The most common reason an ecommerce founder's post is weaker than it should be isn't writing ability. It's a redaction decision made at 11pm, with no rules, by someone who is tired and slightly nervous.
The founder starts with a real story: a return rate that went from 8.1% to 5.4% after a sizing graphic moved to slot two. By the time it's published, it reads "small changes to your listing can have a big impact on returns." Every specific has been sanded off. Not because anyone objected — because there was no policy, so the safest available move was to remove everything that could possibly be a problem.
That's the disclosure ladder problem. And it's a content system gap, not a courage gap.
Why Ecommerce Founders Have This Worse Than Anyone
Most B2B founders talk about frameworks and process. Their raw material isn't especially sensitive.
Ecommerce founders sit on the opposite pile. Your best content is unit economics, supplier friction, ad spend that didn't work, a launch that missed, a return reason nobody predicted. All of it is specific, all of it is verifiable, and all of it is genuinely readable by people whose interests aren't aligned with yours:
- Competitors in your category who would like to know your volume
- Suppliers who would like to know your margin before your next price negotiation
- Retail partners who have opinions about what you say publicly
- Employees who read everything you post
- Acquirers, if you're anywhere near a process
So the founder is being rational. The problem is that the rationality shows up at the worst possible moment — during the draft, when the only two options visible are "publish the number" or "publish nothing."
This Is a Permission Gate, Not a Quality Gate
We've written before about the pre-publish checklist — the kill test that asks whether a post is good enough to ship. This is a different gate answering a different question at a different time.
The quality gate asks: is this worth publishing?
The disclosure gate asks: how much of this am I allowed to show?
Founders collapse the two, and when they do, the disclosure question wins by default — because "make it vaguer" always feels safer than "make it sharper," and a vague post still clears a quality checklist if you're grading yourself generously at 11pm.
Separate them. Answer the permission question first, once, at capture time. Then the quality gate has something specific to grade.
The Five Rungs
Every real number you own can be published at one of five levels of disclosure. The ladder runs from most exposed to least.
Rung 1 — Raw absolute. "We did $83,400 in July on that SKU." Maximum credibility, maximum exposure. Reserve it for numbers that are already public, already historical, or genuinely harmless.
Rung 2 — Banded absolute. "A SKU doing around $80K a month." Rounds off the precision that makes a number traceable. Loses almost nothing in reader trust. This is where most founder proof should live and almost none of it does.
Rung 3 — Ratio only. "Return rate went from 8.1% to 5.4%." Publishes the performance without publishing the volume. A competitor learns your return rate improved; they learn nothing about how big you are. Ratios are the most under-used rung on the ladder.
Rung 4 — Direction plus magnitude. "We cut returns by about a third." No baseline, no volume, still a real claim. Weaker than rung 3, but usable when the ratio itself is sensitive.
Rung 5 — Mechanism only. "The return driver wasn't the product. It was a sizing expectation set by image two, and we found it in the return reason codes, not the reviews." No number at all. And here's the thing most founders get backwards: this is not the weakest rung.
Mechanism Travels Further Than Magnitude
A number proves you were there. A mechanism proves you understood what happened.
Readers forget the number within a day. What gets saved, forwarded, and quoted back to you six months later in a sales call is the how — the specific causal chain that a stranger couldn't have guessed and can now go check in their own account.
We see this in the data constantly. Posts built on rung 5 — pure mechanism, zero disclosure — routinely out-perform rung 1 posts on saves and DMs. The rung 1 post gets more likes. The rung 5 post gets the reply that says "I just pulled our return reason codes and you were right."
That matters for the anxious founder, because it means the ladder is not a trade between safety and impact. You can descend several rungs and lose almost nothing that generates pipeline. What you lose is applause.
What Actually Leaks (And What Founders Guard For No Reason)
Founders tend to guard the wrong things. In our experience the genuinely sensitive items in ecommerce content are narrower than the anxiety suggests:
Actually sensitive:
- Unit volume, which tells a competitor your reorder size and tells a supplier how much leverage you don't have
- COGS and landed cost, for the same reason, one negotiation later
- Anything that identifies a specific live ASIN alongside a performance number
- Timing of an unlaunched product or a channel you haven't entered yet
- Named partners — 3PL, agency, manufacturer — attached to a complaint
Usually not sensitive, but guarded anyway:
- Percentages and ratios with no volume attached
- Historical numbers from a product you no longer sell
- Category-level dynamics everyone in the category already knows
- Your own mistakes, which are the highest-performing content you own and carry almost no competitive value to anyone
The Pairing Rule
One rule prevents most real leaks: publish the number or the identifier, never both.
"An $80K/month SKU in cabinet hardware" is fine. "An $80K/month SKU" plus enough detail that someone can find the listing in ninety seconds is a gift to a competitor. The number is safe because it's unattached. The moment it's attached, it's intelligence.
Set the Ladder Once, at Capture Time
The system fix is to move the decision upstream — out of the draft and into capture, where you're calm and not staring at a deadline.
Write a one-page disclosure policy. It takes twenty minutes and covers four things:
- A default rung per topic. Revenue: rung 2. Margins: rung 5 or nothing. Ad performance: rung 3. Returns and CVR: rung 3. Mistakes and mechanisms: rung 1, publish freely.
- The always-off list. Named suppliers, unlaunched products, anything under NDA, anything involving an active dispute.
- Who else signs off, if anyone — and if the answer is a committee, that's a separate and more expensive problem.
- The pairing rule, written down, because it's the one you'll break under pressure.
Then assign the rung when you log the receipt, not when you write the post. If you're already running a proof bank, add one field: rung. The raw number goes in the file. The publishable version goes in the same entry, pre-cleared.
Now assembly never stalls. You sit down to write, you pull a receipt, and the permission question is already answered by a version of you who wasn't in a hurry.
The Competitor Objection, Taken Seriously
The honest version of the objection is not "someone might see it." It's "the person most motivated to read my content closely is the person competing with me."
That's true. Two things about it.
First, most of what makes founder content work is not extractable into a competitive advantage. A competitor reading your post about a sizing graphic in slot two now knows to check their own sizing graphic. That's it. They still have to have your product, your reviews, your photography, and your judgment about which objection matters most. Insight is cheap to read and expensive to implement.
Second — and this is the part we don't sand off — some things genuinely shouldn't be posted. A supplier relationship you're renegotiating in six weeks. A category you're entering in Q1. A margin structure that's your entire moat. The ladder isn't a trick for publishing everything anyway. Rung 5 exists because sometimes the number stays home and the lesson still ships. And occasionally the right answer is that the whole story stays home. A policy that never says no isn't a policy.
Frequently Asked Questions
Isn't a post without numbers just generic content? Only if it's also without mechanism. Generic content is missing specificity, not digits. "Focus on your customer experience" has no number and no mechanism. "The return driver was an expectation set two images before the buy button" has no number and is completely specific. Digits are one way to be specific. They are not the only one.
How do I know which rung a given story needs? Ask what a competitor could do on Monday with the information. If the answer is "nothing they couldn't already do," you're overprotecting. If the answer is "walk into a negotiation better prepared than me," go down a rung.
What if I already published something at rung 1 that I regret? Leave it up unless it's genuinely damaging. Deleting draws more attention than the post got. Fix the policy going forward — the point of a system is that the same decision doesn't get re-litigated every week.
Does this slow down publishing? It speeds it up. The slow part of founder content isn't writing, it's the stall — the twenty minutes spent deciding whether a sentence is safe, repeated on every post, forever. Deciding once at capture removes that entirely.
My co-founder disagrees with where I set the defaults. Then that's the actual conversation, and it's worth having once in a room rather than implicitly on every draft. Disagreement about disclosure defaults is normal. Disagreement discovered mid-post is what produces content that says nothing.
Founders don't lose to competitors because of what they published. They lose relevance because of what they sanded off — every week, in private, with no rule telling them they didn't have to.
If you'd rather have the ladder built into your content system than rebuilt from scratch at 11pm every Tuesday, that's the kind of thing we set up in the first month. Talk to us about it.