Most go-to-network programs are killed by the wrong scoreboard. Someone pulls up the activity dashboard, sees 40 introductions next to 4,000 outbound touches, and concludes the channel is small. It is small on that axis. It is also the channel that produced the senior meetings, and nobody measured that because nobody defined the metrics.
This post sets out five that do the job, how to define each one so two people get the same number, what good looks like, and the two reports that make the channel visible next to outbound and inbound.
Why outbound's metrics do not transfer
Outbound is a volume channel: more sends, more replies, more meetings, in roughly linear proportion. Its metrics measure volume and conversion of volume.
Go-to-network is a capacity channel. The limit is how many connectors you have mapped and how often each can be asked before they stop saying yes. Sends are not the input; relationships are. So the metrics have to measure the stock of relationships, how well it is being used, and what it produces, not how much activity it generates. The definition of the motion is on go-to-network.
The five metrics
1. Relationship coverage
Definition. The share of target accounts where at least one person on your side has a real relationship with someone at the account. "Real" needs a threshold: a shared employer for a year or more, a board seat, a customer relationship, a calendar history. A LinkedIn connection alone does not count.
Read it two ways. Any path, and path to a decision-maker. The gap between the two is where the work is: plenty of accounts have a path to a manager and none to the person who signs.
What good looks like. It depends on how old the company is and how many customers, investors and partners it has, so the useful number is the trend. A team that maps all four networks for the first time usually finds coverage well above what anyone guessed, because people only remember their own ties.
2. Paths per account
Definition. For covered accounts, how many distinct connectors can reach someone there.
Why it matters. One path is a single point of failure: the connector is busy, says no, or leaves. Three paths is a channel. It also tells you where to invest: an account with one weak path and a $500K opportunity is worth asking a partner or an investor to look at their own network for.
3. Ask acceptance
Definition. Of the introduction requests made to connectors, the share they agreed to make. Count "yes" and "yes, later" as accepted; count "wrong person" as declined with a reason, because it improves the map.
Why it matters. This is the health gauge of the whole motion. Acceptance falls for three reasons: the connector was asked too often, the ask arrived without a drafted note, or the deal was too small for the favour. All three are fixable, and all three are invisible without this number.
What good looks like. Teammates accept most asks. Customers and investors accept a high share when asked rarely and well, and a falling share the moment cadence slips. Track it per network, not blended, or the teammates' yes rate hides the investors' no rate.
4. Intro to meeting
Definition. Of accepted asks, the share that produced a held meeting with the target within a set window, usually 30 days.
Why it matters. It separates connector goodwill from connector standing. A connector who says yes to everything but whose intros do not convert does not know the target as well as the map thinks. Feed that back into relationship strength.
Pair it with cycle time. Days from ask to held meeting. The single most convincing number for a sceptical leader is this one set next to the same number for outbound on comparable accounts.
5. Revenue sourced and influenced
Definition. Sourced: opportunities whose first meeting came from an introduction. Influenced: open opportunities where an introduction added a new stakeholder or moved a stalled deal. Keep the two apart; blending them is how channels lose credibility with finance.
Why it matters. It is the number that funds the channel. Influenced is often larger than sourced, because the most common ask in a mature program is not "get me in" but "get me to the person above my champion".
Two reports worth building
The coverage report, weekly, for the leadership channel. Every opportunity closing this quarter, classified as strong executive path, weak path only, or no path, with the total value in each bucket, the change since last week, and the three largest deals with no warm coverage, each with its owner named. Boomerang users can run this from Claude or ChatGPT over MCP; the prompt is on the MCP page.
The connector scoreboard, quarterly, for the people who open doors. Per connector: asked, accepted, converted. It shows who actually opens doors (often not who you expected), who is being over-asked, and who has stopped responding and needs a thank-you rather than another request. Relationship capital is an asset; this is its balance sheet.
Guardrails
- Per network, always. Blended numbers hide the one network that is being burned.
- Threshold the map. Decide what counts as a real relationship before measuring coverage, or coverage is whatever the loosest definition says.
- Write outcomes to the CRM. If the intro and its result do not land on the account, contact and opportunity, none of the five can be produced reliably, and the channel disappears from every report finance reads. Boomerang writes these back to Salesforce, HubSpot and Attio in the customer's own schema.
- Do not add send volume. It is not an input to this channel, and reporting it invites the comparison that kills the program.
Where to start
If you measure one thing this quarter, measure ask acceptance per network. It is the earliest signal of trouble and the cheapest to collect: every ask gets a yes, a no or a silence, and silence counts as no. If you measure two, add intro-to-meeting with cycle time, and put it next to outbound's on the same slide.
The comparison with cold channels is on go-to-network vs go-to-market, and the tooling categories on go-to-network platforms.
Frequently asked questions
What is the single most important go-to-network metric?
Ask acceptance per network. It is the earliest warning that connectors are being over-asked or asked badly, and it is the cheapest to collect because every ask produces a yes, a no or a silence.
Should go-to-network pipeline be counted as sourced or influenced?
Both, kept separate. Sourced is an opportunity whose first meeting came from an introduction. Influenced is an open opportunity where an introduction added a stakeholder or moved a stalled deal. Blending them costs the channel credibility with finance.
How often should the connector scoreboard be reviewed?
Quarterly for the leadership review, with a monthly glance at acceptance by network. More often than that and the numbers are too small to read.
How does Boomerang produce these numbers?
Every path Rudy finds, every ask and every outcome is written back to Salesforce, HubSpot or Attio on the account, contact and opportunity, so coverage, acceptance, intro-to-meeting and sourced or influenced revenue can be reported next to every other channel. The weekly coverage report can be run from Claude or ChatGPT over MCP.
