Posts
Posts
Build or Buy a Careers Website? A Decision Guide for Enterprise Teams
See other posts from by Jim Taylor
Managing Director
Decision guide
Build or buy a careers website? Compare in-house, agency and platform routes on cost, speed, risk and ownership.
By the Happydance team. Updated November 2026.
Three routes: in-house, agency or platform
Every enterprise team asks the same question: do we build our careers website or buy one? The honest answer is that it depends on three things: who will change it, how fast you need it and how much risk you can carry.
Build
In-house
Your developers build and run it. Full control, full responsibility.
Commission
Agency
A studio designs and builds it, often with a retainer for changes.
Buy
Platform
A careers website platform your own team manages, configured to your brand.
Cost over three years
Compare three-year cost, not the launch price. The build is one line. Developer time, hosting, security, accessibility, languages, markets and every content change are the others.
| Cost line | In-house | Agency | Platform |
|---|---|---|---|
| Build | Developer and design time | Build fee | Set-up fee |
| Run | Developers, hosting, tools | Retainer and hosting | Subscription |
| Change | Developer queue | Per-change or retainer hours | Your own team, in the editor |
| Languages and markets | More build and testing | More fees per language | Managed alongside everything else |
| Compliance | Your own security and accessibility work | Often extra | Part of the platform, check the terms |
Use our careers website cost guide and its editable three-year model with your own quotes.
Speed to launch
Building from nothing takes longest. An agency adds design and approval cycles. A platform starts from a working foundation, so most of your effort goes into content, design decisions and migration. Read how to redesign a careers website for the timeline.
Risk and ownership
| Risk | In-house | Agency | Platform |
|---|---|---|---|
| Key person leaves | High: knowledge sits with a few developers | Medium: the agency holds the code | Lower: configuration lives in the product |
| Content goes stale | High if changes need developers | High if changes need the agency | Lower: your team edits directly |
| Lock-in | Low on code, high on effort | Medium: code and contract | Check export and exit terms |
| Security and accessibility | Entirely yours | Shared, often unclear | Check evidence and audits |
| Roadmap | You build everything | You pay for each change | You rely on the supplier |
Whichever route you choose, insist on owning your content, design and URLs, and on a written exit plan. Our vendor choice guide explains how.
Performance and SEO outcomes
The route does not guarantee the result. A well-run custom build can be fast and accessible; a neglected one will not be. What the data shows is how a sample of large careers websites performs today.
723.0average leaderboard score of the 61 Happydance customer sites
618.6average score of the 40 other sites (38 Fortune 100 companies and 2 wild cards)
41sites at the top of the leaderboard are all Happydance customers
Sample disclosure: 101 careers websites, of which 61 are Happydance customers, 38 are Fortune 100 companies and 2 are wild cards. Scores are Lighthouse results for mobile and desktop, out of 800. It compares customers with large-company careers websites, not the whole market. See the careers website performance leaderboard.
| Measure | Happydance customers | Other sites |
|---|---|---|
| Mean mobile Lighthouse performance | 72.5 | 41.9 |
| Mean desktop Lighthouse performance | 86.8 | 58.7 |
| Mean mobile accessibility | 96.1 | 91.5 |
See the full careers website performance leaderboard. Because the sample is not the whole market, treat it as evidence, not proof.
Decision guide
Five questions point you to the route that fits. It is guidance, not a verdict.
Free tool
Which route fits you?
Answer five questions. Your answers stay in your browser.
| Question | Why it matters |
|---|---|
| Launch date | Speed |
| Who edits | Dependence |
| Languages and markets | Scale |
| Design distinctiveness | Bespoke versus configured |
| Developer capacity | Capacity |
This is the printable version. Switch on JavaScript to use the live calculator in your browser.
What a hybrid looks like
Many enterprise teams end up with a hybrid: a platform for the everyday careers website and custom work for one or two campaigns or experiences. It works when the boundary is written down: the platform is the home for content, jobs and structure, and custom work plugs in around it.
- The platform owns jobs, content and structure.
- Custom work is limited to named experiences.
- Custom work reads content from the platform, not a separate copy.
- A named owner decides what is custom and what is not.
- Both routes meet the same accessibility and speed standards.
How Happydance does it
Customers manage everything themselves, including multilingual content, favicons and basic standard graphics, with no developer needed. We believe that is what a careers website CMS should mean.
See what the Happydance platform does, and how the careers website builder puts content in your team's hands. Read the buying guide for the whole process.
A weighted decision scorecard
Opinions about build versus buy are loud and rarely scored. Score each route from 1 (poor) to 5 (strong) against criteria that matter to you, multiply by the weight and total the result. The weights below are a starting point. Change them, then ask each stakeholder to score separately before you compare.
| Criterion | Weight | In-house | Agency | Platform |
|---|---|---|---|---|
| Speed to launch | 15 | |||
| Three-year cost | 20 | |||
| Editing without developers | 20 | |||
| Design freedom | 10 | |||
| Mobile speed and accessibility | 15 | |||
| Multilingual and multi-market | 10 | |||
| Exit and ownership | 10 |
The totals matter less than the disagreement. If recruiting scores editing without developers as a 5 and IT scores it as a 2, you have found the real question and can settle it with a live test.
Hidden costs of building in-house
| Cost most teams forget | Why it appears | How to estimate it |
|---|---|---|
| Maintenance developers | Frameworks, browsers and security fixes never stop | Allow a fraction of a developer for every year of life |
| Accessibility work | Standards and testing need regular attention | Budget an annual audit and fixes |
| Language and market rollout | Each market adds content, testing and review | Count markets you expect in three years |
| Job feed and ATS change | Systems and fields change over time | List integrations and their owners |
| Analytics and consent | Rules change and tags drift | Allow a recurring review |
| Content bottlenecks | Edits wait in a developer queue | Count monthly changes and multiply by waiting time |
Five questions to ask an agency
- Who can change copy, images and languages after launch, and how long does it take?
- What will you charge for the second year, and for each new template?
- Do we own the code, the design files and the content, and in what format?
- Who maintains the site when standards change?
- What does handover look like if we move on?
Choosing a hybrid sensibly
Some teams combine routes: a platform for the pages and content, with an agency for brand and campaign design, or in-house developers for custom integrations. A hybrid works when each party has a clear boundary, such as who owns the design system, who owns content and who owns integrations. It fails when two parties can change the same thing without a shared process. Write the boundary down before you sign. Use the RFP template to put these questions to any supplier in writing.
When each route is the right one
| Route | Strong fit when | Weak fit when |
|---|---|---|
| In-house build | You have a long-term engineering team with spare capacity and a very specific requirement | You need to change content often or add markets quickly |
| Agency build | You need a bespoke campaign or brand expression with a defined life | You expect many changes after launch and want your own team to make them |
| Platform | You want your own team to manage content, languages and pages at speed | You need a one-off design that no template can express |
A worked example: twelve markets over three years
Consider an employer that launches in three markets and adds nine more over three years. The figures below are invented index points, not prices, to show where cost grows. Replace them with your own quotes.
| Cost line (illustrative) | In-house | Agency | Platform |
|---|---|---|---|
| Initial build | 100 | 80 | 35 |
| Year 1 run | 40 | 30 | 40 |
| Years 2 and 3 run | 90 | 70 | 80 |
| Adding nine markets | 135 | 110 | 30 |
| Changes and new templates | 60 | 65 | 15 |
| Accessibility and standards updates | 35 | 30 | 5 |
| Three-year total (index points) | 460 | 385 | 205 |
The point is not the totals. It is that the line that grows fastest is the one that depends on developers: markets, changes and standards. A route that lets your own team do that work changes the shape of the cost curve. Ask each supplier to price those three lines explicitly.
A risk register by route
| Risk | In-house | Agency | Platform |
|---|---|---|---|
| Key person leaves | High: knowledge sits with a few developers | Medium: depends on the account team | Lower: shared product and support |
| Slow changes | High: developer queue | Medium: change requests and quotes | Lower: team edits directly |
| Standards and security updates | You carry them | Often extra cost | Handled by the supplier |
| Lock-in | Low, but the code is hard to hand over | Medium: depends on the contract | Medium: check export and exit terms |
| Design limits | Low | Low | Medium: bounded by the system |
No route removes all risk. The aim is to choose the risks you can manage. Use the business case template to set the preferred route in front of finance.
Questions to ask your own IT team
- How many developers could we release for this, for how long, and for what else would they be needed?
- Who would keep it secure and compliant for the next three years?
- What is our record on delivering internal web projects on time?
- Which of our standards would the new site have to meet, such as single sign-on or hosting rules?
Timeline comparison
| Stage | In-house | Agency | Platform |
|---|---|---|---|
| Requirements and brief | Long: internal approvals | Medium: agency discovery | Short: guided by templates |
| Design | Depends on design team capacity | Medium to long: bespoke | Short to medium: configured |
| Build | Longest: custom development | Long: development and review | Short: content in place of code |
| Testing | Your team | Agency and your team | Shared, with supplier support |
| Launch | Depends on release schedules | Agreed with agency | Agreed with supplier |
These are relative patterns, not promises. Ask each supplier for the timeline of a comparable recent project, and ask what slowed it down.
Governance and ownership after launch
The decision does not end at launch. Whichever route you pick, name an owner for the content, an owner for the design system and an owner for integrations. Meet once a month for the first year and review requests, issues and results. The cost of a careers website is mostly the cost of keeping it good.
How to run the decision workshop
- PrepareCollect quotes, the scorecard and the key facts about your team.
- ScoreEach stakeholder scores the three routes alone before the meeting.
- DiscussSpend the time on the biggest disagreements, not the agreements.
- TestAgree a live test or reference call to settle open questions.
- DecideWrite down the decision, the reasons and the review date.
Reference calls: ten minutes that save months
Ask every supplier for two customers of a similar size and set-up, then ask the same four questions: how long did launch take against the plan, who makes changes now and how long does a change take, what did the second year cost compared with the quote, and what would you do differently? Listen for what is not said. A confident answer to the second-year question is a good sign.
A fair test for any route
Ask each supplier to show a real change being made by the customer, not by the supplier, on a live site. How long it takes and who does it is the best predictor of how the second year will feel.
Should we build or buy a careers website?
It depends on who will change it, how fast you need it and how much risk you can carry. Use the decision guide on this page for a first view.
Is a platform cheaper than an agency build?
It depends on scope, languages and how often you change content. Compare three-year cost using real quotes and the cost guide.
How long does it take to launch?
A platform starts from a working foundation, so it is usually quicker than a custom build. Your content and migration work will set the pace.
What are the risks of building in-house?
Key person dependence, a developer queue for every change, and sole responsibility for security and accessibility.
What are the risks of an agency build?
Ongoing dependence for changes, unclear responsibility for security and accessibility, and cost per change.
What are the risks of buying a platform?
Reliance on the supplier and its roadmap. Mitigate with clear export rights and an exit plan.
Can I combine routes?
Yes. A hybrid with the platform as the home for content and custom work for campaigns is common. Write the boundary down.
Where can I check performance?
The careers website performance leaderboard is public and lists scores for 101 careers websites.
Is it cheaper to build a careers website in-house?
Often it is cheaper to start and dearer to run once maintenance, accessibility, languages and change requests are counted over three years.
When does an agency build make sense?
When you need a highly bespoke design and have budget for ongoing change requests, or when a campaign site is short-lived.
How do I score the three routes fairly?
Use a weighted scorecard, have each stakeholder score separately and discuss the biggest disagreements with a live test.
What is the biggest hidden cost?
Waiting. When every change needs a developer, the cost is delayed content and missed hiring windows, not only the invoice.
How do I compare the three-year cost of each route?
Price the lines that grow, which are markets, changes and standards updates, as well as the initial build.
What is the risk of choosing a platform?
The main risks are design limits and lock-in. Check the exit terms, export formats and whether the design system fits your brand.
How long does each route take to launch?
Platforms are usually the shortest, agencies longer and in-house the longest, but ask each supplier for a comparable recent project.
Who should be in the decision workshop?
Talent acquisition, employer brand, IT, finance and a market lead if you hire in several countries.
Careers website cost guideHow to choose a careers website CMSRedesign guideCareers website builderCareers website CMSBuying guide
Weighing build against buy?
Book a demo and we will walk through the trade-offs for your team.
Book a demoContact usRelated Articles
Posts
Careers Website Redesign: The Enterprise Guide to Scope, Cost and Timeline
Read more about Careers Website Redesign: The Enterprise Guide to Scope, Cost and Timeline
Posts