. The AI doesn't need to make decisions for me; it just needs to provide answers focus on the objectives and keep things simple. None of that happens. One essential feature missing from the tool—despite its claims of being high-tech—is the ability to improve campaign results based on acquisition performance; it seems not to even grasp what that means. Ultimately, the AI is just a marketing gimmick, as the actual work could just as easily be done using standard algorithms.
Both tools cost roughly the same, both target growth, both use AI as a key selling point, and both are recent creations—yet one delivers better-than-expected results, while the other offers nothing new.
It reminds me of the 90s, when the "Turbo" trend hit the mainstream. Suddenly, everything had a "Turbo" version: hair dryers, blenders, mixers, sound systems, and even computers. Today, "Turbo" has been replaced by AI; everything promises Artificial Intelligence but often falls short due to a lack of human intelligence. What good is an AI that doesn't know what I want or can't tell me what's wrong? What’s the point of an AI that doesn't actually understand the business but simply turns an ad on or off when a certain threshold is reached? That isn't AI; it's just parameter-based logic.
I compared two tools, but there are plenty of others making the same mistakes as OneClick, and very few that do what Temso does. In other words, the "Henry Ford dilemma" now applies to the moment you decide which tool to invest in.
The market for AI solutions is massive, yet there is a clear lack of insight into the actual needs of business operators. It’s reminiscent of the early 2000s, when CRM emerged with the same promise as ERP: process automation. The difference is that ERP processes were based on repetitive physical and legal steps, whereas CRM relied on commercial policies that varied from company to company. Banks—which love spending money on the latest trends—burned through huge budgets on solutions that ultimately only solved problems for the vendors selling them. Folks, let's put AI where it belongs: in delivery, not in the promise.