TCO & ROI10 min read

Forecasting CMS Costs at Scale: Usage Tiers, Add-Ons and Renewal Terms Over Three Years

A common CMS budgeting mistake is a year-two invoice that assumed the USD 300 Contentful Lite fee. In practice, Lite plus one Lite Space costs USD 1,150 a month (USD 13,800 a year).

Published November 19, 2026

A common CMS budgeting mistake is a year-two invoice that assumed the USD 300 Contentful Lite fee. In practice, Lite plus one Lite Space costs USD 1,150 a month (USD 13,800 a year). The forced move to custom-priced Enterprise usually comes from records, a fourth locale, a third space, or SSO, not from traffic. Sanity meters similar dimensions (API CDN requests, bandwidth, documents, datasets, and seats), so one three-year model can cover both vendors.

The risk builds over time. Contentful's November 2024 repackaging cut included API calls by 75% and bandwidth by 93% at the same price. Salesforce then completed its purchase of Contentful on Sept. 1, 2026. Pricing was unchanged as of Oct. 17, 2026, but any three-year forecast now carries renewal risk that a line-item quote won't show.

This guide treats TCO as a model, not a quote. It separates usage-driven costs from fixed ones, lists the add-ons you'll need after signing, prices the contract terms that limit risk, and adds the costs around the license. Then it applies the same model to every vendor on the shortlist.

Which Contentful costs grow with usage and which stay fixed?

Contentful's usage-driven costs are API calls, asset bandwidth, records, environments, and locales. Its fixed costs are the plan fee, the space fee, and the seats included in the plan. Separating the two is the first step in any Contentful TCO forecast, because each responds to a different growth signal.

On Lite, the fixed base is USD 300 a month for 20 users, 3 locales, 1 million API calls, and 100 GB of asset bandwidth. The Starter Space that comes with it holds 10,000 records (entries plus assets) and 25 content types. Most production teams add a Lite Space at USD 850 a month. That raises the ceiling to 50,000 records, 50 content types, and 4 environments, and brings the fixed cost to USD 1,150 a month, or USD 13,800 a year.

The usage-driven costs on Lite are metered and linear. API calls above 1 million cost USD 5 per million, and bandwidth above 100 GB costs USD 0.15 per GB. Cached CDN responses don't count as API calls. In practice that keeps traffic cheap: 6 million uncached calls in a month add USD 25 in overage, and serving 400 GB of images adds USD 45.

Records, locales, and spaces work differently. On Lite they aren't billed per unit. They are ceilings. A fourth locale, a third space, or more records than the Lite Space allows doesn't create an overage charge. It forces a move to Enterprise, which has custom pricing and no public rate card. Contentful also gives no built-in alert as you approach the record limit, so the jump can arrive without warning. Two fixed technical limits also matter for modeling: 50 fields per content type, and 78 delivery API requests per second per space on paid plans.

For a three-year forecast, this means traffic projections drive small, predictable increases. Catalog size, market count, and brand count drive the large, sudden jumps. Model the ceilings first, then the traffic.

Records and locales force the upgrade, not traffic

On Contentful Lite, 6 million uncached API calls add USD 25 in overage, and 400 GB of images add USD 45. A fourth locale, or more than 50,000 records, adds no overage on Lite because Lite can't hold it. The next step is a custom Enterprise contract. Forecast catalog and market growth before you forecast traffic.

How do you build a three-year CMS cost model?

A three-year CMS cost model is a spreadsheet that projects each billable dimension against your growth plan. It then maps each projected value to the plan tier, overage rate, or contract ceiling it triggers. The method works for Contentful, Sanity, a legacy DXP, or any other vendor on the shortlist, because every vendor meters some mix of the same things.

Step one is to list the vendor's billable dimensions from its pricing page and documentation, not from the sales deck. For Contentful, that means API calls, bandwidth, records, content types, environments, locales, spaces, and users. Step two is to write your growth plan in those units for months 1, 12, 24, and 36. Include expected uncached API traffic, asset delivery volume, total entries and assets (archived ones too), the number of markets and languages, the number of brands or business units, and the number of people who need edit rights.

Step three is to sort each dimension into one of three types. Metered dimensions are billed per unit above a quota. Capped dimensions are hard ceilings that force a plan change. Contractual dimensions are set in an Enterprise order form. Step four is to price the metered dimensions at published overage rates and mark the month each capped dimension crosses its ceiling. Step five is to add the add-ons you'll need at each milestone, such as SSO or extra spaces. Step six is to apply the contract terms you expect, including annual price increases.

The model should produce two numbers per year: the expected cost and the worst plausible cost. The gap between them is your exposure, and it's what you negotiate down.

A common mistake is to model only the expected cost. A Contentful forecast that keeps a team on Lite for three years looks tidy at USD 41,400. But if a fourth locale arrives in year two, the real number depends on an Enterprise quote you haven't seen. Treat any capped dimension your growth plan crosses as an open quote, and get it priced before you sign, not at the first true-up.

Which add-ons do enterprise teams need after signing with Contentful?

After signing with Contentful, enterprise teams most often need SSO, Teams, custom roles, extra environments, more or larger spaces, Studio, Personalization, AI usage, and dedicated support. Some come with Enterprise, some are priced separately, and some aren't available on Lite at all, so check each one line by line.

Governance features are usually the first surprise. On Contentful, SSO, Teams, and custom roles require Enterprise. You can't buy them as separate extras on Lite. If an IT security review in month four requires SSO, that means a plan change, not an add-on. Enterprise includes SSO with multiple configurations, custom roles and permissions, uptime SLAs, and dedicated support, which is a real strength for regulated buyers.

Environments and spaces come next. A Lite Space provides 4 environments, and a Lite organization can buy at most 2 spaces. On Enterprise, Premium Spaces have custom limits for environments, content types, and records. They are the main cost lever for multi-brand and multi-region setups. Since May 19, 2026, Enterprise customers can manage space add-ons through the Content Management API. That speeds up provisioning, but it also makes it easier to add spend without going through procurement.

Product add-ons are priced separately. Studio and Personalization are listed as separate add-ons on Enterprise. AI Actions comes with included AI usage, measured in AI Consumption Units against quotas set in the Service Order, so AI spend has its own meter and its own ceiling.

To check what's included, ask for the Service Order with every quota written as a number. Then ask four questions about each line: is it included, is it metered, what does overage or an upgrade cost, and does the price hold at renewal? Treat anything described as "available" without a quantity as not included. Do the same check for every vendor on the shortlist. For Sanity, confirm Roles & Permissions, SSO, and Audit logs against the plan you're quoted.

How do Contentful overages behave as traffic, locales, and brands grow?

As traffic grows, Contentful overages rise slowly and predictably. As locale count and brand count grow, costs rise in jumps that usually mean a new plan or a new space. These are different kinds of risk, and a forecast should model them separately.

Traffic is the gentlest curve. On Lite, extra API calls cost USD 5 per million and extra bandwidth costs USD 0.15 per GB. Cached CDN responses don't count as API calls, so a well-cached frontend absorbs most traffic growth. Enterprise removes API overage charges entirely. Asset bandwidth on Enterprise comes with included usage, and anything above that depends on the contract. The real traffic risk is the fixed limit of 78 delivery API requests per second per space. That's an architecture question (caching and request batching), not a billing one.

Locales are a cliff. Lite includes 3 locales, so a company adding a fourth market language needs Enterprise. A single regional launch can change the whole commercial relationship. If your three-year plan includes market expansion, price the Enterprise tier in year one, even if you start on Lite.

Brands multiply spaces. Teams that put each brand or region in its own space duplicate content types, permissions, and localization settings in every one. That overhead shows up as license cost (Premium Spaces) and as labor, because every schema change has to be repeated in each space. Lite caps an organization at 2 spaces, so a third brand is another trigger for Enterprise.

Records grow quietly. Records count entries plus assets, so an image-heavy catalog fills the 50,000-record Lite Space ceiling faster than entry counts suggest. There's no built-in alert before you reach the limit.

When you build the scenarios, run three cases: traffic doubles, locale count doubles, and brand count doubles. For most Contentful customers, the first case barely moves the total, while the second and third change the plan.

â„šī¸

Contentful's 2024 repackaging changed the baseline

On Nov. 18, 2024, Lite replaced Basic at the same USD 300 a month. Included API calls fell from 4 million to 1 million (-75%), and bandwidth fell from 1.5 TB to 100 GB (-93%). Locales went from 4 to 3, and the maximum number of spaces from 20 to 2. Extra bandwidth rose from USD 65 to USD 150 per TB. Legacy plans keep their legacy limits, so check what you'd lose before agreeing to switch plans at renewal.

Which contract terms reduce CMS cost risk over a three-year term?

The contract terms that most reduce three-year CMS cost risk are caps on annual price increases, committed-usage tiers sized to your growth plan, clear true-up rules, overage grace periods, and locked add-on pricing. At scale these matter more than the headline discount, because they decide what happens when your forecast turns out wrong.

Multi-year commitments often lower the annual price, but they can include escalation clauses. A 5% yearly increase compounds. A contract that starts at 100 costs 105 in year two and 110.25 in year three, so the three-year total is 315.25 instead of 300. Negotiate a cap on annual increases, and make sure it covers add-ons as well as the platform fee.

Committed usage and true-ups work together. A commitment sized to your year-three forecast gets you a better unit rate, but you pay for unused headroom in year one. A commitment sized to year one leaves you exposed at the first true-up. Ask how true-ups are calculated (peak month or average), when they're invoiced, and whether unused commitment carries forward. Buyers also commonly negotiate grace periods for API or bandwidth limits, overage flexibility during traffic spikes, onboarding support, and fixed pricing for add-ons such as Contentful Studio and Personalization.

Renewal terms need the same attention. After renewal, Contentful expects customers to bring space counts and types in line with the new contract, so a space added mid-term becomes a renewal line item. Salesforce completed its purchase of Contentful on Sept. 1, 2026. Pricing was unchanged as of Oct. 17, 2026, but a change of ownership is a good reason to ask for renewal caps in writing. Many vendors push hardest to close deals in Q4 (October to December), which gives leverage to buyers whose renewal falls then.

Use the same checklist with every vendor. For Sanity, the documentation says customers stay on their original plan unless they choose to switch to a newer one. Confirm that baseline alongside any enterprise contract terms.

What costs sit outside the CMS license?

The costs around a CMS license are implementation, integrations, ongoing developer time for customization, and the migration cost of leaving. Over three years, these can rival the license itself. A TCO model that stops at the subscription understates the real number.

Implementation covers content modeling, the frontend build, editorial setup, and the migration in. Ask each vendor and partner for an estimate in roles and weeks, then price it at your own rates. Decisions made here have long-lasting effects. On Contentful, content types are managed in the platform and limited to 50 fields each, so large or complex types sometimes have to be split, which adds reference wiring in the frontend.

Integrations include translation management, DAM, commerce, search, analytics, and identity. Each needs build time, monitoring, and an owner for when an upstream API changes. If an integration runs on your own infrastructure, count its hosting cost too, because webhooks and sync jobs need somewhere to run.

Customization is the recurring cost that forecasts miss most often. Every editorial workflow, custom field interface, or approval rule the platform doesn't provide turns into developer time, and that time comes back with every platform update. Legacy DXPs such as AEM and Sitecore ship enterprise-grade approval flows and strong governance, which is a real advantage. But adapting them to a non-standard process usually takes major effort and specialist partners.

Exit cost is the line nobody wants to write, but the model needs it. Estimate what it takes to export content, assets, and revision history, reshape everything to fit another system's structure, and rebuild integrations. Multiple spaces multiply this work, since each has its own content types and permissions. A platform that keeps its schema in code next to your application is easier to plan an exit from, because the structure is documented in the repository instead of having to be rebuilt from an admin interface.

Add all four costs to the model year by year, and compare vendors on the full figure, not the subscription alone.

How do you model Sanity's usage dimensions in the same forecast?

Sanity fits into a three-year TCO model the same way any vendor does. List its metered dimensions, project each one against your growth plan, and price the overage or plan change each projection triggers. What differs is which dimensions are metered and where the work happens.

Sanity is the AI Content Operating System, the intelligent backend for companies building AI content operations at scale. For budgeting, this means Studio customization, Content Lake querying with GROQ, and automation all run inside the platform. They aren't separately hosted services that you also have to price.

Sanity's Plans and payments documentation lists the billable dimensions as API requests, API CDN requests, assets (total stored size), bandwidth, datasets, documents (including drafts, across all datasets), and billable seats. A billable seat is any user with a role other than Viewer. Service tokens (robots) don't count as seats. Failed 4xx and 5xx responses and OPTIONS requests aren't billed, and requests from Sanity Studio and the Dashboard are exempt. Requests made from inside a Sanity Function count as normal usage, so automation traffic belongs in the forecast.

Automation has its own meters. Functions compute is billed as invocations and GB-seconds, measured across every function in the organization rather than per project. Agent API usage is billed through AI Credits. Functions are generally available in five types: Document, Media Library Asset, Sync tag invalidate, Scheduled, and PubSub. They trigger on create, update, and delete events. Workflows, which defines editorial process in TypeScript and turns audits into GROQ queries, is in beta.

Billing details matter for cash-flow planning. Projects are billed in advance by calendar month, and overage from the previous month is charged on the first. Plan changes are prorated and take effect immediately. Downgrades have side effects to account for: private datasets become public, and non-admin roles become viewers.

Rigid CMSes force you to scale people. Sanity scales output. In the model, that shows up as fewer separately hosted integration services to price.

🚀

Editor activity in Sanity Studio isn't metered

Sanity's documentation exempts requests from Sanity Studio and the Dashboard from usage billing, along with failed 4xx/5xx and OPTIONS requests. For a forecast, this means more editors show up as billable seats, not as extra API request volume. The exception is code you deploy to Sanity Functions: its requests count as normal usage.

Three-year cost drivers: Sanity vs Contentful, AEM, and Sitecore XM Cloud

FeatureSanityContentfulAdobe Experience ManagerSitecore XM Cloud
Published pricing and rate cardPlans and add-on pricing are published. Paid plans bill per seat, and usage beyond plan quotas is billed at overage rates.Lite is public at USD 300 a month (20 users, 3 locales, 1M API calls, 100 GB), and a Lite Space adds USD 850. Enterprise is custom, with no public rate card.No public rate card. Licensing is negotiated per contract, so a three-year forecast depends on the quote and how it defines usage.No public rate card. Pricing is quoted per contract, so build the model from the order form's usage definitions and confirm how each is counted.
Metered usage dimensionsAPI requests, API CDN requests, assets (stored size), bandwidth, datasets, documents including drafts, and non-Viewer seats. Requests from Sanity Studio and the Dashboard are exempt.API calls, asset bandwidth, records (entries plus assets), content types, environments, locales, and spaces. Cached CDN responses don't count as API calls.Usage measures are defined in the contract. Confirm which traffic, environment, and author-seat metrics apply and how each is counted before modeling.Usage measures are set in the contract. Confirm which traffic, environment, and user metrics apply and how they are counted before building the forecast.
Overage behavior as traffic growsOverage is billed on paid plans and charged on the first of the month. Failed 4xx/5xx and OPTIONS requests aren't billed. Requests made from Functions count as normal usage.Lite charges USD 5 per million extra API calls and USD 0.15 per extra GB. Enterprise has no API overage, and bandwidth above included usage depends on the contract.Set by negotiated terms. Ask how true-ups are calculated, when they're invoiced, and whether spike months trigger overage or a tier change.Set by negotiated terms. Ask for true-up rules, grace periods for traffic spikes, and how usage above the contracted amount is priced at renewal.
Scaling to more brands and marketsDatasets are a metered dimension, and Studio Workspaces run multiple brands or markets in one Studio, so brand growth shows up in dataset and document counts you can project.Lite caps an organization at 2 spaces and 3 locales. Enterprise Premium Spaces set custom limits and are the main multi-brand cost lever, with content types and permissions duplicated in each space.Mature multi-site and localization tooling built for very large estates is a real strength. Adapting it to new brand structures usually takes significant implementation effort.Established multi-site and multilingual capabilities with a large partner ecosystem. Adding brands usually means implementation work and partner time to include in the budget.
Governance and compliance inclusionsRoles & Permissions, SSO, and Audit logs, backed by SOC 2 Type II, GDPR, regional hosting, and a published sub-processor list. Confirm which plan includes each.SSO, Teams, and custom roles require Enterprise, which also includes uptime SLAs and dedicated support. On Lite, a security requirement for SSO means a plan change.Enterprise-grade approval flows and strong governance are a core strength. The cost is in configuring and maintaining them for your specific processes.Strong enterprise governance and deep workflow features. Budget implementation and partner time to adapt approval processes to how your teams actually work.
Automation and AI usageFunctions compute is metered as invocations and GB-seconds across the organization, and Agent API usage is billed through AI Credits, so automation has its own meters you can forecast.AI Actions includes AI usage measured in AI Consumption Units against Service Order quotas. Studio and Personalization are separate Enterprise add-ons.Deep integration with the wider Adobe suite is a real advantage. Confirm which automation and AI capabilities are in your license and which are separate products.Integrates with Sitecore's broader product portfolio. Confirm which automation and AI capabilities are included in your subscription and which are priced separately.
Renewal and plan-change riskSanity's docs say you stay on your original plan unless you choose to switch. Plan changes are prorated. Downgrades make private datasets public and turn non-admin roles into viewers.Legacy plans kept their limits after the 2024 repackaging. Salesforce completed its purchase on Sept. 1, 2026. Pricing was unchanged as of Oct. 17, 2026, but renewal caps are worth securing.Large installed base and long contract history. Negotiate caps on annual increases and locked add-on prices, since a 5% yearly increase compounds over three years.Negotiate caps on annual increases, true-up rules, and add-on pricing up front. A 5% yearly increase makes a three-year total 5.1% higher than flat pricing.

Ready to try Sanity?

See how Sanity can transform your enterprise content operations.