Where the money comes from

Chapter 01 · Part I

Where demand comes from

Clients rarely decide to "buy services". They commit to something first: a software licence, a funding round, a growth target, a customer's security requirement, a government mandate, or a system that's falling apart. The services spend follows from that commitment. The trigger is the single most useful way to sort this market. inference

Why the work goes outside

Every case in the research fits at least one of four reasons why building the capability in-house loses: inference

Surveys often cite cost first (59% in Clutch data). Deloitte's 2024 survey of 500+ leaders ranked access to talent as the top reason. The evidence fits "capability and speed first, cost second" for anything above commodity work. primary

Chapter 02

How providers get paid

The pricing structure tells you what is really being sold. Capacity is priced by the hour, a bounded outcome by fixed budget, a continuous function by subscription, a measurable result per outcome. inference

Where in the chain they sit

Whoever owns the buyer when the decision is made keeps most of the margin. Production layers further down compete on cost. inference

Chapter 03

Money trails

Each trail follows the money from the reason it exists to the renewal: driver → budget → buyer → trigger → offer → delivery → outcome → payment.

Platform implementation partner: Coastal Cloud

1. Driver and budget

   A mid-market or enterprise company signs a Salesforce subscription from its IT or software budget.

2. Trigger

   The licence only pays off after configuration, integration and adoption. The CIO or CRO needs certified people to do it.

3. Vendor in the middle

   Salesforce refers and co-sells to certified partners. IDC estimated the partner ecosystem earns ~$5.80 for every $1 Salesforce makes. secondary

4. Offer and delivery

   Multi-cloud implementation, advisory and ongoing services, delivered by 400+ certified staff holding 3,000+ certifications.

5. Payment and expansion

   Project fees, then ongoing services, then more Salesforce products.

6. Capital layer

   Sverica (PE) invested in 2020. TCS bought the firm for $700M cash in December 2025, about 5x its $141M trailing revenue. primary

Why it exists: software vendors earn high margins on subscriptions and low margins on labour-heavy services, and they want to sell into more accounts than their own staff can cover. Handing implementation to partners lets them sell more licences without carrying services margin. The buyer pays again for services because a licence that isn't implemented is wasted money.

Customer-driven compliance: vCISO, pentest, Vanta

1. Trigger from outside

   An enterprise customer's procurement sends a security questionnaire or asks for a SOC 2 report. The startup's deal stalls.

2. Budget appears

   The founder or CTO now has urgent "risk" budget that is really tied to closing the deal.

3. Offer

   A compliance platform (Vanta: $300M ARR, 10,000+ customers), a fractional CISO ($1.5k–$8k/mo), a pentest ($8k–$25k for a typical SOC 2 scope) and a CPA auditor. estimate

4. Channel

   Vanta runs a partner programme for MSPs (since 2022) and auditors, so IT providers resell compliance as a recurring service.

5. Outcome and renewal

   Report issued, deal closes. The audit repeats every year, so the spend recurs.

The key point: a lot of "risk reduction" spending is really revenue spending. The budget comes from the sales pipeline, not the security team.

Solo productized subscription: Designjoy

1. Buyer and trigger

   A founder or head of marketing needs design work every week but can't justify, or wait for, a full-time hire.

2. Offer

   Unlimited requests, worked one at a time, for a flat monthly fee you can pause.

3. Delivery

   One senior designer and an asynchronous request queue. No meetings.

4. Price as a throttle

   The price went up about nine times, from $449 to around $5k–$6k a month, to ration demand. secondary

5. Result

   Reported revenue of $3.1M in 2024 with one person. At ~$5k a month that implies roughly 50 subscriptions running at once. inference

Public-sector digital services: Nava PBC

1. Budget

   Congress funds HHS and CMS. A Medicare data-interoperability policy creates the need.

2. Gate

   A contracting officer controls the award. Access depends on contract vehicles (GSA schedules, IDIQs, BPAs) and small-business set-asides.

3. Awards

   Blue Button 2.0 ($22.6M), the AB2D/BCDA/DPC claims APIs ($34.47M), benefits automation (up to $7.99M). $311.75M across 234 awards in total. primary

4. Renewal

   Option years and recompetes. Past performance compounds.

5. Political risk

   GSA's in-house team 18F was shut down in March 2025 and large consulting contracts were targeted. This budget can disappear for reasons unrelated to delivery. The net effect on small firms is unknown. hypothesis

Nine more trails in brief

Chapter 04 · Part II

Pattern library

Each pattern is labelled by how often and how consistently it shows up across independent examples: structural many quantified examples and a clear mechanism, repeated several independent examples, emerging recent or conflicting evidence.

Chapter 05

Business archetypes

These came out of the evidence rather than being chosen in advance.

Chapter 06

The nine-layer map

Money moves down this stack. At each layer some of it is kept and some passes on, sideways to platform vendors or down to subcontractors.

How it flows: revenue or capital at layer 1 only becomes a budget when a layer-2 problem has an accountable owner at layer 3. That owner pays for an outcome but buys a package, because packages are easier to judge. Providers keep a margin, pass some down to freelancers and white-label shops, and pass some sideways: media spend to Google and Meta, licences to HubSpot and Salesforce, API usage to model vendors. Vendors send part of it back up as revenue share. inference

Chapter 07

Prices, margins, multiples

Price is not profit. The only margin data is benchmark-level (Promethean Research, SPI Research). Single-firm margins are unknown.

What sets willingness to pay

Willingness to pay = closeness to revenue or a deadline × cost of failure × scarcity of credible providers − the buyer's doubt about you

The same capability can sell at very different prices. A healthcare MVP is offered from about $10k by offshore shops and for $50k–$1M+ by US studios. Price follows the buyer's risk tolerance and trust, not the provider's delivery cost. inference

What buyers of firms pay for

• Platform partners: Silverline sold for ~1.75x revenue (2023); Coastal Cloud for ~5x (2025). primary
• Agencies: brokers cite 4.5–5x EBITDA for retainer-heavy agencies vs 3.5–4x for project-heavy ones; ~47% of deals include earnouts. estimate
• MSPs: 169 deals in 2025, mostly PE-backed consolidators; large platforms near 20x EBITDA. secondary

Chapter 08 · Part III

What's really happening underneath

A business pays an outside provider when four things line up. An outcome tied to revenue, an obligation or a deadline creates a budget. Building the capability in-house costs more than renting it. The provider lowers the buyer's doubt enough to be chosen. And the provider has a route to that buyer at the moment of need. Providers then grow their share by sitting closer to the decision, separating revenue from their own hours, and attaching to needs that recur.

Skill, offer, business

Chapter 09

Seven principles

The smallest set that explains almost every case in the research.

1. Commitment creates budget

   Services spending follows something the client already committed to: a target, a platform, a capital event, an obligation or a decaying system.

2. Buy outside when in-house loses

   On intermittent need, urgency, scarce skills or wanting someone else to own the risk. Cost is secondary except for commodity work.

3. Willingness to pay has four parts

   Closeness to revenue or a deadline, cost of failure and scarcity raise it. The buyer's doubt about you lowers it.

4. Margin goes to whoever owns the buyer

   At the moment of decision. Production is passed down the chain at lower prices.

5. Leverage means separating revenue from hours

   Through process, other people's labour, software or taking on principal risk. Otherwise revenue per head stays near $150k–$300k.

6. Durability is a recurring need plus a switching cost plus a renewal trigger

   And buyers of firms pay for it.

7. Platforms create, route and reshape service markets

   Each new platform wave opens a premium for scarce skills at the top while automating the bottom.

Where the principles strain

• Lock-in instead of value. Some SMB providers appear to keep clients through switching costs, for example websites on a proprietary CMS, rather than results. These reports come from competitors, so they may be biased.
• Political shocks. 18F's shutdown breaks the link between need and budget.
• Solo firms that refuse growth. Designjoy uses price to cap demand, not to scale.
• Pattern stops being a service provider and becomes the merchant.
• Unexplained exit prices. Why Coastal Cloud sold for ~5x revenue and Silverline for ~1.75x isn't explained by the evidence here.

Chapter 11

What this page knows

The research covers 50 US providers and 27 market, platform and benchmark data points, gathered in October 2026. The full evidence tables, with buyer, trigger, offer, pricing, channel, money trail and sources for each provider, are kept alongside this site's source.

• How it was gathered. Facts come from search-engine extracts of the cited pages rather than full page reads. Re-check any figure before relying on it.
• Nothing invented. No revenue, margin or price was estimated without a source. Figures like "~5x revenue" or "~50 subscriptions" are simple arithmetic on cited numbers and are tagged as inference.
• Not a census. The sample mixes famous and obscure firms on purpose. "Structural" means a pattern shows up often and consistently, not that it dominates by market share.
• Self-reported numbers (Designjoy, Draft.dev, Lead Cookie) are marked as such.

Open questions

• Are productized and solo firms really more profitable than agencies, or just higher revenue per person?
• Why did similar Salesforce partners sell for such different multiples?
• What did the 2025 federal changes do to small digital-services firms' revenue?
• Is AI shrinking total services spending, or just moving it between categories?
• How large is subcontracting and white-label work beyond SPI's 10.9% figure?

Key sources

• HubSpot FY2024 10-K: ir.hubspot.com; partner programme policies: hubspot.com
• Microsoft/IDC partner study: partner.microsoft.com; Shopify ecosystem: shopify.com; Salesforce/IDC: diginomica.com
• TCS–Coastal Cloud: consultingmag.com; Mphasis–Silverline: mphasis.com; Cognizant–Thirdera: cognizant.com
• Fractional AI acquisition: blackstone.com; Ode with Anthropic: techcrunch.com
• Upwork FY2025: Q4 2025 call; MBO Partners: State of Independence; Gartner: IT spending 2026
• SPI Research: PS Maturity Benchmark; Deloitte: Global Outsourcing Survey; inTandem: SMB survey
• AI and freelancing: bloomberry.com; 18F: siliconangle.com
• Provider pages: Atomic Object, FastRuby.io, Planet Argon, RevPartners, Kruze, Hook Agency, WP Buffs, Draft.dev, Focus Lab, Designjoy
• Public sector: Nava PBC awards, Fearless, City of Tacoma