Careers Website Personalisation That Turns Visitors into Applicants

What careers website personalisation is

Personalisation means a visitor sees what is most relevant to them, without having to dig for it.

Candidates arrive with different goals. A graduate wants to know about early careers. An experienced engineer wants to see technical problems and flexible working. A nurse in Leeds wants roles in Leeds. A single homepage can serve all of them only by being generic. Personalisation lets the page do the work of a good recruiter: ask, listen and show the right thing.

Illustrative structure. The last figure reflects the approved proof point that customers manage content themselves.

How it works: signals, rules and content

Every personalisation set-up has three parts. Keep them simple and each rule will be easy to explain, test and switch off.

1. 1. SignalsWhat you know: location, referrer, device, pages viewed, language.
2. 2. RulesIf this signal, then show that content. Written by your team.
3. 3. ContentThe variants: a hero, stories, roles or calls to action.
4. 4. MeasureCompare with the default version and keep what works.

Personalisation depends on content you can change quickly. If every variant needs a developer, rules die before they start. See how a careers website CMS makes that practical.

Three visitors, three homepages

Here is what one careers website might show three different visitors. The page stays on brand, and only the parts that matter move.

Privacy and consent

Personalisation should earn trust. Use the least data you need, be clear about it and let people say no. Rules based on page context, such as the page a visitor is on or the campaign link they used, need far less personal data than rules based on tracking across visits.

• Agree with legal what data you will use and where it is stored.
• Make consent choices clear and honour them.
• Prefer context signals to behavioural tracking.
• Keep a default version that works for everyone.
• Document every rule and review them quarterly.

Rules vary by country. Take advice from your own legal and data protection teams.

What it needs from your team

Start with employee stories. They are the easiest content to vary by audience, and our storytelling course can help you collect them.

Results and what to measure

Test one rule at a time against a default version. Judge the result on applications, not clicks.

Our ROI calculator turns a small uplift in application rate into an estimate of hires and savings. Always show a cautious case first.

Eight personalisation ideas to try first

You do not need advanced technology to start. These ideas use signals you already have and content you can write in a day.

A 30-day plan to launch your first rule

1. Week 1: ChoosePick one audience and one signal. Agree consent rules with legal. Record today's baseline.
2. Week 2: CreateWrite the content variant and check it on a phone. Keep the tone and images on brand.
3. Week 3: TestRun the rule against the default for at least two full weeks, ideally four.
4. Week 4: DecideKeep it, change it or switch it off. Write down what you learned and pick the next rule.

Treat every rule as an experiment, with an owner and an end date. Rules that nobody reviews slowly turn into clutter.

Common mistakes

• Too many rules at once, so nobody knows what worked.
• Personalising the message but not the jobs.
• Using personal data when context would do.
• Forgetting to keep a good default version.
• Judging on clicks instead of applications.
• Leaving rules running for years without review.

Personalisation works best when the basics are strong. If your careers website is slow or hard to use, fix that first. Our careers website performance leaderboard shows how large employers compare on speed and accessibility.

Availability by tier

Related: conversational navigation

Personalisation chooses what to show. Conversational navigation lets a visitor ask. They work well together. Read our chatbot guide and the wider Happydance platform overview. For design thinking, see careers website design best practices.

A measurement plan for your first rule

Personalisation is easy to switch on and hard to prove. Decide how you will judge a rule before it goes live, or you will argue about the result afterwards.

Small sites may not get enough traffic to prove a small change. In that case, test bolder changes, run them longer, or treat the result as a signal rather than proof. Be honest about uncertainty in what you report.

A worked example: one rule, start to finish

A retail employer wants more applications for store roles from people near a store. All figures are invented to show the method.

1. SignalVisitor location, with consent where required.
2. RuleShow the three nearest store roles above the fold.
3. ContentA short local headline and a map link, written once by the team.
4. MeasureJob search starts per visit, with bounce rate as the guardrail.
5. DecisionKeep it if starts rise and bounce does not, then try the next rule.

Notice what is not in the example: a long list of rules. One clear rule, with one owner, beats ten rules nobody maintains. See the conversational navigation guide for how a guided route can complement it.