Metrics, customers and the organization · Answered by Golden Section from more than 400 B2B software companies observed
Yes, when services deliver the outcome the software promises and are sold as repeatable packages; no, when they replace the product. In vertical software, implementation, training and change management are often how domain expertise reaches the customer, and they are the part an internal IT group cannot copy. Price implementation from tracked hours, report services on their own revenue and margin lines, and never sign hourly not-to-exceed contracts. Judge services on customer outcomes and retention first, and on margin second. An engagement at modest margin that puts a customer in the top quartile of his field is worth more than a high-margin one that leaves him at the median, so treat margin as a floor to protect. Services at 40% of revenue is heavy; if customers spend more time with your team than your product, you have become a services firm. Start by tracking implementation hours by customer.
Offer services that are packaged, priced from measured hours and tied to a customer outcome, and keep them on their own revenue and margin lines. Hold services margin above a written floor, judge the function on whether engaged customers retain and expand better than the rest, and stop any service that substitutes for product work.
| Metric | Value | What it means | Source |
|---|---|---|---|
| Outcome vs margin | 25 points of margin that reaches top quartile beats 55 that leaves the customer at the median | services gross margin compared with the customer result it producesmargin is a floor to watch, not a number to maximize, where the moat is domain mastery | Golden Section playbookValue Pinnacle Services play |
| Services-heavy threshold | 40% of revenue | services revenue ÷ total revenueat this share a software company starts to look like a services firm to investors and buyers | Golden Section operating viewGolden Section operating view |
| Gross margin we look for at investment | above 65%, path to 75%+ | blended gross margin after implementation and support costsservices sit inside this figure, so a heavy, low-margin services mix pulls it down | Golden Section, publishedGrowth Equity |
| What counts as ARR | committed, fixed recurring fees only | excludes one-time implementation and consulting feesreport services revenue beside ARR, never inside it | Golden Section playbookARR Schedule play |
| Implementation phases | 5, each gated by client sign-off | planning and requirements, design and configuration, user acceptance testing, deployment, close-outfixed-scope packages are only possible when the phases are standard | Golden Section playbookImplementation Project Lifecycle play |
Vertical software usually needs services, because the customer is changing how his business runs, not just installing a tool. We underwrite total annual revenue rather than ARR alone for that reason. But services are also how software companies quietly turn into consultancies: each custom engagement looks profitable, the team grows to match, and one day the product is a side effect of the services. We treat services at 40% of revenue as heavy, pass on services-heavy models at investment, and buyers price them down at exit.
The line between the two is packaging. The value pinnacle services play builds small, repeatable packages against a published maturity rubric, sold many times at consistent prices, and tracks the share of services revenue that is packaged rather than bespoke. Implementation hours tracking prices those packages from data, and the implementation lifecycle makes them predictable enough to fix the price. For customers whose commitment is uncertain, core and provisional segmentation uses a paid services engagement before any subscription as the test.
On profitability, our view is that services exist to make the subscription succeed. Keep a floor so services never subsidize customers, but optimize for the outcome. If the engaged cohort does not outperform the rest of the base on net retention, the services are not working, whatever their margin.
The numbers are fictional. A $5M company earns $1.2M from services, about a quarter of revenue, almost all of it scoped from scratch, at a 12% margin. Tracking hours for a quarter shows most implementations follow one of three patterns. The company turns them into fixed-price packages priced from the measured hours, adds two packaged services tied to a five-level maturity rubric, and moves services onto its own revenue and cost lines. It stops quoting bespoke work over a set size. A year later services revenue is slightly lower, its margin is 30%, most of it is packaged, and customers who bought a package retain noticeably better than those who did not.
Products simple enough for customers to set up alone should productize onboarding rather than sell services. Where a strong implementation partner ecosystem exists, letting partners deliver services can be right, provided someone tells the customer which costs he would incur whether he built or bought.
From the Golden Section mistakes list, each paired with the play that prevents it.
Services that replace product turn a software company into a consultancy.
The opposite error: starving services by optimizing for margin when services are how domain mastery reaches the customer.
Hourly not-to-exceed contracts put all the overrun risk on you and cap the upside.
A partner's services quote can make your software look expensive unless someone separates costs the customer incurs either way.
In the order we would run them. Each is on its own page, most with a free Excel template.
Tracks implementation hours by customer so services are priced from data.
Standardizes implementations into gated phases that can be sold at a fixed price.
Packages services against a maturity rubric and judges them on outcomes.
Uses a paid services engagement to test commitment before a customer joins the core book.
Customer plays Implementation, onboarding and services plays that keep services in support of the product.
Profitable enough that services never subsidize customers, with a written margin floor, and no more than that as a goal. Services in vertical software exist to deliver the outcome the subscription promises, so judge them first on whether engaged customers retain and expand better. A margin well below your floor usually means the package is mispriced or badly scoped, so fix the package rather than the outcome.
No. ARR should include only committed, fixed recurring fees. Report implementation and services revenue on their own lines, with their own margin, beside ARR, because buyers will separate them anyway.
Usually, when implementation takes real work. Price it from tracked hours plus margin, sell it as a fixed package where you can, and make it visible at kickoff rather than during the first scope argument. A free implementation teaches the customer that the work has no value.
We underwrite annual revenue including services because durable vertical software carries it, and we pass on services-heavy models, where services reach about 40% of revenue. A company moving its services from bespoke to packaged is making exactly the kind of change that raises what both investors and buyers will pay.
Growth equity →Reviewed by Dougal Cameron, CEO & Co-Founder on 2026-10-22. Golden Section observations are labeled separately from external benchmarks and illustrative arithmetic.