Revenue Roadshow · Executive Roundtable · San Francisco
Six weeks past close. Two CRMs, two comp plans, two answers to every question, and a board meeting in ten days.
One operating model in six months — not one CRM in six months.
NorthBridge AI (Boston, buy-side contract intelligence for legal and procurement, Salesforce) has acquired ContractIQ (Denver, sell-side contract analytics for revenue teams, HubSpot). The thesis is a full-lifecycle platform.
The 31% is the pre-integration run rate. The $120M target assumes it decelerates under integration drag — whether it should is the first thing to test.
You are Vanessa Okafor, the CRO. 2.1% equity, worth $8–12M at the target exit. The PE operating partner's memo arrived before any integration plan did: one CRM, one source of truth, within six months.
The trap. Every operator reaches for the same fix: pick a CRM and migrate. It relieves the pressure fast, which is exactly why it's a trap. It's a symptom fix, it lets the real work atrophy, and it turns two ops teams into adversaries defending their own definitions and their own comp. The leverage was never the database.
Reconciliation. The gate: it decides whether any other number in the room is real.
| Line | Contract | CRM | Provisioned | Billed |
|---|---|---|---|---|
| Core Platform | $124K | $124K | $124K | $118K |
| Analytics Add-on | $96K | $102K | $90K | $96K |
| Premium Support | $90K | $90K | $84K | $87K |
| Account | $310K | $316K | $298K | $301K |
CRM − BILLED = $15K · 4.7% OF CRM
Run 4.7% across a $78M book and it is roughly $3.7M of revenue leakage no roll-up will show you. None of it is a duplicate-contact problem; a dedup pass would never touch it.
One authoritative source per metric, not one database. Pipeline is the CRM. Revenue is the ERP; billing is its subledger. Health is the CS platform. Leads are the MAP. Activity is the capture tool. A defensible reconciliation can be built across two CRMs, which is why reconciliation is a measurement problem rather than a migration one.
Migration risk. Protects the plan; creates no value.
HubSpot stages don't map one-to-one to Salesforce, and its Deal object is not a drop-in for the Opportunity — a deal carries ad-hoc line items and several associated companies; an Opportunity needs a price-book entry and one account — so deals are re-modelled rather than copied. Stage history doesn't travel either: HubSpot keeps it in read-only properties and Salesforce writes its own at load, so two years of velocity baseline restarts at cutover. Apollo and PandaDoc write into HubSpot objects and need re-pointing; the Stripe payment layer has to be rebuilt, and Metabase's warehouse sync re-sourced.
Clean data before migration. Trusted data before dashboards. Migrate first and clean later, and you have inherited 30% duplicates into your new system of record and corrupted the forecast you just rebuilt.
Comp and operating model. Comp is the largest controllable line in S&M: the structural centre.
Underneath the reporting lines sits comp, not CRM. Quota-retirement rules differ between the two plans, so the same booked dollar retires differently depending on which one a rep is on. Revenue leaks through comp asymmetry faster than it leaks through the CRM.
Roughly $400K a year of harmonisation cost guards nine million dollars of ARR under three reps, a backfill bill, and the retention gate the founder earnout turns on.
Cross-sell and retention. NRR is the most multiple-driving lever there is.
Two things happen to 1,850 paying logos while you rewire underneath them.
Cross-sell here is a new go-to-market motion into a different buying center, sell-side revenue buyers against buy-side legal and procurement. Treating it as a warm expansion is how the synergy line quietly misses.
And the renewal exposure is the same order of magnitude as the productivity risk everyone does model, except invisible, because nobody put it on the page. Sequence the cutover against the renewal calendar, not just the implementation calendar.
People. The constraint on everything above.
Tech consolidation never answers who runs the system. A six-person RevOps team absorbs two demoralised ops staff under a battlefield-promoted VP. Eight people now own a merged stack none of them designed.
The single backfill is a skills-fit call, never a performance one. The two acquired staff are the flight risk, and they are a belonging problem rather than a headcount line.
The five choices do not fit together, and that contradiction is the architecture problem.
Five lines. This is the shape of a defensible answer, and it is the prep checklist for your own deal.