Creative studios sell judgment. Too many also rent a second career as SaaS administrators. Notion boards, Webflow seats, HubSpot contacts, and a Zap that posts Slack when a form fires. Here is a clean split: what to own, what to rent, and how to price the difference with the own vs rent comparator and stack cost calculator.
01 · The typical agency rent stack
What studios usually pay
| Project management SaaSAsana, Monday, Notion, per seat | ~$50–$200 |
| CRM seatsHubSpot, Pipedrive, often underused | ~$100–$400 |
| Studio marketing siteWebflow, Squarespace: CMS you rent | ~$50–$300 |
| File delivery / proofingFrame.io, Dropbox Business | ~$30–$150 |
| Email marketing toolMailchimp, ActiveCampaign | ~$50–$200 |
| Loaded monthly rent sketch | ~$280–$1,250+ |
| Three-year rent totalLow-to-mid range × 36 months | ~$10,000–$45,000 |
That range is before a mid-cycle tool switch, a new PM system the team decides to try, or a CRM tier upgrade when the contact list crosses a threshold. Three-year totals at the higher end of a mid-size studio clear $45,000 on tools alone, software the studio owns none of.
02 · Own: pipeline and presence
The studio OS you should control
Three things belong on the owned side of a creative agency’s infrastructure.
The studio site. Your portfolio is the primary sales asset. If it lives in Webflow and you cancel the account, the case studies go with it. Own the content store. An Astro site on Cloudflare costs near-zero to host after delivery, serves fast from any geography, and the code lives in a repository you control. Every case study, every SEO ranking, every backlink is on infrastructure you can migrate at will, not infrastructure that migrates you when pricing changes.
The deal CRM. Agency pipelines are not generic. The stages are lead → proposal sent → revision → production → invoice → retainer, and the notes on each deal include creative brief decisions that matter six months later. A generic CRM rented at $100–$400/month gives you an object model designed for SaaS sales. An owned CRM gives you the fields your PMs actually fill in. When the deal CRM matches how the studio operates, PMs stop being middleware between tools.
The proposal-to-project handoff. The moment a signed proposal converts to an active project is where studios lose the most coordination time. If the proposal lives in a PDF sent by email and the project starts in Notion, someone is copying client details and brief notes by hand. Own enough structure in the handoff that this step is automatic: the approved proposal creates the project record, client details flow with it, and the PM does not start day one re-entering information the salesperson already collected.
03 · Rent: commodities at the edge
Tools that should stay replaceable
Ad platforms and media buying tools. Google, Meta, LinkedIn, and their native tooling are not infrastructure. You use them on the client’s behalf and the platform relationship is often the client’s anyway. Rent freely here. These are not strategic assets.
Design suites. Figma, Adobe CC, and equivalents are tools your team is trained on and competitive pressure keeps pricing rational. They are commodity collaboration tools, not strategic infrastructure. Rent them, use them, and replace them when a better option appears.
Short-lived experiment tools. Every studio picks up a new tool for every new type of project: a video tool, a 3D renderer, a new analytics platform. These belong on the rent list and in the audit checklist for annual review. If you have not used a tool in 90 days, it is a cut candidate in the next billing cycle.
04 · The seat math
Thirty-dollar seats add up fast
Agency stacks do not usually explode on one large invoice. They die by accumulation: $30 for this tool, $15 for that one, a $50 tool someone on the team added last quarter and nobody cancelled. The pattern repeats across project management, CRM, file storage, time tracking, and email, and each tool carries multiple seats.
| Rented stack (mid estimate)$600/mo × 36 months | $21,600 |
| Mid-cycle tool switches and migrationsAverage one major switch per 18 months | $2,000–$5,000 |
| Three-year rent total | ~$23,600–$26,600 |
| Owned studio OS buildSite + deal CRM + proposal handoff | $15,000–$35,000 |
| Run rate post-build$0–$100/mo hosting × 36 months | $0–$3,600 |
| Three-year owned total | $15,000–$38,600 |
The ranges overlap in the middle. At the lower end of both, rent and own are comparable over three years. At the upper end, a fuller studio build versus a fully-loaded rented stack, owned is significantly cheaper. The stack cost calculator and own vs rent comparator let you run this on your actual numbers rather than illustrative midpoints.
05 · What clients want and what you need
Client rent is not studio rent
Clients frequently request specific platforms. “Build it in Webflow.” “We want to use WordPress.” “We need it to work with our HubSpot.” These are legitimate client constraints and your job is to deliver on them.
But the client’s platform preference is not a constraint on the studio’s own infrastructure. You can deliver work on Webflow for clients without running your own studio site on Webflow. You can work inside a client’s HubSpot without your own deal pipeline living in a HubSpot account. Client rent and studio rent are separate decisions made for separate reasons.
The studio OS (site, CRM, handoff flow) can be fully owned while every client deliverable is built on whatever platform the client specifies. One is your infrastructure; the other is their infrastructure, for which you are providing a professional service. These do not have to be the same stack.
Conclusion
You sell judgment, not SaaS seats
The creative studio’s competitive advantage is not the tools. It is the thinking behind the work. Tools that compound complexity, require a monthly fee to stay live, and hold your data hostage on cancellation are a tax on that advantage.
Own the studio OS. Rent what is genuinely commodity. Run the three-year math on your seat list and you will find the crossover between renting and owning is closer than the individual invoice lines suggest.