Northwind Labs — GTM diagnostic
A real report structure with an anonymised company. Seed-stage B2B SaaS, nine customers, eleven months of runway. Generated in four minutes from a fifteen-minute intake plus public research.
Your ICP is three companies wearing one label.
You describe a single customer profile, but your closed-won accounts fall into three groups with materially different buying behaviour. Budget, messaging and outbound are spread evenly across all three, which makes every channel look mediocre and no channel look repeatable.
Overall
Choose one segment for the next two quarters.
Segment A converts fastest and churns least. Everything else becomes inbound-only until you have twenty accounts there. This is the decision that unblocks all the others.
Delay the GTM lead hire by one quarter.
Hiring before the segment is chosen means writing a job description for a role you cannot yet define. The wrong hire here costs six months and roughly $180k against an eleven-month runway.
Instrument the funnel before adding to it.
You cannot currently tell which experiments worked. Three days of tracking work makes the next three months readable — without it, every result stays an opinion.
Run outbound to Segment A only for six weeks.
Hypothesis: reply and close rates rise sharply when messaging matches one buying context instead of averaging three.
Rewrite the homepage around one job-to-be-done.
Hypothesis: demo requests are suppressed by a page that describes a category rather than a problem the buyer already has language for.
Quote Segment C at 3× and refer out anyone who says no.
Hypothesis: Segment C is not a targeting problem, it is a pricing-signal problem. The accounts that accept 3× are your real ICP hiding inside a segment you were about to abandon. The ones that refuse were never going to be profitable at 96 days.
Why this is the dark horse — every instinct at seed says discount to close a slow segment. Doing the opposite tests willingness to pay and segment quality in a single move, at zero acquisition cost. No competitor will copy it, because it looks like turning down revenue.
Every Oracle report contains one dark horse. Two of your three experiments will be things a good operator would eventually reach. The third is deliberately the one nobody in your market is running — drawn from patterns across companies at your stage, in adjacent categories, that your competitors have no reason to have seen. It is not there to be clever. It is there because at seed, the experiments everyone runs produce the results everyone gets.
Same constraint: three segments, one label.
Cut two of three segments in month one. Revenue dipped for six weeks, then outbound reply rates roughly tripled because every message finally assumed one context.
Tried to serve all three at once.
Hired a GTM lead before choosing. The hire optimised the wrong segment for two quarters, then left. They chose a segment nine months later than they needed to.
Ran the pricing dark horse.
Re-quoted their slow segment at 2.5× expecting to lose all of it. Kept a third — and those accounts became the highest-margin cohort in the business.
Narrow
Cut the surface area. One segment, one message, a funnel you can actually measure.
Prove
Run the three experiments that can be read against a threshold you set in advance.
Repeat
Turn what worked into a motion your team can run without you in the room.
Book a GTM Clarity Call.
Segment choice is the highest-stakes decision on this list and the hardest to reverse. It is worth pressure-testing with someone outside the company before you commit two quarters to it. If we thought you could run this plan alone, this section would say so instead — roughly a third of reports do.
Fifteen minutes for a report like this one.
Free, no card, no call required. Answer a structured set of questions and the Oracle reads them alongside public research on your company.
10–15 min · One bottleneck named · Three experiments, one dark horse