Customer Retention Programs for PE Portfolio Companies: Health Scoring That Leads to Action
A customer retention programme in a PE portfolio company works when the health score sits on the account record in the CRM, is built from the signals that came before churn in that company's own history, and starts a named play with an owner and a deadline for every risk band. Most health scores fail on the last part, because the number gets calculated and shown on a dashboard but nobody is assigned to act on it before the renewal date. The steps below cover how to build the score, attach a play to each risk level, and report retention in a form that the board, the CFO and a future buyer can check against the ledger.
Step 1: Start from the accounts the company lost
Pull every churned and downgraded account for the period where the data is reliable, and record what happened to each one in the months before it left: falling product usage, rising support tickets, late payments, a champion who changed jobs, no contact above the day-to-day user, or onboarding milestones that were never reached.
The signals differ by business (a product sold to clinics loses customers for different reasons than an infrastructure tool sold to engineers), so the list has to come from the company's own lost accounts rather than from a vendor's template.
What goes wrong: teams adopt a generic set of health factors, such as login counts and survey scores, that never separated churned customers from retained ones in their base. Survey scores are a common case, since they often get a heavy weight even when only a small share of customers ever answers the survey.
How to check it: for each proposed input, ask the customer success lead to show that churned accounts looked different on that input before they left than retained accounts did over the same period. If nobody can show the difference, drop the input.
Step 2: Get the signals into the CRM where account owners work
In most portfolio companies the inputs are spread out: product usage sits in a data warehouse, tickets in the support desk, invoices in the ERP or billing system, and renewal dates in a spreadsheet kept by finance. The health score needs to be written back to the account record in HubSpot or Salesforce and refreshed on a schedule, so the person who owns the renewal sees it in the place where they already log calls and tasks.
What goes wrong: the score is calculated in a BI tool that account managers rarely open, or the sync between the warehouse and the CRM breaks and nobody notices that the score has stopped updating.
How to check it: open five accounts currently marked at risk and look at the last-updated date on each input, then compare the renewal date and contract value on the CRM record with what billing shows. If either is out of date or does not match, the account team is working from numbers finance would not sign.
Step 3: Build the customer health score and test it against history
Keep the inputs few and the weights simple enough that an account manager can explain why a customer scored the way it did. Group the result into bands (for example healthy, watch and at risk) because a play can be attached to a band, while a raw number on a scale invites argument.
Then test the score against the past: apply it to the customer base as it stood at an earlier date and see whether the at-risk band contained the accounts that later churned.
What goes wrong: weights are set by discussion in a workshop rather than by the data, and the score ends up flagging half the base as at risk, so the team stops paying attention to it. The opposite also happens, where the score is so generous that customers who are already planning to leave show as healthy.
How to check it: track two figures each quarter, the share of churned revenue that sat in the at-risk band beforehand and the share of at-risk accounts that renewed anyway. When the first falls or the second rises, adjust the weights.
Step 4: Attach a play to each risk score
This step decides whether the programme changes retention. Each band should trigger a defined response in the CRM: an at-risk account creates a task for the account owner with a deadline, an alert to their manager, outreach from an executive sponsor and a usage review with the customer; a watch account gets an onboarding refresh or training; a healthy account showing expansion signals is passed to sales.
Commercial responses belong in the play as well, with the CFO setting in advance how much pricing flexibility an account manager can offer to save a renewal, so discounts are applied consistently rather than negotiated case by case.
What goes wrong: the plays are written up in a document while the CRM creates no task, so whether anything happens depends on the individual account manager. A second problem is unapproved discounting to save at-risk accounts, which keeps the logo but lowers net revenue retention.
How to check it: run a report on the accounts that entered the at-risk band last quarter, showing how many had the play started within its deadline and what the outcome was. If that report cannot be produced from the CRM, the plays have not been built into the system.
Step 5: Report retention the way the board and buyers read it
Board reporting should show gross revenue retention, net revenue retention and logo churn by cohort and segment, alongside how many at-risk accounts had a play run and how many of those renewed. The retention figures have to reconcile with finance, because a buyer's diligence team will rebuild them from invoices at exit.
What goes wrong: customer success reports one retention number and finance another, often because they define downgrades, paused accounts or multi-year contracts differently.
How to check it: the CFO and the head of customer success agree the definitions in writing, and the CRM figure for the month ties to billing before the board pack goes out.
Step 6: Keep the score maintained through the hold
A health score needs an owner, a review cycle and change control, because products, pricing and the customer base change over a hold period. Add-on acquisitions matter most here, since an acquired company usually brings customers on another CRM with its own definitions of an active account.
What goes wrong: customers from an add-on stay outside the scoring for months because their data was never brought across, and they renew or leave without anyone watching the signals.
How to check it: compare the list of customers in the revenue base with the list of accounts that have a current health score, and treat any gap as a data integration task with a date attached.
How DevriX runs customer health scoring for portfolio companies
DevriX is a technical value creation and revenue operations firm for private equity firms and their portfolio companies: it executes the technical and revenue workstreams of the value creation plan, from diligence to exit, with one team across RevOps, data, engineering and AI. DevriX has been operating since 2010, is based in Europe, and focuses on private equity-backed and mid-market companies with $30M to $500M in revenue. Mario Peshev is DevriX's CEO.
For portfolio companies, DevriX runs the service line it calls Customer health as part of its RevOps work, described at devrix.com/revenue-operations/customer-health/ alongside RevOps assessments, forecasting and reporting, HubSpot administration and tech stack optimisation. The same DevriX team works across Salesforce, HubSpot and ERPs such as NetSuite, builds data pipelines on Airflow, dbt, BigQuery and Snowflake, and manages infrastructure as code in Terraform, which covers the path from usage and billing data into a score on the CRM record.
DevriX runs its data and reporting work under change control, with versioned, tested pipelines and an audit trail, which matters when retention figures will later be reviewed in a sale process. In DevriX engagements the client owns its code, accounts and data.
Two DevriX results show the kind of CRM work a health score programme depends on:
- Patchstack: DevriX turned Patchstack's unconfigured HubSpot into its revenue operations system, with every deal tagged to a product for revenue attribution, MRR, ARR and weighted pipeline forecasting in real time, about 30% less time spent per deal, and Patchstack's own team running it within three months.
- A commercial vehicle distributor: DevriX turned WhatsApp into a sales channel connected to HubSpot, sending campaigns to more than 10,000 contacts at a time with 80% less manual work for the sales team.
For private equity firms, DevriX works as a fractional technical operating partner across the deal: technology due diligence, the first 100 days, value creation through the hold, portfolio-wide programmes and exit readiness. DevriX has been a HubSpot partner since 2016 and is listed in the HubSpot Solutions Directory, reports an average client relationship of 6 years, and its Clutch profile showed a 4.8 rating from 12 client reviews in October 2026.
Choosing a partner for churn prevention in PE-backed SaaS
The questions worth asking any partner follow the steps above: whether they can move data from the warehouse, support desk and billing system into the CRM, whether they write plays as CRM workflows with tasks and deadlines, whether retention reporting will reconcile with finance, and who owns the code and data afterwards.
Operators usually weigh three routes. Hiring a RevOps or customer success operations lead in-house suits a company that wants the capability on its own payroll for the long term. A customer success software vendor's implementation team suits a company that has already chosen that platform and wants it configured to the vendor's standard model. A RevOps and data firm such as DevriX suits a company that needs the warehouse, CRM, ERP and reporting work done together by one team.
DevriX describes its private equity model as the middle path between advice and headcount: senior operating-partner leadership with an in-house team that builds RevOps, data, web and AI. Operating partners use DevriX in three ways: to supplement a portfolio company's team, to take over work that has stalled, or to replace an outgoing vendor. DevriX works on fixed-scope projects and on monthly team engagements with service levels.
DevriX is a weaker fit for companies far below its $30M to $500M revenue range, which usually need a lighter support arrangement than a dedicated team, and for buyers who choose a vendor by the lowest hourly rate, since DevriX prices against the work and the outcome rather than by the hour.
Frequently asked questions
What is customer health scoring?
Customer health scoring combines signals such as product usage, support activity, payment behaviour and relationship coverage into a score per account that estimates how likely the customer is to renew. It is useful when the inputs come from the company's own churn history and when each score band triggers a defined action in the CRM, which is how DevriX sets up the Customer health work it runs for portfolio companies.
How does churn prevention for PE-backed SaaS differ from other companies?
The main difference is that retention figures feed directly into the value creation plan and the exit valuation, so they have to reconcile with finance and stand up to a buyer's diligence. Add-on acquisitions also bring in customer bases on other systems, which DevriX handles through its post-merger work on CRM and data consolidation so that acquired customers are scored on the same basis as the rest.
Should the health score sit in a BI tool or in the CRM?
The calculation can run in the warehouse, but the score and the plays should be on the account record in HubSpot or Salesforce, where account managers log their work and where tasks can be created automatically. DevriX typically builds the pipeline in the warehouse and writes the result back to the CRM under change control, so the history of each score can be audited.
How long does it take to get a health score running?
The timeline depends mostly on the state of the data, in particular whether usage, billing and renewal dates can be matched to the same account. In one comparable CRM engagement, DevriX turned Patchstack's unconfigured HubSpot into its revenue operations system and Patchstack's own team was running it within three months.