The Shift · By Norm deSilva

Editorial discipline as competitive moat.

Every B2B marketing page on the internet now sounds the same. We can tell, and so can buyers. The work IG does to remove that sameness is not a style guide. It is a strategy.

Norm deSilvaExecutive Partner, Innovative Group2026-05-24

A B2B marketing director shared three competitor homepages with me last week. She asked me to identify which one belonged to which company without the logos. I could not. None of us could. The pages were interchangeable. Same hero structure, same three-up benefit grid, same testimonial carousel, same "demo" CTA in the same teal button.

The pages were not written by the same agency. They were written by four different agencies serving four different companies in the same category. What they shared was the editorial template that AI tooling now defaults to. Em dashes. Binary contrast filler. Parallel-list rhythm. "It is not X. It is Y." stock transitions. Every page sounds like every other page because every page was written through the same model with the same default outputs.

This is the editorial wave we are sitting inside. And it is the competitive moat we are taking the time to build at IG.

The moat is the difference between sounding generic and sounding like a person

A buyer who lands on a page that sounds like every other page in the category does not remember the page. The page does its job (descriptions, capabilities, social proof) and the buyer moves on. The brand has no purchase in the buyer's memory because nothing on the page sounded like a specific operator with a specific point of view.

A buyer who lands on a page that sounds like someone in particular wrote it has the opposite experience. They remember the line. They quote it back in the discovery call. They send the URL to their colleague with a note that says "this one is different." The brand has purchase because the writing did.

This is not new. Every advertising era has its version of the lesson. The new wrinkle in 2026 is that the default agency output sounds AI-generated, even when an AI did not write it, because the people writing it grew up reading AI-generated content as their reference set. The reference set produces the output. The output produces the indistinguishability.

What the discipline looks like

We strip four things from every piece of writing that ships from IG.

First: em dashes. The em dash is the canonical AI-tell. Every model overuses it. Most human writers do not, when given a clean reference set. We use commas, periods, parentheses, or restructure the sentence. The result reads as more deliberate, which it is.

Second: binary contrast filler. The "Not X. Y." pattern. "This is not a tool. It is a partner." The pattern feels rhetorical and is in fact filler. Most binary contrasts collapse to a single declarative sentence that says Y. Cut the X half.

Third: stock transitions. "Let us dive in." "Here is the thing." "The truth is." These phrases consume attention and deliver no content. The piece moves faster without them. Readers stay with you longer.

Fourth: parallel lists masquerading as prose. The three-item-rule list with each item the same length and structure ("Strategy. Execution. Outcomes."). The pattern reads as algorithmic even when a human wrote it. Where the meaning requires three things, name three things specifically; do not force them into the same syntactic shape.

Why this is a moat

A competitor can copy our pricing model. A competitor can copy our specialty-team structure. A competitor can copy our product (Beat, NBA) given enough engineering effort. None of them can copy our editorial voice without rebuilding the team that produces it, because the voice is what the team writes when no one is editing them.

The moat compounds. Every Shift piece, every blog post, every landing page is in the same voice. A buyer who reads three IG pages in a session forms an opinion about the company that does not depend on which page they read first. Coherence becomes the brand. Coherence becomes the differentiator. Coherence is hard to fake at scale because the second-tier copy reveals the seams.

The cost of the discipline

Writing this way takes longer. A first draft that includes em dashes and binary contrasts is a normal first draft. Removing them adds a pass that most agencies do not budget for. The trade-off is real.

I think the trade-off is worth it. The cost of one extra editorial pass is small. The cost of having a homepage that sounds like every other homepage in the category is large. The math runs in favor of the discipline.

What I tell new operators inside IG

Two things, every time.

Read your work aloud. Anything that does not sound like a sentence you would say to a colleague gets rewritten. The test is unforgiving and it works.

Cut every adjective that does not earn its keep. Most adjectives are decorative. The ones that survive are the ones that change the meaning of the noun. Everything else is noise the reader has to skip past.

The voice is not a style. It is the absence of the patterns that most agency writing has been trained on. The discipline is removing those patterns, not adding new ones. What is left after removal is the operator's actual point of view, which is what we have been hired to surface.