Designing the Metrics Tree Investors Expect
Round Ready Academy — Lesson 6 of 11
A Series A or Series B investor does not read your metrics. They read your metrics tree — the structured hierarchy that shows how top-of-funnel activity resolves into retained, paying customers, how that retention converts into contribution margin, and how the resulting unit economics justify the fund you are asking to write the cheque.
This lesson covers what that tree looks like, the metrics each layer contains, and how to present it so that defensibility — a word often used and rarely evidenced — becomes a conclusion the IC can draw from the data rather than a claim they have to take on trust.
Your metrics are the inputs. The tree is the structure that turns inputs into a thesis. A well-designed tree lets a Series A partner move from "top-of-funnel is X" to "the round prices at Y" in a single continuous logical path. Founders who present disconnected metrics leave the work of drawing that path to the investor, and pay for it in the pre-money.
The Four Layers of a Series A Metrics Tree
A well-designed tree has four layers. Each layer answers one question and feeds the next.
The Four Layers
| Layer | Question it answers | Example metrics |
|---|---|---|
| 1. Top-of-funnel | How do prospects reach you, and at what volume and cost? | Branded search volume, inbound volume, outbound activity, MQLs per channel, CAC by channel |
| 2. Conversion and quality | How well does the funnel convert, and which segments close? | Stage-by-stage conversion rates, win rate by segment, sales cycle, deal size distribution |
| 3. Retention and expansion | How do closed customers behave after the first contract? | Gross logo retention, NRR by cohort and segment, expansion ARR, churn reasons |
| 4. Contribution margin and unit economics | What is the economic shape of each customer? | Contribution margin, LTV by segment, payback period, CAC-to-ARR ratio |
The tree is an argument: top-of-funnel produces qualified prospects who convert into customers who retain and expand and produce contribution margin that justifies the capital being raised. Every layer has to hold up on its own and feed the layer below.
Layer 1 — Top-of-Funnel
The question an investor is asking at this layer is not "how big is your funnel?" but "how repeatable and cheap is your funnel?"
Repeatability shows up in channel diversity. A pipeline that is 90% one channel — be it paid, outbound, or founder-led inbound — concentrates risk. A pipeline where three or four channels each produce 15 to 30 percent of qualified volume is structurally safer and typically priced more favourably.
Cheapness shows up in CAC by channel. Blended CAC is almost useless at Series A. CAC per channel, per segment, over the last four quarters, with trajectory — that is the useful view.
Layer 2 — Conversion and Quality
Layer 2 is where most founder decks over-present and most diligence partners under-trust. The overall win rate tends to be flattering; the stage-by-stage conversion is the rigorous view.
A metrics tree that holds up at Layer 2 typically shows:
- Conversion from MQL to SQL (stage 1)
- Conversion from SQL to opportunity (stage 2)
- Conversion from opportunity to closed-won (stage 3)
- All broken out by segment, not blended
This level of detail lets the partner identify whether your funnel is volume-constrained (low top of funnel) or conversion-constrained (low stage-3 conversion). That difference shapes the Series A thesis: a volume-constrained funnel is typically fixed with capital (the thesis for raising); a conversion-constrained funnel is usually an organisational problem and a harder sell.
Layer 3 — Retention and Expansion
This is the most-read layer in every Series A IC memo. We covered retention as a driver in Lesson 3 and as a question set in Lesson 5; this section is about how it sits in the tree.
A strong Layer 3 presentation has three things:
- Cohort retention curves — gross logo retention and NRR, per segment, with at least eight quarters of history where the business permits
- Cohort maturation evidence — showing that second-year retention for later cohorts is equal to or better than earlier cohorts (evidence of improving product fit or motion)
- Expansion story — what drives NRR above 100%, which segment generates it, and what the expansion motion is (usage-based, seat-based, tier-based)
Net revenue retention (NRR) at Series A is typically expected at 100%+ in best-performing segments. At Series B, the bar rises to 120%+ average across segments. These are not absolutes — sector matters — but they are the benchmarks against which your numbers will be read.
Founders who lead with headline NRR often present a blended figure that obscures segment variation. The honest view breaks it out: if one segment does 134% NRR and another does 82%, that is the view the IC wants, not the blended 108%.
Layer 4 — Contribution Margin and Unit Economics
Layer 4 is where the metrics tree resolves into the economic shape of the business. This is the layer where LTV, CAC, and payback period live — but only in honest form.
Three rules for honest unit economics at Series A:
1. LTV uses observed retention, not assumed retention
A 36-month LTV assumption built on a 10-month observation window is a forecast, not a fact. Present the observed retention explicitly and state the LTV sensitivity to different retention assumptions.
2. CAC is fully loaded
Sales salaries, marketing spend, contracted SDRs, tooling, events, content — all of it. Founder-led CAC is often understated because founder time is not costed. A fully loaded CAC is harder to justify; it is also the one the IC will calculate regardless.
3. Payback period is cash-on-cash, not gross-margin-on-CAC
The useful definition of payback is "months until the cumulative contribution margin from a cohort equals the CAC spent to acquire it." Shortcuts that use gross margin tend to present a more flattering figure and are corrected in diligence.
Present unit economics honestly and the partner trusts the rest of the tree. Present them flatteringly and every other layer is read with scepticism.
Defining "Defensibility" — in Evidence Terms
The word "defensibility" is used in every pitch and evidenced in very few. For a Series A investor, defensibility is not a claim. It is a specific question: what makes it harder to compete with you next year than it was this year?
The answer lives in the metrics tree, layered against the Opagio 12 asset base.
Defensibility — What Evidence Looks Like
| Claim | Evidence that supports it | Evidence that does not |
|---|---|---|
| Network effects | Later cohorts retaining better than earlier cohorts | "Our platform has two sides" |
| Data moat | Growing labelled dataset, year-over-year | "We have a lot of data" |
| Brand | Branded search up 40% YoY, NPS 62 | "Customers love us" |
| Switching costs | Gross churn <8%, average contract 24 months, top-tier integration depth 11 systems | "It's sticky" |
| Regulatory | FCA permission obtained 2023, blocking 18 months of competitor runway | "We're regulated" |
The pattern is the same throughout: defensibility is proven by data, timelines, and trajectories, not by adjectives.
The Test for a Metrics Tree Being Round-Ready
If a Series A partner can read your tree without you in the room and arrive at a pre-money range that matches what you are asking for, the tree is round-ready. If they have to guess at a link between layers — if they have to call you to understand how top-of-funnel connects to retention, or how retention connects to contribution margin — the tree is not round-ready yet.
Common Tree-Design Mistakes
Three mistakes tend to recur in Series A tree design:
- Collapsing layers. Showing only a blended MQL-to-revenue conversion rate hides whether volume or conversion is the constraint. The partner will ask for the breakdown anyway.
- Cherry-picking the segment. Presenting NRR for only the best segment and treating it as the business NRR. Every segment has to be visible.
- Unit economics without retention. LTV cannot exist without an observed retention base. A tree that presents LTV/CAC without the underlying cohort maturation chart will be discounted.
Fix these and the tree tends to support the round.
What to Present and When
The tree is a diligence document, not a pitch document. Do not put it in the first-meeting deck; do put a summary version in the appendix so the partner knows it exists. Present the full tree in the first diligence session, ahead of any other workstream. It anchors the rest of the diligence conversation.
The Tier Decision — Diagnostic or Full Valuation Defence?
Most founders preparing the tree are best served by running the Round Readiness Diagnostic first, which is free. The diagnostic tells you which layers of the tree are likely to be thin, and therefore which work to prioritise.
For founders already inside an active process and wanting a full Valuation Defence Report — the document that combines the tree, the asset register, and the comparable analysis into one board-ready output — the right tier is Growth.
Primary CTA: Run the Round Readiness Diagnostic (free) — or generate a Valuation Defence Report preview on the Growth tier (£1,499) for the full output institutional investors will read.
For the broader unit-economics framework, see LTV and CAC and Cohort Analysis. For the next lesson in the course, continue to Lesson 7: The bridge round decision tree.
Ivan Gowan is Founder and CEO of Opagio. Previously founded and scaled two regulated financial platforms through multiple institutional rounds. Meet the team.