The pages were not converting, and every instinct in the room said rewrite them. But the evidence needed to fix a single headline was scattered across three teams that never spoke, and the team that owned the page held none of it.

I was brought in to consult with the web marketing team at one of the largest companies in marketing software. The website drew the bulk of the company's inbound prospects, and it was underperforming.

Trial starts from the highest-traffic pages had gone flat, and the brief that greeted me was the obvious one: rewrite the copy, tighten the layout, ship a cleaner page.

Before touching a word of it, I went looking for why the pages actually failed. Two findings pointed away from the page itself.

In focus groups, prospects could not tell whether the product would solve their particular problem. They were not tripped up by the words on the screen; they simply could not map the product to their own situation.

And the traffic data showed that most of the volume landing on those pages came from paid media rather than organic search.

A conversion problem with no owner

Put those two findings together and the answer moved off my desk.

If paid traffic dominated, then the promise that pulled a visitor onto the page was written by the paid-media team, inside their ad themes. If prospects could not see whether the product fit, then the truth about what it could honestly claim sat with the product managers. And the record of exactly where visitors gave up sat with analytics.

Three teams each held one piece of the evidence I needed to fix a page none of them owned.

One funnel, three owners, one blind spot Paid media holds the ad promise that brought the click Product holds what the product can honestly claim Analytics holds where the visitor gives up Web marketing team owns the page, holds none of it
The gaps in the dashed lines are the whole story. Every piece of evidence a rewrite would need sat one team away from the people holding the pen.
The conversion problem lived in the org chart. No landing page rewrite could fix the org chart.

None of this was anyone's failure. It is the natural physics of a large organization.

Paid media optimizes for the click. Product optimizes for the roadmap. Analytics optimizes for clean measurement.

Each team does its own job well, and doing it well is what builds a wall around a different slice of the funnel.

The bigger the company, the taller the walls, because that specialization is exactly what lets a big company move at all. Deloitte Insights research on analytics-driven enterprises found that most companies take a fragmented, siloed approach to data and analytics, one that correlates with diminished business success.

I could redraft the page from my own assumptions, but without the ad themes, the product truth, and the drop-off data in one place, I would only be redecorating a room while the leak stayed inside the wall.

Thirty conversations before a single test

The move I made looks slow on paper, and it was the fastest option available.

Instead of opening a document to redraft the page, I set a multi-month goal to build a working relationship with every person who touched the funnel: roughly thirty people across product, web, and paid media.

One conversation at a time. No agenda beyond understanding what each of them knew, what they were measured on, and what they wished the web team would do differently.

This is uncomfortable to justify to yourself when a page is underperforming and someone is paying you to fix it. The pressure is to produce a visible artifact fast.

But the artifact everyone wanted, a rewritten page, was the one thing that could not work yet. The relationships were the real deliverable, and they had to come first.

Thirty relationships, then one room The conversations were the work. The workshop was the capstone. About thirty one-on-ones, over months One cross-functional workshop seven functions in one room The quarter's test roadmap
Run this room with thirty strangers and you get a polite meeting. Run it with thirty people who already trade data with each other and you get a roadmap.
Thirty coffees outrun one reorg. Influence built one relationship at a time beats a mandate nobody asked for.

I had no standing to reorganize anyone, and I would not have wanted it.

An order to share data across teams buys you the letter of compliance and none of the substance. Rob Cross, who founded the Connected Commons research consortium, maps how collaboration actually moves through companies, and the patterns often come back very different from the formal hierarchy leaders rely on.

What I actually wanted was for the paid-media analyst to think of the web team the moment she noticed an ad theme outperforming, unprompted, because we had talked and she knew it would matter to me.

How the resistance came down

The early conversations met a reasonable wariness. People were not sure this would be worth their time, and most had watched cross-functional efforts curdle into standing meetings that produced slides and nothing else. Behnam Tabrizi's Harvard Business Review research on cross-functional teams found 75% of them dysfunctional.

I did not argue the point. I brought something instead.

Every conversation, I shared a piece of what I already had: the web traffic patterns, the focus-group findings, the drop-off points I could see from my seat.

That changed the temperature. When I showed a product manager where prospects stalled on the page describing their feature, it was immediately useful to them, and they answered with what they knew about the feature the page was failing to explain.

Data flowing out of my hands pulled data back into them.

Resistance melts when the data flows both ways. Share yours first and the silo opens from the inside.
Three half-answers, one brief SILO WHAT ONLY THEY KNEW THE PAGE DECISION IT SETTLES Paid media which ad promise brought them what the page must live up to Product what it can honestly claim what the page may promise Analytics where the visitor gives up which part to fix first No team could brief the page alone. Combined, they wrote the brief.
Each row is useless by itself and decisive in company. The silo was not a communication nicety. It was the missing brief for the page.

Turning a network into a workshop

After several months, the thirty separate relationships had become a working network, and I could do something that would have been impossible on day one.

I pulled about two dozen of them, across seven functions, paid media, product, design, web marketing, analytics, copy, and web ops, into a single structured workshop.

Because the trust was already there, the room was not a cold introduction. It was a group of people who believed the exercise enough to do real work inside it.

That workshop is where the scattered evidence finally combined into a ranked set of experiments, and its output became the quarter's test roadmap.

I have written up how that session is actually run, its structure, the breakout design, and the way it turns a plan into something the room owns, in the learning agenda.

The part worth keeping here is the sequence. The workshop worked because the relationships came first, and no amount of facilitation skill can manufacture that on the day.

This all feeds the broader growth operating system, whose planning machinery decides which experiments to run, while the weekly growth review keeps them honest once they are live.

But none of that machinery turns over inside a large company until the people holding the evidence are willing to put it on the same table. Breaking the silo is what makes the rest of the system possible.

Build your own version

A reorg is not required to pull this off. If the evidence that would fix your funnel is spread across teams that rarely talk, the work is relational before it is analytical. The steps:

  • Find where the evidence actually lives. Ask which team knows the ad promise, which knows the product truth, and which knows where prospects drop off.
  • Accept that the fix is upstream of the page. If three teams hold three pieces, no solo rewrite can assemble them into a brief.
  • Build the relationships one at a time, as a genuine multi-month goal, before you ask anyone for a deliverable.
  • Lead every conversation with your own data, so the exchange starts with you giving rather than asking.
  • Only once the network exists, bring it into one room and let the combined evidence write the roadmap.

The lever here is easy to underrate, because it looks like having coffee. But the reason those pages kept losing prospects was that the knowledge to fix them was locked in separate rooms, and the highest-leverage thing I could do was open the doors between them.

Once the evidence could move, the page nearly wrote itself.

If the fix for your funnel is stuck between teams that do not share what they know, that is exactly the kind of problem I am brought in to solve. Let's talk.