Every ecommerce founder we work with has three stories they reach for. The 8-figure exit. The tariff scramble. The SKU that went from nothing to category leader in nine months.
Those stories work. That's why they get reached for.
Then one day they stop, and nobody can tell you when it happened — because proof doesn't fail loudly. It expires. The post still gets likes. The story still gets nods on a podcast. But it stops doing the one job it was hired for: making a stranger believe you know what's happening right now.
We've written for enough operators to see the pattern clearly. The founders who lose credibility rarely lose it by being vague or by overclaiming. They lose it by running on receipts with a date they never say out loud.
Undated proof reads present-tense to your reader
This is the whole mechanism, and it's worth slowing down on.
When you write "we cut our return rate from 14% to 9%," you know exactly when that happened. It was 2023. You remember the meeting. To you, that sentence is clearly historical.
To a stranger reading it in 2026, that sentence has no date on it at all. They read it as a description of your current business. Not because they're careless, but because unmarked claims default to the present tense. That's how language works.
So you and your reader are holding two different documents. Yours says "here's something I did once." Theirs says "here's how this person's business operates today."
For a while, that gap is free money. Your best year is doing PR for your current year. Then something moves — a fee structure, a platform mechanic, your own involvement in the day-to-day — and the gap becomes the liability. Now you're not just quoting old proof. You're implicitly claiming something that isn't true anymore, and the reader who catches it doesn't email you about it. They just quietly reclassify you.
The proof bank tells you to capture the receipt and date it. Nobody tells you what to do when the date gets old.
Different proof decays at different speeds
Not all receipts age the same, and treating them as one pile is what gets founders in trouble. Four tiers, fastest-rotting first:
1. Platform-mechanic proof (6-18 months). "Here's the campaign structure that got our ACOS to 18%." "Here's the title format that ranked us." This is the most engaging proof you own and the most perishable. Amazon moved the title field in July 2026. The Buy Box eligibility gate is gone. Anything you proved about how the machine behaves has a shelf life measured in quarters. Post a 2024 mechanic as current advice and you don't sound experienced, you sound absent — like someone describing a road that's been rerouted.
2. Unit-economics proof (12-24 months). "Our contribution margin on that SKU was 31%." True when you calculated it. Since then referral and fulfillment costs moved, storage got more expensive in Q4, and the ad line went up because a competitor raised money. The number was real. The math underneath it isn't the math your reader is doing today.
3. Outcome proof (3-5 years, sometimes longer). "We took returns from 14% to 9% in one quarter." This travels. The how may be dated, but the outcome and the fact that you produced it don't rot quickly. This is your durable inventory.
4. Judgment and pattern proof (barely decays). "Across the accounts I've worked in, the brands that fix this first are the ones that survive Q4." Nobody can date this and nobody needs to. It's a claim about how things behave, backed by volume. It's also the hardest to fake, which is why it holds.
Most founders spend their proof in exactly the wrong order — leading with tier 1 because it performs, hoarding tier 4 because it feels less impressive. Tier 1 is the stuff that will make you look stale in eighteen months. Tier 4 is the stuff that made you worth reading in the first place.
Four tells that your proof has expired
You will not notice this from your dashboard. Reach holds up on name recognition long after credibility starts thinning. Look for these instead:
You're reaching for the same three stories. Not because they're your best, but because they're the ones you've told enough times to tell smoothly. Smoothness is a decay signal. When was the last time you wrote from something that happened this quarter?
You've started hedging the date. "A while back." "In a previous cycle." "When we were scaling that brand." Every one of those is your own instinct telling you the receipt is old, routed around instead of dealt with. If you can't say "in Q1 2024" comfortably, you already know.
Somebody asked if it still applies. The comment reads "is that still the case in 2026?" and it feels like a small clarifying question. It isn't. It's a reader telling you the claim didn't survive contact with their reality.
The specific posts stopped getting forwarded. Old proof still gets liked — it's a fine sentence, people are supportive. But nobody sends a colleague a tactic they suspect is out of date. Likes survive expiry. Forwards don't.
Why founders keep spending expired receipts
Two reasons, and neither is laziness.
The first is that the proof performed. You posted it, it did numbers, and you learned that this story works. Content that gets rewarded gets repeated. Nobody builds a habit around a post that flopped. So your highest-performing receipts become your defaults precisely because they were good, and they stay defaults long after the ground moved under them.
The second is more uncomfortable: fresh proof requires being in it. Tier 1 and tier 2 receipts come from the weeds — from pulling a report, running a test, sitting in the account. As founders succeed, they get further from the desk. The story from 2023 isn't chosen over a 2026 story. It's chosen because there is no 2026 story. The inventory ran out and the shelf was never restocked.
That's the connection to the altitude problem: drifting into abstraction and running on expired specifics are the same disease at different stages. First your proof gets old. Then, because you have nothing current and specific to say, you go general.
The fix: date it, audit it, convert it, restock it
Date every receipt in public. "In Q1 2024, we cut returns from 14% to 9%." One clause. It costs you nothing and it closes the gap between your document and your reader's. Founders resist this because dating it feels like weakening it. The opposite is true — a dated claim reads as confident, and an undated one reads as evasive to exactly the skeptical operator you're trying to convince. Precision about when is a credibility signal, not a discount.
Run a quarterly expiry audit. Open your proof bank. Add one field: as of. Then go through and mark anything platform-mechanic that's more than a year old as expired — don't delete it, flag it. Twenty minutes, four times a year. You'll usually find two or three things you've been saying confidently that stopped being true a while ago.
Convert expired proof instead of retiring it. This is the move most founders miss. An expired tier-1 receipt is raw material for durable tier-4 proof. "Here's the title structure that worked in 2024, here's what it was actually doing, here's why the 2026 version of that logic looks different." Now the old receipt is evidence you've watched a mechanic change across two eras — which is worth more than the tactic ever was. Nobody with only current knowledge can write that post.
Restock deliberately. If your content is going to run on current specifics, something in your week has to produce them. One live account. One test you personally read the output of. One report you pull yourself. Founders who stay credible for a decade aren't the ones with the best archive — they're the ones who never stopped generating tier-1 and tier-2 material, even at small scale.
FAQ
Doesn't putting a date on it make my best result look old? It makes it look true. The result is already old; the only question is whether you're the one who says so. Readers discount undated claims automatically because everyone in their feed is doing the same thing. Being the person who dates the receipt is a differentiator, not a concession.
My best number is from three years ago. Do I stop using it? No. Date it, and pair it with something current — even something small. "In 2023 we took returns from 14% to 9%. Last month I pulled return reason codes on a client's hero SKU and saw the same top driver." The old number provides the scale. The recent detail provides the proof of life. Alone, either one is weaker.
I'm not in the day-to-day anymore. What's left? Tier 4, and you should say so plainly. "I don't run the ads day to day now — what I see is the pattern across the accounts my team runs." That's an honest, strong position. What kills credibility isn't being off the tools; it's being off the tools while writing as though you're on them.
How often does this actually need attention? Quarterly for the audit, and once whenever a platform changes something structural. When Amazon moved the title field, every founder with a title-optimization receipt in their bank had inventory to re-date that week. Most didn't do it.
If you're a founder whose best stories are getting older than your audience realizes, that's usually a systems problem, not a memory problem. We build the capture and audit layer that keeps founders writing from this quarter instead of their best one. Get in touch if you want a look at what's still current in yours.