01The Direct Answer
We have a site, SEO, media, CRM and content. Why is there still no predictable pipeline?
Direct Answer
Because a predictable pipeline is a property of the agreement between the pieces, and the agreement was never bought — it has no supplier, because it has no invoice.
Each purchase was made in a different year, for a different reason, from a different supplier, and each was briefed by whoever happened to be responsible at the time. So the site describes one customer, the media targets another, the CRM has stages that match a third, and the content speaks to a fourth. Every piece works. What does not exist is a single definition all of them were built against — and predictability is exactly the thing that requires that definition.
02What Each Piece Answers
Five local questions, well answered
It is worth writing this out, because the list is the strongest evidence that the problem is not any individual purchase.
The Local Questions
Site
Answers: does the company look credible and can someone contact us? A modern site answers that well, and it answers nothing about who should be arriving.
→
Search
Answers: are we findable for these terms? It does not decide whether these are the terms a buyer with budget uses — that decision came in as an input.
→
Media
Answers: can we buy attention efficiently towards this event? It optimises towards whichever event it is given, and it is almost always given the form submission.
→
CRM
Answers: where is each deal and what was agreed? It records the process that exists. It does not create one, and it will faithfully digitise an undefined one.
→
Content
Answers: is there material to publish on this subject? Whether the subject is the one a buyer needs before deciding is, again, an input somebody supplied.
Notice the repeated phrase. Four of the five received a decision as an input — and in most industrial operations nobody can say who made that decision, when, or on what basis. It was made implicitly, five times, by five different suppliers doing their best with what they were told.
03The Four Answers Test
Ask the same question to four people this week
The Test
Send the same message, separately, to whoever built the site, whoever runs the media, whoever implemented the CRM and whoever runs sales: "in one paragraph, who is our ideal customer and which application matters most?"
Do not brief them first and do not explain why. Four different answers is the normal outcome, and it is also the diagnosis: you own four internally coherent systems that were built against four different customers. Nobody was careless. Each answered from what they were given, and what they were given came from a different conversation in a different year with a different person.
The test also gives you something a report cannot: it shows which of the four answers is closest to the customers that actually generated good orders last year. That comparison is usually more uncomfortable and more useful than any of the four answers on its own. [EVIDÊNCIA NECESSÁRIA: uma conta real em que este teste foi aplicado — as quatro respostas obtidas e o quanto elas divergiram entre si e da carteira que gerou margem. Sem isso, apresentar como procedimento, não como achado.]
04The Single Document Test
Did every supplier receive the same page?
The second test is faster and it explains the result of the first. Ask yourself whether there exists one document — approved by whoever runs the business, not written by a supplier — that every one of those suppliers received before starting.
The common case
Each supplier ran its own briefing
Each arrived, asked good questions, took notes and built a picture of the customer. Five suppliers, five pictures, five internally consistent implementations. The divergence is not a mistake by any of them — it is the arithmetic of five independent briefings.
What is missing
A specification the company owns
Not a briefing, which collects what the company already knows how to say. Not onboarding, which is access and calendar. A written statement of who is bought from, why, in which words, and what counts as a good inquiry — approved internally and handed to every supplier as an input rather than reconstructed by each of them as an output.
The absence of that document is not a failure of organization or discipline. It is that nobody sells it, so nobody buys it, so the work of producing it never gets scheduled — while every individual purchase has a supplier lining up to make it easy.
05Why One More Piece Does Not Close It
Each addition adds a boundary, not only a capability
The reasonable instinct at this point is to add the piece that is missing — automation, a new channel, an intelligence tool, a specialist agency for the one front nobody covers. Sometimes that is right. But it is worth seeing what an addition does to the arrangement as a whole.
What an Addition Really Adds
One more local target
The new supplier will be measured by a number it controls, and it will pull the arrangement towards that number, exactly as the others do.
→
One more translation boundary
Something has to be handed from the new piece to the existing ones. Every handover is a place where context is lost, and the number of handovers grows faster than the number of suppliers.
→
One more implicit answer
The new supplier will also form its own picture of the ideal customer, because it needs one to work. That is now five pictures instead of four.
→
One more coordination load
Paid in the hours of whoever inside your company has to keep all of them consistent — usually one person, usually already the busiest one.
This is why the arrangement can keep growing while the result stays flat. Capability accumulates and coherence does not, because coherence is not something any of the pieces can supply.
06What the Spending Bought
This is not an argument for throwing anything away
The stack you bought is capability, and capability is real. The site exists and can be changed. The pages that rank are earned positions that a competitor would need time to take. The media accounts have history, which is what makes any future campaign start faster. The CRM has the record of who bought and who did not. The content is published and is being read.
What none of it bought is the sentence that all of it should have been built against. That sentence is cheap compared to any of the pieces, it does not require replacing anything, and once it exists most of the stack can be pointed at it without being rebuilt — a page rewritten, a form changed, a campaign retargeted, a CRM field added. The expensive part was never the correction. It was buying five times and specifying zero times.
07The Next Question
From diagnosis to what to keep, fix and drop
This page closes the diagnosis and stops there on purpose. Deciding which parts of the stack to keep, which to correct, which to end and which are missing altogether is a different piece of work, and doing it before the specification exists just reproduces the same buying pattern with more information.
The question that comes next is the recovery one — what to do with everything already hired when the feeling is that something is still missing. It belongs to a different cluster and will be linked here once it is published.
What comes before it, in every case, is the specification itself: who is worth selling to, in which words, what counts as a good inquiry, and what has to come back. Producing it, approving it internally and handing it to every supplier as an input is the Prebound Marketing method — the layer that has no invoice, and the only one that would have made the five purchases add up.
Prebound Marketing is cheaper before.
08FAQ
FAQ
We have a site, SEO, media, CRM and content. Why is there still no predictable pipeline?
Because predictability is a property of the agreement between the pieces, and the agreement was never bought. Each purchase happened in a different year, from a different supplier, briefed by whoever was responsible at the time. So the site describes one customer, the media targets another, the CRM has stages matching a third and the content speaks to a fourth. Every piece works on its own. What is missing is a single definition all of them were built against.
What can explain a complete stack producing nothing?
The most common is divergent specification: five suppliers, five independent briefings, five different pictures of the customer. Next is the missing return path — nothing about what happened after the inquiry reaches whoever produces the next campaign, so every campaign is a first campaign. Third is optimisation targets that compete, where each supplier is pulled towards a number it controls and the sum drifts away from the buyer you want. All three can run at once.
With five suppliers, how do I find out where the problem actually is?
Send the same question separately to whoever built the site, runs the media, implemented the CRM and runs sales: in one paragraph, who is our ideal customer and which application matters most? Do not brief them first. If the four answers differ, the problem is structural and no single contract contains it. If they converge and the result is still flat, then the diagnosis moves to what happens after the inquiry arrives, which is an execution question and can be inspected.
How do I prove the specification is divergent without commissioning an audit?
For divergent specification: the four written answers, plus the list of customers who actually generated good margin last year, compared against each answer. For the missing return path: whether loss reasons exist as a closed list and whether they reach whoever writes the campaigns. For competing targets: what event each supplier is optimising towards, written down side by side. All of it is inside your company and none of it requires a new tool to collect.
Should we throw the stack away and start again?
No, and starting again would repeat the pattern that produced this. What you bought is real capability: a site that can be changed, earned search positions a competitor would need time to take, media accounts with history, a CRM holding the record of who bought and who did not. Once the specification exists, most of that can be pointed at it without being rebuilt — a page rewritten, a form changed, a campaign retargeted, a field added. The expensive part was buying repeatedly and specifying never.
Would hiring one agency to run everything solve the coordination problem?
It reduces the number of handovers, which is a genuine gain, and it concentrates the picture of the customer in one supplier instead of five. It does not produce the specification — it only means one supplier now forms the implicit answer instead of several. The improvement is real and it is fragile: it lasts exactly as long as that contract does, and the day it ends the picture leaves with it. A specification you own survives that; a consolidated supplier does not.
Who inside the company should own this specification?
Whoever can approve it and make it stick, which in most industrial companies means whoever runs the business together with whoever runs sales. It is not a marketing document and it should not be delegated to whoever manages the suppliers, because the decisions in it — which customer is worth having, which application to pursue, what to decline — are commercial decisions with consequences in production and in pricing. A supplier can help write it. It cannot own it.
How long does it take to see a change after the specification exists?
Two different clocks, and confusing them is what produces premature switches. The composition of what arrives can change quickly, because a form, a page and a campaign target can be changed in days — you can read that shift in the inquiries themselves. Revenue moves on the clock of your own commercial cycle, from first inquiry to signed order, which in industrial sales is long. Agree at the start which of the two you will be reading, and when, or the argument returns.
Specify once, point the stack at it
Nothing here needs to be thrown away. What is missing is the sentence every piece should have been built against — who is worth selling to, in which words, what counts as a good inquiry, what has to come back. Producing it and keeping it with the company is the Prebound Marketing method.
