Two EDC systems with the same features can feel completely different to run, because in one the vendor builds and changes your study and in the other your team does. This page sets out who each model suits, what each really costs in time and change orders, and how to decide.
Free sandbox · No credit card · 21 CFR Part 11 aligned
| Full-service EDC | Self-serve EDC | |
|---|---|---|
| Who builds the eCRFs | Vendor team | Your team |
| Mid-study change | Change order | Edit and approve yourself |
| First build lead time | Weeks, typically | Hours to days |
| Managed data management included | often | |
| Needs in-house data skills | ||
| Direct control of timelines |
The short answer
Definitions
The terms are used loosely, so pin them down before you compare quotes. In a full-service model the vendor receives the protocol, specifies and builds the database, tests it, hands it over after your user acceptance, and then handles changes through a request process. Many vendors also sell data management as a service: query handling, data review listings, coding, reconciliation and database lock. In a self-serve model you receive an account and tools. You (or a colleague, or a CRO you contract separately) build the forms, set the checks, add the visits and edit them as the protocol evolves.
There is a middle: assisted self-serve, where the platform is yours to operate but the vendor will do the first build or answer build questions. When a vendor says "we do the setup for you", ask whether that is a one-off, whether it is priced, and what happens when you need a change in month nine.
This is a separate question from whether to use a CRO. A CRO can run a trial on a self-serve system, and a sponsor can run a trial alone on a full-service one. For the CRO question, see clinical trial software without a CRO; for what self-serve platforms look like in practice, see self-serve EDC platforms and self-serve clinical trial software.
Side by side
Descriptions of how each model typically behaves. Individual vendors differ, so test each line against the quote in front of you.
| Dimension | Full-service EDC | Self-serve EDC |
|---|---|---|
| Up-front cost | Build fee or paid service hours on top of licence; quoted from the protocol. | Licence only, with the build cost being your team's time. |
| Time to first build | Specification, build, test and UAT cycles; weeks is common. | A first draft can exist within a day when forms come from a protocol or template. On Capture the pricing page says a study can be live in under 24 hours. |
| Protocol amendment | Change request, estimate, approval, build, test, release. Each step has its own lead time. | Edit the form or visit, test in a draft, approve and release. Turnaround is your own review time. |
| Cost of change | Usually billed per change order or as service hours; the unknown line in the budget. | No vendor charge for routine edits; your reviewer's time is the cost. |
| Data management labour | Can be bought in the same contract. | You, an in-house data manager or a separate provider. |
| Skills needed in-house | A reviewer who can specify and accept the build. | Someone who can design forms and checks, and own the data management plan. |
| Control of timeline | Shared with the vendor's queue. | Entirely yours. |
| Accountability for build errors | Contractual; the vendor fixes defects in their build. | Yours, which makes testing in the sandbox more important. |
| Scale-up | Capacity grows by buying more services. | Capacity grows with your team; the platform scales on its own. |
Market description, not a statement about any named vendor. Cost figures differ by vendor and contract.
The money line
A full-service quote is a forecast of the build, not of the study. The build is the part that is easy to specify. What is not is everything the protocol will do after the database is live: amendments that add visits, a safety request for one extra lab, a field someone forgot to make required, a new site that needs a different consent version. Every one is a change to a validated build, and in a vendor-built model each normally goes through a request, an estimate and a release.
Before you sign, ask for the change-order rate card in writing, a turnaround time for a standard change, and whether a change after first participant in is priced differently. Then model your own protocol: the number of amendments the study will likely see is a good predictor. A short, stable protocol amends rarely; an adaptive or first-in-human study amends often. The same logic is in protocol amendments without change orders.
Self-serve moves that cost from the vendor to your calendar. An amendment still needs impact assessment and testing; what disappears is the queue, the quote and the fee.
Protocol amendment approved
Adds a Week 6 visit and one lab
Self-serve: edit draft form and visit
Same day, in the study team
Self-serve: test and approve
Sandbox run, then release
Full-service: request, estimate, approval
Vendor queue, then billed
Full-service: build, test, release
After the vendor's lead time
Upload a protocol to the free sandbox and see how much of the build the AI drafts. Every feature, no credit card, no time limit.
Decision rule
Answer in order. The first "no" that matters usually decides it.
If yes, self-serve is realistic. If the only candidates are clinicians with no database experience, budget for either training or a vendor build.
Edit checks, query handling and database lock need an owner. A data manager can be an employee, a consultant or a services contract. Self-serve software does not remove the role.
Many changes expected means a change-order model is risky. Few changes expected shrinks the difference.
A first-patient date weeks away favours self-serve. A date months away makes the vendor's lead time acceptable.
Some sponsors prefer a vendor-run build with contractual accountability. Others accept in-house builds with documented testing. Ask before you commit.
Fixed-price full-service quotes can still move through change orders. Per-participant software pricing moves with enrolment. Decide which variance you can absorb.
Who fits where
Invented profiles for reasoning, not recommendations.
| Team | Situation | Likely fit |
|---|---|---|
| University department | Investigator-led Phase 2, a research coordinator who has built REDCap projects, no budget for services | Self-serve. Add a biostatistician review of the build and edit checks. |
| Small biotech | First-in-human, 40 participants, protocol likely to amend after dose escalation, one clinical operations lead | Self-serve for control of amendments, with a contracted data manager if no one on staff has done a database lock. |
| Mid-size sponsor | Phase 3, 600 participants, 40 sites, in-house biometrics department | Either. Self-serve works if biometrics can specify and test; many sponsors in this position still choose a vendor build to share accountability. |
| Device startup | Post-market study, simple forms, no data staff | Self-serve with templates, or a small services contract for the first build. |
| Sponsor with no data capability | Rare disease trial, outsourced to a CRO that supplies its own EDC | Full-service through the CRO, with the data export terms written into the contract. |
Where Capture sits
Capture is a self-serve clinical trial platform. Your team builds the study in the same tool you will run it in, tests everything free in the sandbox, and goes live when ready. The pricing page lists study setup by the Capture team at $0 on Core, and no per-site fee or setup fee line.
Stage 1
DraftBuild and edit freely
Stage 2
ApprovedLocked for live use
Stage 3
In the casebookOnly approved forms
SDV requirement changed on Systolic BP
Data manager · Reason: critical safety value
Form approved: Vital Signs (VS) v2
Study owner · Locked for live use
Answer changed: Weight 68.0 to 68.5 kg
Site coordinator · Reason: transcription error
Quote questions
Written scope for the first build, including number of forms, visits and edit checks.
Hourly or per-change fees, minimum charges and turnaround, in writing.
Is a change after first participant in treated differently, and who validates it?
Query handling, listings, coding, reconciliation and lock: included, priced or excluded?
What your team gets to run the system, and the response time for build questions.
Full data export with audit trail and codebook if you leave, and what it costs.
Mixing the two
You can split the labour without splitting the platform. A common pattern is a self-serve system with an external data manager or statistician hired for the first build and the lock, and used by the study team for everything in between. Another is the reverse: an in-house team that builds routine studies alone and buys help for a complex one. Because Capture has a free sandbox with every feature, a consultant can build and test with you before you pay anything.
Whichever pattern you choose, write it into the data management plan: who builds, who reviews, who approves a release, and who may edit a live form. For cost context, the EDC software pricing guide shows how market quotes are typically structured, and the 500-patient cost worksheet shows Capture's published rates at scale. The wider vendor field is in the comparison hub.
In full-service EDC the vendor builds your database, often handles changes through change orders, and may supply data management. In self-serve EDC your team builds and edits forms, visits and checks in the platform itself.
Usually the licence-plus-services total is lower because there is no build fee and no change-order charge, but your team's time is the cost. If nobody can do the work, hiring a data manager or buying services can erase the saving.
Yes, if someone owns the data management plan, edit-check design and database lock. See clinical trial software without a CRO for the checklist.
The Capture site describes a self-serve platform and does not list managed data management as a service. If you need it, contract a data manager or CRO, or ask on a demo what the Capture team can help with.
The pricing page lists study setup by the Capture team at $0 on Core, with Core at $1,000 per month including one live study plus $50 per participant per month. The study fee starts when the first participant is screened.
There are none from the vendor for routine edits. Your team edits a draft form, tests it and approves it. The amendment still needs your own impact assessment and documentation.
Capture is suitable for Phase 1 to 3. Whether your team should self-build a Phase 3 database depends on the skills in biometrics and data management, not on the phase.
It is a migration project, not a toggle. See how to switch EDC vendors mid-study before deciding.
Keep exploring
Self-serve EDC platforms
What self-serve platforms offer.
Self-serve clinical trial software
Running a study without a vendor build.
Clinical trial software without a CRO
The CRO side of the decision.
Protocol amendments without change orders
How edits work when you own the build.
Hidden EDC fees
The lines that grow after signing.
Pricing
Core, Full Platform and Enterprise.
Compare EDC vendors
Head-to-head pages.
Upload a protocol to the free sandbox and judge the first draft. Every feature, no credit card, and you pay only once you go live.