Choose recruiting software by starting with your own process, not with the vendor demo. Write down the 10 tasks your team does every week, set a budget range and a data protection requirement, then ask every vendor the same 25 questions and score the answers on one sheet. This guide gives the questions, the scorecard and a short pilot plan.
Key takeaways
- Define your process and must-have needs before you see a demo.
- Ask every vendor the same 25 questions and record the answers in writing.
- Check data protection, security and automated-decision features early. They can rule a product out.
- Ask for a pilot with your own roles, and set success measures before it starts.
- Be careful with any tool that ranks, scores or rejects people. Ask how a person can see and challenge each result.
Step 1: Write down what you need
Vendors sell features. You need outcomes. Before you speak to anyone, spend an hour with your team and write a short brief. It should include the following.
- The 10 weekly tasks. For example: post a job to several boards, receive applications, review CVs, message candidates, schedule interviews, collect interview feedback, make an offer, and report numbers.
- The three biggest pains. For example: applications arrive by email, interview feedback is lost, or candidates are not told the outcome.
- Volume. How many roles and how many applications per year, as ranges. Use your own records. If you do not have them, see our guide to a hiring dashboard in a spreadsheet.
- Countries. Where do you hire and where will candidate data be stored? Rules differ for the US, UK and EU.
- Budget range and contract length you can accept. Do not invent a figure to impress anyone. Use your own finance team’s number.
- Who will use it. Recruiters, hiring managers who rarely log in, and an administrator. Hiring managers should need no training.
Step 2: Separate must-haves from nice-to-haves
Use the same discipline you use for a job. A must-have is a need where “no” means you cannot use the product. A nice-to-have is a feature that helps but has a workaround. Keep the must-have list to about eight lines.
| Area | Example must-have | Example nice-to-have |
|---|---|---|
| Applications | Receives applications from our career page and by email | Mobile app for candidates |
| Collaboration | Hiring managers can leave structured feedback without training | In-app chat |
| Communication | Templates for acknowledgement and decline messages | Text message sending |
| Data | Export all our data in a standard format at any time | Custom report builder |
| Compliance | Retention rules and deletion per candidate | Built-in consent pages in several languages |
| Control | No automated rejection without a person’s decision | Duplicate detection |
Step 3: Ask the 25 questions
Send the questions in writing before the demo, and ask for written answers. Spoken answers are easy to forget and hard to compare. Keep a copy of the vendor’s reply for your records.
Process and daily use (questions 1 to 6)
- Which of our 10 weekly tasks can your product do without workarounds? Show each one in the demo, using our example role.
- How does a hiring manager give feedback? How many clicks, and can they do it from email?
- Can we create our own stages, scorecards and message templates, or do you set them?
- How do candidates apply? Is there a mobile-friendly form, and how long does it take a candidate to finish?
- How do we reach applicants who are not a good match? Can we send a decline in bulk, and can we personalise it?
- What reports can we run without paying extra? Can we export raw data to a spreadsheet?
Data protection and security (questions 7 to 14)
- Where is candidate data stored, in which country, and which companies process it as sub-processors?
- Do you act as a data processor for us? Will you sign a data processing agreement before we start?
- How do you support lawful transfers of data between regions, if we hire in the EU or UK and you store data elsewhere?
- How can we delete a single candidate’s data on request, and how fast does it disappear from backups?
- Can we set automatic retention periods per country or per role?
- What security certifications or independent audits do you hold, and can you share the reports under an agreement?
- Who at your company can see our candidate data, and how is access logged?
- What happens to our data if we end the contract? How long do you keep it, and in what format can we take it?
Fairness, automation and AI (questions 15 to 20)
- Does the product rank, score, filter out or recommend candidates in any way? Describe each feature in plain language.
- Can every automated feature be switched off? Is any enabled by default?
- For every score, ranking or recommendation, can a person see the reasons, and can they override it?
- What data was used to build or tune the feature? Do you test for different results between groups, and can we see the test summary?
- How do you help customers meet the rules that apply to automated decisions and AI in hiring, such as local laws in some US cities, the UK and EU data protection rules, and the EU AI Act?
- Do candidates see any notice that automated tools are used on their application, and can we edit it?
Cost, support and contract (questions 21 to 25)
- What is the total cost over three years for our volume, including set-up, training, extra users, extra roles and support? Please give it in writing.
- What is the contract length, the notice period and the price increase rule at renewal?
- What support is included, in which languages and time zones, and what is the response time for urgent problems?
- What does the first 90 days look like? Who helps us import data and set up templates?
- Can you give us two customers of our size and type to speak to, who were not selected by your sales team alone?
Questions 15 to 20 need special attention. In the United States, some cities and states regulate automated tools in hiring. For example, the New York City Department of Consumer and Worker Protection explains the city’s rules on automated employment decision tools. In the EU, the AI Act sets rules for AI systems, and hiring uses are a regulated area. We do not state application dates here. Read the official text for current dates, and see our guide on the EU AI Act and hiring tools. In the UK, the Information Commissioner’s Office explains data protection duties in its recruitment and selection guidance.
Step 4: Score the answers
Create one table. List vendors across the top and your criteria down the side. Give each criterion a weight from 1 to 3 based on how important it is to you. Score each answer from 0 to 3 as follows: 0 means it cannot be done, 1 means it needs a workaround, 2 means it works, and 3 means it works well and is shown in the demo.
| Criterion | Weight | Vendor A (0 to 3) | Vendor B (0 to 3) | Vendor C (0 to 3) |
|---|---|---|---|---|
| Fits our 10 weekly tasks | 3 | |||
| Hiring manager ease of use | 3 | |||
| Data protection and deletion | 3 | |||
| Automation is optional and explainable | 3 | |||
| Reports and export | 2 | |||
| Total cost over three years | 2 | |||
| Support and onboarding | 2 | |||
| Contract terms and exit | 2 | |||
| Weighted total |
Use the table to reduce the choice to two vendors, not to make the final choice. A score cannot capture how a product feels in daily use. Use the pilot for that. And note the irony: you are scoring software, not people. When you screen candidates, the same discipline applies, but a person should make each decision.
Step 5: Run a short pilot
Ask for a trial or a paid pilot of two to four weeks with one or two real roles. Do not test only with demo data. Agree the success measures in writing before you start.
| Measure | How to check |
|---|---|
| Setup time | Hours from start to the first role live |
| Hiring manager feedback | Share of requests answered without a reminder |
| Candidate completion | Share who start the application and finish it, compared with your current form |
| Data accuracy | Spot-check 20 records against the source |
| Support quality | Number of questions asked and time to a useful answer |
| Data export | Export everything and open it. Check that no fields are missing |
At the end of the pilot, ask your hiring managers and two or three candidates for comments. Then decide. If the vendor will not allow a pilot, that is information about the contract you will get.
Red flags
- No written answers. If a vendor will not put data protection answers in writing, do not buy.
- “Our AI picks the best candidates.” Ask how, with what data, and what a person can see and change. Be careful with any promise of removing bias with software.
- No data export. You must be able to leave.
- Pressure to sign this week. A good tool will still be there next month.
- Pricing that depends on a call. This may be normal, but ask for a three-year total in writing.
- Reference customers chosen only by the vendor. Ask for one more you choose from the vendor’s public customer list.
FastHire is coming soon and is being built for the screening step. It shows each job requirement next to the exact CV passage that supports it. It does not rank, score or reject candidates. People make every decision. If you compare tools, ask FastHire and every other vendor the same 25 questions.
Contract points to check before you sign
| Point | What to ask for |
|---|---|
| Data ownership | A clause that says candidate data belongs to you, and the vendor may use it only to provide the service |
| Data processing agreement | A signed agreement that lists sub-processors and gives you notice of changes |
| Exit | Export of all data in a standard format within a fixed number of days, then deletion confirmed in writing |
| Price changes | A cap on increases at renewal, or a fixed price for the term |
| Service problems | A written commitment on availability and a remedy if it is not met |
| Changes to features | Notice before the vendor adds or changes any automated feature that affects candidates |
Ask your legal or data protection adviser to read the contract. If you hire in the EU or UK, also ask whether a data protection impact assessment is needed before you start. Our guide on impact assessments for hiring tools will cover this in more detail.
A 60-day rollout plan
- Days 1 to 10: set up. Import open roles, build stages and message templates, and connect the career page. Name one owner.
- Days 11 to 20: train a small group. Train recruiters first, then two friendly hiring managers. Collect questions and fix templates.
- Days 21 to 40: run live roles. Use the tool for new roles only. Do not move every old role at once.
- Days 41 to 50: review. Compare your pilot measures with the targets. Fix problems with the vendor in writing.
- Days 51 to 60: expand. Move the remaining roles, switch off old inboxes and tell candidates where to apply.
Plan one more step: set a review date six months after launch. Ask whether the tool still solves the three pains you wrote at the start. If not, say so while you still have contract options.
Frequently asked questions
How long should the selection process take?
For a small team, three to six weeks is often enough: one week for the brief, one to two weeks for questions and demos, and two to four weeks for a pilot. Adjust to your own approval process.
Do I need an applicant tracking system at all?
Not always. A small team that hires a few people a year can use a spreadsheet and email. Consider software when volume, collaboration or compliance records become hard to manage.
Who should be in the decision?
At least one recruiter, one hiring manager who will use it, someone responsible for data protection or IT security, and whoever owns the budget.
What if the vendor says its AI is compliant?
Ask for the supporting documents and read them. Compliance is a shared responsibility, and you as the employer remain responsible for your decisions.
Should I choose an all-in-one suite or a specialist tool?
It depends on the pains you wrote down. A suite reduces the number of systems. A specialist tool may fit one task better. Score both with the same questions.
Related guides
- Hiring guides for recruiters and HR teams
- How to build a hiring dashboard in a spreadsheet
- Screening at volume: a process for 500 applications
- EU AI Act and hiring tools
- All hiring guides for employers
- Applicant tracking system compliance checklist: GDPR and US records
- Data protection impact assessments for hiring tools
- How to work with agencies: a briefing and fee checklist
This is general information, not legal advice.