7 de outubro de 2026
Top Posts
1
CATEGORIA • 17 NOV, 2025
2
CATEGORIA • 17 NOV, 2025
3
CATEGORIA • 17 NOV, 2025
4
CATEGORIA • 17 NOV, 2025
5
CATEGORIA • 17 NOV, 2025
stephanieamatus
Seguir
2 posts
69 seguidores
356 seguindo
Stephanie Amatus she/her
I like thinking about stuff, and sometimes I talk about them.
Diagnosing the Funnel: Do You Have a Traffic, Offer, or Conversion Problem?
Start with the system, not the tactic
When sales slow down, one of the first instincts in marketing is to look toward acquisition. Maybe we need to increase the ad budget, publish more content, expand into another channel or finally test the strategy someone has been suggesting for the past six months. It makes intuitive sense: if the business needs more customers, put the offer in front of more people.
Sometimes, that is exactly what you should do. The problem is that “we need more traffic” has become such a convenient explanation for disappointing growth that companies often start pouring more people into a funnel without first checking whether the funnel was doing anything useful with the people already there.
I’ve always found this a slightly strange way to approach marketing. If a store had a thousand people walking through its doors every day and almost nobody bought anything, our first question probably wouldn’t be how to get another thousand people through the door. We’d want to know what was happening once they got inside. Were they finding what they came for? Did they understand what was being sold? Were the prices reasonable? Did they pick something up and then put it back? Did they walk all the way to the register before changing their minds?
Digital marketing shouldn’t be any different.
Before increasing acquisition, rewriting the offer or redesigning the landing page, I want to know where people stop moving and, more importantly, why. Traffic problems, offer problems and conversion problems can eventually produce the same disappointing outcome at the bottom of a funnel, but they are very different problems to solve. Treating them as interchangeable is how businesses end up spending months optimizing things that were never broken.
A funnel is really just a way of describing a series of decisions made by another human being.
Someone discovers you, becomes interested enough to pay attention, considers what you’re offering and, perhaps, decides to do something about it. Depending on the business, that final action might be buying a product, starting a trial, booking a demo, requesting a quote or talking to someone in sales.
Marketing dashboards turn this journey into numbers, which is useful, but numbers can also make us forget what we’re actually looking at. A falling conversion rate isn’t simply a percentage moving in the wrong direction. It represents people arriving somewhere in the customer journey and deciding, for one reason or another, that they don’t want to continue.
The job of diagnosis is to find that moment.
If very few qualified people are entering the funnel, you may have an acquisition problem. If the right people are arriving but don’t seem particularly interested in what you’re selling, the offer deserves attention. If they clearly want what you’re offering but disappear somewhere between interest and purchase, then you’re probably looking at a conversion problem.
The distinction sounds obvious when it’s laid out neatly on a page. Inside a real business, where five campaigns are running at once, sales has its own opinion, the CEO wants growth this quarter and someone has already decided the website needs to be redesigned, things tend to become considerably messier.
That’s exactly why diagnosis matters.
Do you actually have a traffic problem?
The easiest way to misunderstand traffic is to confuse volume with opportunity.
You can have hundreds of thousands of visitors and still have an acquisition problem if very few of those people have any realistic chance of becoming customers. On the other hand, a business with relatively modest traffic can perform extraordinarily well when most of the people arriving have the right intent.
Imagine a company selling enterprise cybersecurity software. A video from the company unexpectedly goes viral and brings 200,000 people to the website. The marketing dashboard looks fantastic. Traffic has exploded, engagement is up and the brand has reached more people in a week than it normally reaches in months.
There’s only one inconvenient detail: most of those visitors are students and technology enthusiasts who have no authority, budget or intention to buy enterprise cybersecurity software.
Technically, the company has traffic. Commercially, very little has changed.
This is why I wouldn’t diagnose acquisition by asking only how many people are arriving. I’d want to know who they are, where they came from, what brought them there and whether their behavior resembles the behavior of people who eventually become customers.
Search traffic can be particularly revealing because intent varies enormously between queries. Someone searching for a definition is behaving very differently from someone comparing vendors, looking at pricing or searching for a solution to a specific business problem. The same applies to paid media and social content: reach from the wrong audience can make a campaign look much healthier than it actually is.
A genuine traffic problem becomes more likely when the people who do reach your offer are qualified and convert reasonably well, but there simply aren’t enough of them to support the growth the business wants.
That’s a very different situation. You already have evidence that people want what you’re selling and that the journey can turn them into customers. At that point, investing in SEO, paid media, partnerships, organic distribution, outbound or new acquisition channels can make perfect sense because you’re feeding a system that has already demonstrated that it works.
But when qualified people are already arriving in meaningful numbers and consistently choosing not to continue, buying more traffic is often an expensive way of avoiding a more uncomfortable question.
What if the problem is what we’re selling?
When the offer is the problem
This is probably one of the hardest marketing problems to diagnose honestly because nobody particularly enjoys discovering that the market is less excited about an offer than the company that created it.
So we change the ads.
We test another headline, create new visuals, adjust the targeting, rewrite the landing page, bring in another agency and start wondering whether TikTok might finally be the answer.
Sometimes those changes work because communication really was the problem. A great product can absolutely be presented badly.
But there are also situations in which the marketing is communicating perfectly well and people simply aren’t interested enough.
An offer isn’t just the product sitting underneath it. It includes the problem you’re solving, the outcome you’re promising, the price you’re asking people to pay, the effort required from them, the alternatives available and the amount of risk involved in saying yes. All of those things contribute to a customer’s perception of whether the exchange is worth making.
A product can therefore be objectively good while the offer around it remains weak. Maybe the problem it solves isn’t particularly urgent. Maybe the difference between your solution and a cheaper competitor is obvious internally but almost invisible from the customer’s perspective. Maybe the price requires a level of trust the brand hasn’t earned yet, or the company has built something people find interesting without giving them a compelling reason to buy it now.
This is where marketing sometimes gets assigned an impossible job. Instead of helping the market understand the value of a strong offer, it is expected to manufacture enough excitement to compensate for the absence of one.
Good marketing can certainly influence perception. It can make value easier to understand, give context to a problem, build trust, position a product against its alternatives and make something desirable.
What it cannot reliably do forever is make people want something they don’t value enough to buy.
If qualified people are reaching the business, understanding what is being offered and still showing very little meaningful intent, I would investigate the offer before increasing the acquisition budget. Otherwise, all you’re doing is paying to introduce the same unconvincing proposition to a larger audience.
Or maybe people want it, but something gets in the way
A conversion problem looks different because the customer has already given you evidence of interest.
They spend time on the product page. They compare plans. They start filling out the form. They add the product to their cart. They request information, begin a free trial or click the button to schedule a demo.
Then, somewhere along the way, they disappear.
This is the part of funnel analysis I find especially interesting because the reason can be surprisingly small compared with the amount of revenue it affects. A checkout reveals an unexpected shipping fee. A form asks for information the customer doesn’t understand why you need. The mobile experience is frustrating. Pricing is unnecessarily difficult to find. The page answers every question except the one people actually have before buying. A lead requests a demo on Monday and receives a response on Thursday.
None of these things require the customer to suddenly stop wanting the product. They simply introduce enough friction, uncertainty or effort to make continuing less attractive.
That’s why an overall conversion rate is useful, but it isn’t much of a diagnosis by itself. Knowing that 2.4% of visitors convert tells you the outcome. It doesn’t tell you what happened to everyone else.
Sometimes marketing isn’t where the funnel is broken
The more interesting information is usually found in the transitions.
If people arrive and leave almost immediately, I want to know whether the acquisition message matched what they found on the page. If they spend time exploring but rarely demonstrate purchase intent, I’d look more closely at the offer and positioning. If they begin checkout and abandon it at a particular step, I’d investigate that step. If they happily become leads and then disappear after speaking with sales, the investigation needs to follow them there.
The point is to stop treating “conversion” as one event and start looking at the series of decisions that precede it.
This becomes particularly important in B2B businesses because the customer journey often crosses several departments before revenue ever appears.
Imagine marketing generates 300 qualified leads in a month. They match the company’s ideal customer profile, come from relevant companies, engage with the offer and actively request conversations.
Marketing celebrates.
Then sales closes six of them.
The instinct might still be to ask marketing for more leads because more leads should, mathematically, create more customers. And technically, it might. If you double the number of leads while everything else stays the same, perhaps sales closes twelve instead of six.
But that’s a very expensive interpretation of the problem.
I’d want to know what happened to the other 294 people first.
How quickly were they contacted? Did sales agree that they were qualified? What happened during those conversations? Were prospects consistently raising the same objection? Did the promise that attracted them match the conversation they had afterward? Was follow-up consistent? Were opportunities being lost to competitors, pricing, timing or something else entirely?
Marketing can do an excellent job generating demand while the business does a terrible job capturing it.
This is why I don’t think a funnel should end at a form submission simply because that’s where the marketing dashboard ends. The customer doesn’t care where your departmental responsibilities change. From their perspective, they’re still having one continuous experience with the same company.
If the purpose of the funnel is eventually to create revenue, diagnosis has to follow the customer far enough to understand whether that value was actually captured.
Read your funnel metrics as clues, not grades
There’s a temptation in marketing to look at metrics as though they were report cards. Green means good, red means bad, and anything below an industry benchmark becomes something to optimize.
I find them much more useful when they’re treated as clues.
A high bounce rate, for example, doesn’t automatically tell you that a landing page is bad. It might mean the page attracted the wrong audience, the ad created an expectation the page didn’t fulfill, the visitor found the answer they needed immediately or the experience genuinely was terrible. The number tells you where something interesting may be happening, but you still have to investigate what caused it.
The same is true of conversion benchmarks. Knowing that another company converts at 4% while yours converts at 2% can give you context, but those numbers may come from completely different acquisition channels, audiences, price points, buying cycles and levels of brand awareness.
Benchmarks can tell you where to look. They cannot tell you what you’ll find.
Your own funnel becomes much more useful when you start looking at relationships between metrics rather than evaluating each one in isolation. What happens to visitors from different acquisition sources? Which campaigns produce customers rather than simply leads? Where do high-intent users behave differently from everyone else? Which objections appear repeatedly in sales calls? At what point does a customer who seemed interested begin behaving like someone who isn’t?
Those patterns tell you far more about the health of the business than whether a single percentage happens to be above or below an industry average.
How I would diagnose a funnel that isn’t working
If I were brought into a business tomorrow and told that the funnel wasn’t producing enough revenue, I wouldn’t begin by deciding what needed to change. I’d begin by following the customer from the moment they first encountered the company until the moment they either bought something or disappeared.
I’d want to understand how people were discovering the business and what they expected before arriving. Then I’d compare that expectation with what they encountered afterward. I’d look at which audiences stayed, which left, what interested prospects did differently, where intent appeared and where it vanished.
I’d also talk to the people closest to the customer. Analytics can tell me that someone abandoned a page; a sales conversation might tell me that customers consistently misunderstand the pricing model. A CRM can show that leads aren’t closing; looking at response times might reveal that half of them wait two days before hearing from anyone. Customer interviews may expose an objection that no dashboard was ever designed to measure.
Before you scale, find the bottleneck
One of the most expensive things you can do in marketing is scale a problem you haven’t identified yet.
More traffic sent to a strong funnel can create more revenue. The same traffic sent to an offer nobody wants creates more rejection. A larger lead generation campaign connected to a broken sales process creates a larger pile of neglected leads. A bigger advertising budget directed toward a checkout that frustrates customers simply gives more people the opportunity to abandon it.
This is why growth isn’t always about adding something.
Sometimes the highest-leverage decision is discovering what is preventing the system you already have from working properly.
So the next time a funnel isn’t generating enough revenue, don’t begin with the assumption that you need more people at the top. Look at the people who are already there. Follow what they do, pay attention to where their behavior changes and try to understand what that change is telling you.
If they aren’t arriving, you may have a traffic problem. If they’re arriving but don’t care enough, you may have an offer problem. If they care but can’t quite make it to the other side, you may have a conversion problem. And if they make it through marketing only to disappear somewhere afterward, keep following them.
The numbers will tell you where to look, but the real diagnosis comes from understanding the people behind them.
Once you know where they’re getting stuck, deciding what to optimize becomes considerably easier.
Until then, you’re mostly guessing.
How AI Is Changing SEO, Content, and Customer Acquisition
Eventually, the funnel begins to tell a story.
Maybe qualified visitors convert beautifully, but there simply aren’t enough of them. That’s when I start thinking seriously about acquisition.
Maybe plenty of the right people arrive but almost nobody demonstrates meaningful interest. Then I want to understand the offer, the positioning and whether the value proposition matters enough to the market.
Maybe people clearly want what we’re selling and repeatedly disappear at the same point. That’s a strong reason to investigate conversion friction.
Or maybe marketing is producing healthy opportunities and the real loss happens after the handoff to sales.
The important part is that the solution comes after the diagnosis, not before it.
For most of the internet’s history, finding something online followed a fairly predictable ritual. You had a question, typed a few words into a search engine, received a page full of links and started opening them until one seemed useful.
For businesses, this created an equally familiar equation. If you wanted to be discovered, you needed to appear somewhere in that journey. So we learned how people searched, built pages around those searches, fought our way up the rankings and tried to earn the click before a competitor did.
AI is beginning to interfere with that equation in a very interesting way.
Increasingly, people don’t necessarily want a list of places where the answer might exist. They want the answer.
They ask ChatGPT to compare software, use Google’s AI results to understand a complicated topic, ask Perplexity to research a purchase or have an AI assistant summarize several sources before they ever visit one of them. The information that once required ten tabs can increasingly be synthesized into a single conversation.
That doesn’t mean search is disappearing, nor does it mean SEO suddenly stopped mattering. It means the journey between having a question and choosing a business is changing, and with it, some of the assumptions marketers have relied on for years.
The interesting question is no longer simply whether you rank.
It is whether your company becomes part of the answer.
SEO is moving beyond the click
Traditional SEO was built around a relatively straightforward exchange. Someone searched for something, Google presented results and websites competed for visibility. Ranking mattered because visibility produced clicks, and clicks brought people into an environment the company controlled.
That environment was valuable.
Once someone reached your website, you could introduce the brand properly, guide them toward another page, collect an email address, retarget them later or eventually turn them into a lead.
AI search complicates this because some of that discovery now happens before the visitor ever reaches you.
From ranking in search engines to appearing in answers
This is where terms such as GEO, or Generative Engine Optimization, and AEO, Answer Engine Optimization, have started entering marketing conversations.
The terminology will probably continue evolving, and marketers will undoubtedly invent several more acronyms before we’re done, but the underlying shift matters more than what we decide to call it.
For years, SEO asked some version of:
How do we make this page rank for what people are searching for?
AI discovery adds another question:
How do we make our business understandable and credible enough to become part of the answer?
That changes how I think about content.
If you want an AI system, or a search engine increasingly powered by AI, to understand what your company knows and does, publishing vague content simply because a keyword has search volume becomes much less interesting.
Your website needs to contain clear information about your products, expertise, customers, use cases and the problems you solve. Your content needs to make relationships between ideas understandable. Claims need evidence. Expertise needs substance.
In other words, the web increasingly rewards something marketers occasionally managed to forget while chasing rankings: actually being useful.
Imagine someone researching CRM software for a growing B2B company. A few years ago, they might have searched “best CRM for B2B companies,” opened several articles, visited comparison pages and slowly assembled their own shortlist.
Today they can ask an AI system:
“What are the best CRM platforms for a 50-person B2B SaaS company with a small sales team, a long sales cycle and a budget under $2,000 a month?”
That’s not really a keyword anymore. It’s a conversation containing context, constraints and intent.
The answer may compare several companies, explain their strengths and weaknesses and narrow the market down to three options before the person has visited a single vendor.
From the customer’s perspective, this is wonderfully convenient.
From the marketer’s perspective, it means something important has happened: part of the consideration process has moved outside the website.
Your potential customer may be learning about you, comparing you with competitors and forming an opinion about your brand in an environment you don’t control.
SEO therefore isn’t becoming irrelevant. If anything, discoverability becomes more complicated because businesses now have to think about both traditional search visibility and how their information is understood, retrieved, cited and represented by AI systems.
7 de outubro de 2026
17 de agosto de 2026