What this system is
It is two connected loops. The acquisition loop nurtures good-fit leads in HubSpot until they show buying intent, then hands them to Sales with context and a deadline. The customer loop brings product usage from Mixpanel onto the HubSpot company record, so customer success and account managers act on what customers actually do.
Both loops run on the same idea: write the definition once (what is an MQL, what is a dormant seat), then let workflows enforce it.
- Nurture workflow
- A HubSpot workflow that sends a sequence of emails and actions to enrolled contacts, with branches by persona and an exit when they qualify or go cold.
- Fit score
- Points for who the contact is: industry, company size, role. Disqualifiers such as personal email addresses subtract points.
- Engagement score
- Points for what the contact did, such as clicks, webinars and page depth. It should decay over time so old activity stops counting.
- Intent signal
- A high-intent action that shows buying interest, such as repeat pricing page visits, an ROI calculator, a demo video or a trial signup.
- MQL (marketing qualified lead)
- A lead that meets the agreed definition and is ready for Sales. Here: score of 60+ and at least one intent signal in 14 days.
- Handoff SLA
- The agreed time Sales has to work a new MQL. Two business days is a common standard, tracked per owner.
- Group analytics
- A Mixpanel feature that analyzes events by account instead of by user, using a group key such as the HubSpot Company ID.
- Cohort sync
- Sending a Mixpanel cohort (a saved group of users or accounts) to HubSpot as a list or property so workflows can act on it.
Why it matters: the symptoms it fixes
- Sales says the leads are bad, and nobody can point to a written MQL definition.
- Everyone gets the same drip, whatever their role, so nobody reads it.
- MQLs sit untouched for days, and there is no report showing who let them sit.
- Leads Sales rejects disappear instead of going back into nurture with a reason.
- Customer success learns about risk at renewal time, because product usage lives in a tool CS never opens.
- Expansion is reactive: customers at seat capacity or doing manual workarounds are never flagged to an account manager.
The playbook, step by step
Build in this order. Definitions come first because every workflow after them is just the definitions, automated.
Agree lifecycle stages and the MQL definition
Write down each lifecycle stage (Subscriber, Lead, MQL, SQL, Opportunity, Customer) and the exact rule that moves a record between them. Get Sales and Marketing leaders to sign the MQL rule before you build anything.
Done when: a one-page definition both teams have signed.Build the lead score with fit, engagement and intent
Score fit (industry, size, role) and engagement (clicks, events, page depth) separately, with engagement decaying over time. Define a short list of intent signals and require one of them for MQL on top of the score threshold.
Backtest the score against the last two quarters of closed-won deals before turning it on.
Done when: the score separates won deals from lost or unworked leads in the backtest.Set strict enrollment criteria
Enroll contacts at Lead stage with Fit grade A or B, who are not customers and have no open deal. Turn on re-enrollment for leads Sales recycles, so they come back into nurture instead of going cold.
Done when: every good-fit lead is in exactly one nurture track, and no bad-fit lead is in any.Write one track per persona, one job per email
Branch on a Persona property. Give each email a single job: name the problem, show the product, share peer proof, or move the reader to an intent signal. Measure each email by the MQLs it creates, and rewrite or cut the weakest every quarter.
Done when: each email has a written job and an MQL-attribution number.Automate the MQL handoff
At the end of each track, branch on the MQL test. On yes: set lifecycle to MQL, rotate an owner, add the contact to the HubSpot–Salesforce sync, and create a task due in two business days. On no: keep nurturing, and after a set number of emails with no intent, move them to a newsletter list.
Done when: MQLs appear in Salesforce with an owner and an SLA task within minutes.Measure the handoff and close the loop
Build a dashboard with SLA compliance, median time to first touch, MQL→SQL rate, aged MQLs by owner, and Sales disposition (accepted, recycled, disqualified) with reasons. Review it weekly with both teams.
Done when: SLA is reviewed every week and recycled leads carry a reason.Send product usage to the account with Mixpanel group analytics
Track core product events in Mixpanel and set up group analytics keyed to the HubSpot Company ID. Roll up seats bought vs active users, last active date and feature adoption to the company level, then sync those values to HubSpot company properties.
Done when: usage properties are populated on at least 95% of customer companies.Turn cohorts into plays, with guardrails
Define cohorts such as dormant seats over 25%, heavy manual workaround with no add-on and at seat capacity. Sync them to HubSpot daily and route each to an owner: renewal-risk flag and adoption call for the CSM, cross-sell task for the AM, or an Expansion opportunity in Salesforce. Add guardrails: no duplicate tasks, one play per account per 30 days, and no marketing email while a support ticket is open.
Done when: the first month of plays has been reviewed with CS and AM, and the rules adjusted.
What good looks like
Use these thresholds to grade your current funnel and customer signals. They are JayOh working targets for B2B SaaS teams with a sales-led motion.
| Measure | Weak | Acceptable | Strong |
|---|---|---|---|
| MQLs worked within SLA (2 business days) | Under 60% | 60–89% | 90%+ |
| Median time to first sales touch | Over 3 days | 1–3 days | Same business day |
| MQL definition | Score only, unwritten | Written, score only | Signed, score + intent |
| Recycled MQLs with a reason | None | Some | All, re-enrolled automatically |
| Customer companies with usage data in CRM | None | Under 80% | 95%+ |
| Lead time on renewal risk | Found at renewal | 60–90 days out | 120+ days out |
Common mistakes
- MQL on score aloneNewsletter readers rack up points without buying intent. Require one intent signal on top of the threshold.
- One drip for everyoneOwners and operators buy for different reasons. Branch by persona or nobody reads it.
- No SLA, or an SLA nobody measuresWithout a deadline and a report, MQLs age out. Put the clock on the task and the dashboard.
- Rejected leads vanishIf Sales can't recycle a lead with a reason, marketing never learns what "bad" means.
- Usage data in a tool CS never opensCharts in Mixpanel don't change behavior. Cohorts synced to the company record do.
- Plays without guardrailsWithout suppression rules, one account gets five tasks and three emails in a week.
Frequently asked questions
What is a good MQL definition in HubSpot?
A good MQL definition combines fit, engagement and intent: for example, a lead score of 60 or more plus at least one intent signal (repeat pricing page visits, an ROI calculator, a demo video or a trial) in the last 14 days. Write it down and have Sales and Marketing leaders both sign it.
What lead score threshold should trigger an MQL?
There is no universal number. Pick a threshold by backtesting your score against the last two quarters of closed-won deals, then adjust so that most won deals would have qualified and few unworked leads would have. Scores only mean something relative to your own model.
What is a reasonable MQL to SQL handoff SLA?
Two business days is a common standard, with a target of 90% or more of MQLs worked within it. Faster is better: the median time to first touch should ideally be the same business day.
How does Mixpanel data get into HubSpot?
Mixpanel can sync cohorts to HubSpot through its integrations, and account-level data needs group analytics with a group key that matches a HubSpot identifier such as the Company ID. Some teams route usage through a data warehouse and a reverse-ETL tool instead. Either way, the goal is usage properties on the HubSpot company record.
What product usage signals predict churn?
Common early signals are a high share of dormant paid seats, falling weekly active users, declining use of the core workflow, and a drop in logins by the admin or champion. Define each as a cohort with a threshold, such as dormant seats over 25% for 30 days.
What usage signals indicate an expansion opportunity?
Customers at seat capacity for several weeks, heavy use of a manual workaround that a paid add-on automates, and new teams or locations appearing in the product are strong expansion signals. Route them to the account manager, not marketing.
Should nurture leads sync to Salesforce?
Usually only at MQL. Syncing every lead clutters Salesforce and makes sales reports noisy. Use a HubSpot inclusion list such as lifecycle stage of MQL or later to control what syncs.
How long should a nurture sequence be?
Long enough to give a good-fit lead several chances to show intent, typically four to eight emails over a few weeks. After that, move non-engaged leads to a lower-frequency newsletter rather than emailing them forever.
Want nurture, handoff and usage signals built for your team?
JayOh writes the definitions with your sales and marketing leaders, builds the HubSpot workflows and Salesforce sync, and connects product usage to the accounts your team already works.
Related playbooks
HubSpot, Mixpanel, Salesforce and related marks are trademarks of their respective owners. JayOh is not affiliated with or endorsed by them unless stated. Walkthrough screens are illustrative recreations for education only. Rates, thresholds and benchmarks are JayOh working guidance, not vendor defaults; set your own with your leadership and finance teams.