Pipeline Reporting at Seed: A Report You Can Trust
Nat Guy's useful split still holds: GTM engineering leans toward tools that create pipeline; RevOps leans toward tools that manage and measure it. At seed, you often cannot hire the second person. Commenters on pay surveys have said, in so many words, that they are not sure the average company will be able to afford a specialist. You still need a report you can trust — one a founder can read on Friday without cleaning rows for an hour first.
Pramod Kumar, co-founder of Kingdom of Kumar. On X
·
This is that report. It depends on CRM hygiene. The role split is in GTM engineer vs RevOps. The weekly habit is in A weekly GTM operating rhythm. The work is on GTM engineer and GTM engineering. If you are not ready to sell, take the App Launch Score and do not invent a forecast.
What "a report you can trust" means
It means every number can be traced to a list of records you would defend out loud. If "pipeline" is a sum of hopeful amounts on stale deals, it is not a report. It is a mood.
I would rather show a smaller, true set of conversations than a large number that would collapse under a question. Investors and cofounders can deal with a small number. They cannot deal with a number that changes meaning every time they ask how it was built.
Stages have to mean one thing. "Closed-won" is won. "Open" is a living conversation with an owner and a next step. "Closed-lost" is lost, with a reason if you have one. If you need a "maybe later" bucket, name it, and do not add its value into the number you call pipeline.
The seed report, on one page
Pull this weekly, by hand if you must:
- Conversations started this week — outbound replies that were real, plus inbound requests, counted as people, not as mail opens
- Meetings held — not meetings booked and cancelled
- Open conversations — the list, with owner, stage, next step, and next-step date
- Value, if you already price in a way that is not theatre — only on open conversations where a number has actually been discussed, or a published price you are honestly using. If you do not have that, skip value and count deals.
- Won and lost this week — with a short reason
- Inbound response misses — anything that broke the written time limit
- Outbound activity as context, not as vanity — approved first touches and follow-ups, under or over the cap
That is enough for a seed Friday. A funnel graphic with eight stages and a conversion percentage to two decimals is not.
If two people need to see the same page, put it in the CRM or a sheet that is fed from the CRM, not in a private tab.
Forecasts I would not make yet
I would not multiply a made-up close rate by a made-up value and call it a forecast. I would not import an enterprise forecast category into a five-deal pipeline. I would say, in words: these conversations are live, these are the next dates, this is what we have won, this is what we need to learn.
When you have a longer history, you can look at how deals like these have actually moved. Until then, a narrative plus a clean list beats a model.
Hygiene is the reporting system
If customers are not marked, if stages are a suggestion, if "open" includes people you have not spoken to since last quarter, the report is a fiction. The Friday hygiene pass and the Friday report are the same meeting. Fix the records first, then write the numbers.
A GTM engineer can help with both, but if you hire them only to build outbound and you never ask for this page, you will still be surprised at month end. If you cannot afford that hire, the founder still owns this page. That is the affordability point, in practice: the work does not vanish because the salary does not fit.
How to talk about the page with a board or a cofounder
Send the list, not only the totals. Say what you are unsure about. Do not pad the week with meetings that were really intros of intros. If a number moved because you cleaned data, say so: "the total dropped because we closed stale rows," which is a good change.
If someone wants a more "professional" dashboard, ask which decision they will make with the extra chart. If there is no decision, keep the one page.
A worked Friday, without invented figures
Imagine you held four conversations this week, started two new ones, lost one that had been sitting at "open" since last month, and missed one inbound reply by a day. The report is that list, plus the one missed reply, plus the outbound cap you did or did not keep. It is not a hockey-stick of "influenced pipeline." If a cofounder asks "are we on track," you answer with the next-step dates and what you learned from the loss, not with a made-up coverage ratio.
If this feels too small to be "reporting," that is the stage talking, not a failure of professionalism. The professionalism is in the honesty of the rows. When the list is longer, the same page still works. You add a count, maybe a value column you can defend, and you still attach the list.
I would also keep a short "definition of the numbers" at the bottom of the page, dated. When you change what "open" means, you change the date and you say so in that week's note. Silent definition changes are how year-on-year comparisons become fiction.
Metrics I would leave off the seed page
Open rates, LinkedIn impressions, "leads generated" that were really list uploads, and any dollar number that was typed in to look serious. Those can live in a scratch note if you are curious. They should not sit next to won and lost, where they will be added up by accident.
If a tool auto-builds a dashboard with those metrics, hide them until you have a reason. Default dashboards are designed for a different company.
When the CRM and the report disagree
Believe the list you are willing to read out loud, then fix the CRM to match it the same day. Do not maintain a side spreadsheet "because the CRM is wrong" for more than a week. A side spreadsheet becomes the new source of fiction. Either the CRM is the source, or the sheet is, and the other must be updated or retired.
FAQ
Should I include verbal interest as pipeline?
Not in the number I would call pipeline. You can have a second list called "not yet a conversation." Mixing them is how trust dies.
Do I need a BI tool?
No. You need true dates and stages, and a weekly habit. A BI tool on a dirty CRM will only make the fiction prettier.
Can this be part of the $500/month system?
Kingdom of Kumar builds and runs a scoped motion on Grok and Claude for $500 a month, three-month minimum. You own it, and you approve what goes out. We can leave you with a Friday page that matches the motion, not a fake forecast. We are not a RevOps department, and this is not a $135,000 specialist.