Higher-ed web procurement rewards the vendor with the best RFP response, not always the best team. This comparison lays out the four categories of providers universities choose between, what each realistically costs, where each one breaks, and the scoring model that survives a committee review.
What are the options for higher-ed web development?
| Provider type | Typical engagement | Best for | Common failure mode |
|---|---|---|---|
| In-house web team | Salaries plus tooling, $180k to $600k/yr | Continuous improvement, institutional knowledge | Capacity collapses during a full rebuild |
| Boutique senior agency | $60k to $300k per project | Rebuilds, design systems, accessibility remediation | Limited bench for a 12-campus system |
| Large digital agency | $300k to $1.2M per project | Multi-campus platforms, formal governance | Senior staff pitch, junior staff build |
| Higher-ed SaaS platform | $40k to $200k/yr license | Small colleges without technical staff | Template ceiling, weak data portability |
Most institutions end up blending: an agency for the rebuild and design system, in-house staff for ongoing publishing and iteration.
How do we compare providers fairly?
Use a weighted scorecard and fill in the weights before proposals arrive.
| Criterion | Weight | Evidence to require |
|---|---|---|
| Higher-ed track record | 20% | Two live .edu builds of comparable page count |
| Accessibility practice | 20% | VPAT, manual testing process, remediation SLA |
| Integration experience | 15% | Reference for your SIS and CRM pairing |
| Performance results | 15% | Core Web Vitals field data from their live builds |
| Team composition | 10% | Named seniors with time allocation per week |
| Governance and training | 10% | Editor enablement plan, template limits |
| Five-year cost | 10% | Build plus hosting plus expected change requests |
Ask for Core Web Vitals field data, not lab scores. Anyone can pass Lighthouse on a laptop.
What do higher-ed web services cost?
| Service | Small college | Mid-size university | Multi-campus system |
|---|---|---|---|
| Full rebuild | $60k to $140k | $150k to $420k | $420k to $1.1M |
| Design system only | $25k to $50k | $50k to $110k | $110k to $240k |
| Accessibility remediation | $12k to $35k | $35k to $90k | $90k to $220k |
| CMS migration | $18k to $45k | $45k to $160k | $160k to $400k |
| Ongoing care and support | $1.5k to $4k/mo | $4k to $12k/mo | $12k to $30k/mo |
Budget 15% to 20% of the build as an annual figure for care, iteration, and accessibility monitoring. Institutions that skip this line spend it anyway, just in emergency change orders.
Which services matter most for enrollment?
Ranked by measured impact on inquiry and application volume:
- Program page speed and clarity, since these are the pages prospects actually land on
- A short, mobile-first RFI form wired into the CRM
- Search that returns useful results for program names and abbreviations
- Accessible, scannable program content with outcomes and cost stated plainly
- Structured data so programs surface in AI answers and search features
Everything else, including the homepage redesign everyone argues about, moves fewer numbers than these five.
Should we hire one vendor or split the work?
Splitting design, build, and accessibility across vendors sounds like risk management and usually creates finger-pointing. A defensible split is a single build partner plus an independent accessibility auditor. That keeps accountability clean while giving the committee an objective third-party check.
How do we avoid the bait-and-switch staffing problem?
Put it in the contract. Name the individuals, specify their weekly hours, and require written approval for substitutions. Ask for the resumes of the people who will be on your project in month three, not the pitch team. Boutique senior shops rarely have this problem because the people pitching are the people building.
What should the RFP require?
- WCAG 2.2 AA with a named testing method and remediation window
- Core Web Vitals targets as an acceptance criterion
- Content migration scoped by page count with a redirect map deliverable
- Editor training hours, recorded, plus a written governance model
- Full ownership of code, design files, hosting, and analytics
- A 90-day post-launch warranty
- Optional annual care priced separately so it can be compared
How long should the engagement take?
Twelve to twenty weeks for a small college, six to nine months for a mid-size university, and a year or more for multi-campus systems. The slowest phase is always content ownership decisions, so start the inventory in week one, not after the design is approved.
What does DoodleWeb bring to higher ed?
Senior-only staffing, Drupal and WordPress depth, accessibility written into the contract, and performance budgets enforced in the build pipeline. We also publish our approach openly, from what universities need in website development to the higher education CMS guide. See our work and pricing.
Where to go next
Fill in the scorecard weights with your committee before you read a single proposal. It is the single change that most improves higher-ed vendor selection.
Related reading: why university websites get slow and hard to manage and WordPress to Drupal higher-ed case study. Want an outside review of your shortlist? Contact us.
Web Strategy Lead · Seattle, WA
Nick leads discovery and platform selection at DoodleWeb. He has run CMS evaluations and rebuild roadmaps for universities, health systems, and mission-driven organizations across the Puget Sound.
