You announced a $55M Series B yesterday and said you're hiring across go‑to‑market. Here's the plan first. The reasoning behind every line of it is one click away, underneath.
A benchmark is the only asset that's proprietary, repeatable every year, quotable by press, rankable in search, and self‑qualifying when someone fills it in. You're the only company that can publish this one.
It's the qualification form that doesn't feel like one. Four inputs, one number about them — and the answers become next year's dataset.
You're leaving about $225K on the table — not because the charges are valid, but because nobody has time to work them inside the retailer's dispute window.
4.1% median, ~33% invalid share, recovery scaled to disputes worked — all invented to make the mechanism testable. The real ones come out of your first data pull. If the actual spread is narrower, the campaign's headline number isn't surprising any more and I rewrite the copy around whichever figure is. The asset survives either way; only the lead changes. That's the job, not a setback.
Whether your customer agreements permit publishing aggregated, anonymized deduction data. I'd ask before writing a line of copy. If the answer is no, the fallback is a blinded index — retailers grouped as national mass, natural channel, club — plus a 200‑brand survey. Weaker, still publishable, still the best asset in the category.
| Proprietary | Nobody can copy it without your transaction volume. |
| Repeatable | It's an annual. Year two is a trend line, which is a better story than year one. |
| Linkable | Trade press cites benchmarks. Ebooks earn no links, and links are the only durable SEO asset. |
| Self‑qualifying | The calculator collects retailer mix and dispute volume — a lead score and next year's dataset in the same field. |
The Series B date and detail are verified. Plan‑season and line‑review timing is category knowledge I'd sanity‑check against your own closed‑won dates in week one.
| Segment | The filter I'd build | Accounts | Trigger that puts them in | Entry channel |
|---|---|---|---|---|
| A · Just got the door | $15–75M gross, won a new national retailer or distributor authorization in the last two quarters, finance team of three or fewer | 610 | Door count jumps; deduction volume follows 60–90 days later and nobody is staffed for it | Cold email + DM |
| B · Plan season | $75–250M, demand planner on staff, broker network, still on spreadsheets or a legacy trade tool | 430 | 2027 AOP is due and the deduction and trade assumptions in it are guesses | Paid ABM + webinar |
| C · Foodservice entrant | Retail brand adding foodservice or c‑store distribution — the expansion your Series B explicitly funds | 280 | New lane, unfamiliar deduction codes, different promo mechanics | Earned + partners |
New‑item velocity in SPINS or Nielsen, trade‑show exhibitor lists, broker announcement posts, job postings for "Demand Planner" or "Deductions Analyst," ERP migration chatter, and the retailers named on your own integrations page. Account counts are modeled from the CPG brand population above $10M — they're the number I'd most want to replace with your closed‑won data in week one.
| Persona | Judged on | The sentence | Proof I'd send |
|---|---|---|---|
| Controller / AR | Unapplied cash, backlog age | "Your backlog isn't a staffing problem, it's a matching problem." | Feastables reclaimed $500K in invalid and duplicate charges |
| VP Finance / CFO | Gross‑to‑net, margin per point | "You're negotiating the 4% you can see and eating the 1.4% you can't." | Dude Wipes: 2× revenue on the same team, ~$300K headcount avoided |
| VP Sales | Forecast accuracy, broker reporting | "Your forecast and your trade spend live in different files. That's the whole variance." | Deep Indian Kitchen: a 2027 annual plan across 24,500 stores |
| COO / Supply | OTIF fines, weeks of cover | "Beat plan all spring, overstocked by summer." | Your own newsletter headline — already the best line you've written |
All four proof points are yours. You have four hard dollar outcomes sitting under a Resources menu with nothing pointed at them.
Nothing in week three needs writing from scratch. The retailer cards, the microsite pages and the cold‑email first lines are all cuts of the same dataset — that's the only way one person runs this.
Subject lines lead with a number, because a number is the only thing in an inbox that can't be skimmed past. The ask at every step is for a document, not a call.
{{first}} — across $1B of retailer deductions we've processed for 250+ brands, the median brand gives back 4.1% of gross sales. At {{brand}}'s scale that's roughly {{$leak}} a year, and about a third of it is invalid: duplicate charges, promo that was authorized, shortages that shipped complete.
We just published the breakdown by retailer. {{retailer}} is the worst line on the page.
Want the two‑page cut for {{retailer}}?
We looked at $1B of retailer deductions across 250+ brands.
The median brand loses 4.1% of gross sales. The best‑run brand on our platform loses 1.6%.
That gap isn't negotiating skill. It's documentation — whether you can produce the PO, the BOL and the promo authorization inside the retailer's dispute window.
Full breakdown by retailer, free, no form for the headline numbers. The Deduction Index 2026, link in comments.
Same retailers as you, same problem. Feastables found $500K in invalid and duplicate charges in year one. The work wasn't negotiation — it was matching remittance lines to promo authorizations automatically instead of by hand.
Two‑minute version here.
If someone on your team is still exporting the retailer portal to Excel on Fridays, that's the whole story.
Last one from me. If you're locking the 2027 plan in the next six weeks, the deduction assumption in it is probably last year's number plus a guess.
I'll send the retailer‑level medians so you can put a real one in. No call needed — reply "send" and it's yours.
Saw you pulled the Index — the {{retailer}} page is the one people screenshot most.
Genuinely curious: are you working disputes in the portal yourself, or has someone taken it over? I'm collecting how teams handle the 30‑day window for next year's edition.
Today: "Plan trade without the clicks" — a product update.
Instead: "Trade planning: what 250 brands budgeted vs. actually spent in Q3." Same product news, delivered as the reader's benchmark, with the product appearing in paragraph three as the mechanism.
Keep the changelog — as a fixed What shipped block at the bottom.
3,600 cold sends at 5.5% reply, 35% of replies positive, 45% of those to a held demo. 108,000 paid and organic impressions at 3.5% CTR. Calculator start‑to‑complete at 60%, which is the number I'd defend with field count. Close rate 22% on a 90–120 day cycle.
What breaks it: if your average cycle runs longer than 120 days, or deals need a CFO who doesn't read LinkedIn, the demo number holds and closed‑won slips a quarter. I'd rather be wrong on the timing than on the mechanism.
A test tells you which of two things you already thought of did better. A loop changes what you send, who you send it to, and eventually what you're selling — on a clock. So the number that matters here isn't reply rate, it's cycle time.
Thresholds, written before the sends: reply ≥6% scale, 3–6% iterate, <3% kill, nothing judged under 200 sends. LinkedIn CTR ≥2.5% scale. Calculator completion ≥55%, or cut a field. Kill any channel above $75 per Index read.
| Variant | Test | Result | n | Call | What we do Tuesday |
|---|---|---|---|---|---|
| Hook | "4.1% of gross" vs. "the 30‑day window" | 2.9% / 1.4% | 41k imp | scale | All paid spend to the percentage hook; the window becomes body copy |
| Subject | "your 2027 plan" vs. "4.1% of gross" vs. "quick question re: Kroger" | 7.1 / 5.8 / 2.2% | 600 ea | scale | Deadline subject moves to step 1; retire "quick question" for good |
| Segment | A new‑authorization vs. B plan‑season vs. C foodservice | 6.9 / 5.1 / 2.4% | 840/720/140 | hold C | C is under n — no conclusion. Finish the cohort before deciding |
| Asset gate | Gated PDF vs. ungated + calculator | 31% / 69% | 1,180 | scale | Remove the form everywhere — the calculator collects more than it did |
| Channel cost | LinkedIn ABM vs. podcast read vs. search | $41 / $118 / $22 | per read | kill podcast | Budget moves to search; keep one show that reaches ops, not finance |
| CTA | "Book a call" vs. "Want the two‑page cut?" | 0.9% / 6.2% | 1,200 | scale | Document‑ask everywhere, including the ads. The call comes one step later |
Segment C is held, not killed — 140 sends is noise, and killing a segment on noise is how a company decides a market doesn't exist.
The CTA result is the largest effect on the board, which should make us suspicious: it's probably measuring intent rather than persuasion. The real test is whether document‑askers close at the same rate. That's a 90‑day question, and it goes on the board with a date, not a shrug.
Caveat: v1→v2 changed the subject and drifted the segment, so part of that +2.5pp is list quality, not copy. I'd hold the segment constant next round and accept a slower answer.
| Signal | Where it fires | What we keep | The decision it feeds |
|---|---|---|---|
hook_impression | LinkedIn organic + ads export | hook_id, segment, format | Which of the twelve retailer cards gets paid spend next week |
index_view | Index microsite | utm_campaign, retailer page, scroll depth | Which retailer page to build next — readers pick the roadmap |
calc_complete | Calculator | revenue band, retailer mix, dispute volume | Two things at once: a lead score, and a row in next year's Index |
reply_classified | Inbox → CRM | objection_tag, from a fixed list of nine | Message library — three of the same tag rewrites the sequence |
demo_held | CRM | source hook, first objection on the call | Next month's subject lines come from here, not a swipe file |
deal_lost | CRM | stage, reason, competitor named | Segment definition — losing the same way twice means the filter is wrong |
closed_won | CRM | days from first touch, full touch path | Channel budget, and the only number that settles a channel argument |
The part people skip: calc_complete is the whole design. Someone tells us their retailer mix and dispute volume in exchange for a number about themselves. That row scores the lead, writes the first line of their email, and becomes a data point in next year's Index. The loop feeds the asset, not just the copy — which is why year two costs a fraction of year one.
The copy isn't the problem, the segment is. Escalate to a list rebuild, not a fourth rewrite — which is where most outbound programs quietly die.
That's positioning, and it belongs on the website and in the sales deck, not in a subject line. It leaves the loop and goes to the product marketing backlog with the three quotes attached.
People typing something we didn't ask for is the cheapest product research there is. It goes to the product team weekly, verbatim, with counts.
| 1 | What crossed a threshold, up or down? Numbers only, no stories yet. |
| 2 | What did people actually say? Three real replies, read aloud. |
| 3 | What are we changing, by Wednesday? Three decisions, named owner. |
| 4 | What did we learn that isn't about copy? Goes to product, sales or the roadmap. |
Whoever ran the sends, one AE who took the calls, someone from product every other week. Thirty minutes, hard stop.
One page leaves: three decisions, the threshold each was judged against, and the one thing we got wrong last week. That last line is why anyone trusts the page by week six.
The AE isn't there as a courtesy. Marketing's fastest signal is what an objection sounded like on Thursday, and no dashboard carries that.
I read everything you publish before writing any of this. None of it is a knock — it's what marketing looks like when founders are doing it between customer calls, which is the correct trade until you hire for it.
Of the nine posts on The Confido Aisle, three are Engineering — two founding‑engineer interviews and an AI hire announcement — and exactly one is filed under CPG Playbooks. There's a gap from Apr 1 to May 27, and another to Jul 7.
New on the Shelf describes itself as "Confido's weekly roundup of product updates, upcoming events, and industry highlights," and the cadence has slipped — Aug 5, 12, 20, then Sep 9. Your best headline, "Beat Plan All Spring, Overstocked by Summer," is buried in it.
$500K reclaimed at Feastables, $325K at Every Man Jack, ~$300K of headcount avoided at Dude Wipes with 2× revenue on the same team, a 24,500‑store annual plan at Deep Indian Kitchen, and a competitive win at Evergreen — all sitting under a Resources menu.
The integrations page lists Ahold, Albertsons, Aldi, BJ's, Costco, AWG, Certco, Core‑Mark and C&S — then abstracts the rest to "Accounting, ERPs, and Sales Data."
The matrix exists in the nav, and the industry pages are the only ICP‑shaped entry points. Nothing lands a "$40M beverage brand that just got into Kroger" anywhere specific.
| Event | Key properties | Lands in |
|---|---|---|
hook_impression | hook_id, segment, format, spend | warehouse |
index_view | utm_*, retailer_page, dwell | warehouse + CRM |
calc_start | session, entry_page | warehouse |
calc_complete | rev_band, retailers[], disputes_pm, computed_leak | CRM + Index dataset |
reply_classified | variant_id, objection_tag, sentiment | CRM |
demo_held | source_hook, first_objection, segment | CRM |
closed_won | touch_path, days_to_close, acv | CRM + warehouse |
| Legal | Can aggregated, anonymized deduction data be published? Tab one rests on this. |
| CRM | HubSpot or Salesforce — and is the reply‑to‑opportunity path already joined? |
| Baseline | Site traffic, list size, inbound demo volume. I modeled a funnel without knowing your starting point. |
| Budget | Paid budget for Q4, and whether events already claim most of it. |
| Ownership | Who owns SEO and the website today, and what they'd want to keep. |
| Capacity | Sixty demos a quarter is only a good plan if the team can hold them. |
If inbound is already most of your pipeline, the cold‑email lane is the wrong place to spend a new hire's first month. The Index and the site work still hold; I'd shift the 30‑day plan to conversion rather than acquisition.
If your buyer is more often a CFO than a controller, LinkedIn weight drops and the webinar and partner channels carry more. The message map changes; the loop doesn't.
If the deduction spread turns out narrow, the headline isn't surprising and the Index needs a different lead — probably recovery rate rather than deduction rate.
The part I'd defend hardest is the loop. The campaign is a guess until the first 400 sends come back; the system that turns those replies into a better next 400 is the thing worth hiring for.