A Google Forms alternative for hiring, without buying an ATS

Most screening processes start the same way. A role opens, applications arrive faster than anyone expected, and someone builds a Google Form: five questions, a link in the job ad, responses into a sheet. It takes twenty minutes and it costs nothing.

That is a good decision. It stays a good decision for longer than most vendors will admit. This page is not an argument that you should have known better. It is about the specific point where the form stops carrying the process, what your actual options are at that point, and how to tell which one you need.

What Google Forms genuinely gets right

Worth saying plainly, because most pages on this topic skip it and go straight to the sales pitch.

Candidates can answer it. There is no account to create, no software to install, no slot to book, no camera to face. It works on a phone on a bus. The completion rate for a short form beats almost anything else you could put in front of a person who is applying to several places at once.

You already have it. No procurement, no security review, no seat licences, no annual contract, no conversation with finance. For a team hiring a few people a year, that is often the whole argument, and it is a fair one.

It is structured. Everyone answers the same questions in the same order, which is already more consistent than a set of phone screens run by three different people on three different days.

So the problem is not the form. Collecting the answers was never the hard part.

Where Google Forms stops working for candidate screening

Four things go wrong, and they go wrong in a specific order.

Response one hundred and eighty is not read like response three. This is the one nobody says out loud. When the sheet has two hundred rows of free text, there is no reading of two hundred rows. There is a fast pass down the column, a few seconds each, and the rows that get real attention are the early ones, the short ones, and the ones that happen to open with a familiar company name. That is not carelessness. It is what fits in the day.

Scoring is manual, so it quietly stops happening. A form can mark a quiz answer right or wrong. It cannot tell you whether a paragraph about handling a production incident shows someone who has actually handled one. So you read, you form an impression, and the impression lives in your head or in a colour on a cell. A week later you cannot reconstruct why candidate forty was rejected, and neither can the hiring manager who asks.

Comparing candidates is the part that breaks first. Two answers to the same question, written by two people, sitting in two rows of a spreadsheet, are genuinely hard to weigh against each other. Not because the answers are unclear, but because you are holding the earlier one in memory while reading the later one. Nobody does that reliably at row sixty.

Now every answer reads well. This is the change that turned a slow problem into an urgent one. Candidates paste the job ad into an assistant and get back a fluent, structured, on-topic answer in seconds. Fluency used to be a weak signal of effort and competence. It is now free. If your screen is a form and your filter is how the writing reads, the filter no longer filters.

Here is what the difference looks like when something does read the answers. Two candidates, the same five questions for the same sales development role.

In a spreadsheet those two answers sit in adjacent rows and look about the same.

There is a fifth problem that is not about screening at all, and it tends to surface later, usually from someone in legal.

Google Forms, form builders, an ATS, or a scored written screen

Every alternatives page collapses into one of three recommendations, and they solve different problems. Picking the wrong one is how teams end up paying for software that does not touch the thing that hurt.

Read that as the honest version rather than a scorecard. A form builder is a real improvement if the form itself is your complaint, and several have usable free tiers. An ATS is the right answer to a pipeline problem, and an expensive, heavy answer to "I have too much text to read". Eneya is the third option, so treat the last column as an interested party describing their own product.

What changes when the answers are scored

Nothing changes for the candidate. They get a link, they answer in writing, they do it whenever suits them. No account, no recording, no scheduled slot. Whatever completion rate your form gets, this is the same shape of thing.

What changes is the step after submission.

You write the questions, tied to competencies. Three to five is usually enough. When you write a question you also say which competency it is meant to test, which is the part that makes scoring mean anything later. You can also give a private brief about the role, the team and what a strong answer would actually contain. Candidates never see it.

Every set of answers is read in full. Not a faster skim, and not only the promising ones. Candidate one hundred and eighty gets the same attention as candidate three, because the reading is no longer rationed by your afternoon.

The score comes with its reasoning. Each competency gets a level and a short note pointing at what in the answer earned it. The final number is computed by a fixed rule in code from those levels, not invented by the model, and the breakdown is shown in every report. This matters more than it sounds: a score you cannot inspect is a score you cannot defend to a hiring manager, or to a candidate who asks.

Each report says whether the answers name anything checkable. This is not AI detection, and it is worth being precise about the difference, because the detection promise is everywhere and it does not hold up. Nobody can reliably tell writing that a person produced from writing that a person produced with help. What can be read is whether an answer contains things that could be false: a named system, a specific decision, a trade-off with a cost attached, a number the person would have to stand behind. Generated text is fluent and unfalsifiable, because the assistant has nothing to be specific about. So the report tells you where to push in the interview, and never issues a verdict on the person.

Answers are deleted on a schedule. Candidate data is processed on EU infrastructure, removed automatically after a set retention period, and covered by a published data processing agreement. This is the one difference a spreadsheet cannot close with any amount of discipline.

What this does not do

Setting the limits out early saves everyone a demo call.

It is not an ATS and does not want to be. It covers one step. Whatever you use to track applications, you keep using.

It does not replace the interview. It changes who gets one, and gives you a first ten minutes that start from something the person already wrote.

It does not tell you who to hire. There is a score and there is reasoning behind it. The decision is yours, and a tool that pretends otherwise is selling you a liability, not a shortlist.

It does not detect AI, and any tool promising that is describing something it cannot do.

Moving a screening form over to a written screen

If you already have a form, most of the work is done. The questions are the asset.

1. Take the three questions from your form that actually change your mind about a candidate. Not the ones about notice period and location, those belong in a form and always did.
2. For each one, write down which competency it is testing. This usually takes longer than expected, and the difficulty is informative: a question you cannot map to a competency is a question you were reading for tone.
3. Write the private brief. A few sentences on the role and what a strong answer would contain in your context.
4. Send it to everyone, including the applications you would have discarded after ten seconds. That is the whole point of making the step cheap.
5. Read from the top of the scored list, and read a few from the middle too. That is how you find out whether the ordering is any good before you trust it.

Keep the form for the administrative fields if you like. The two do not conflict, and plenty of teams collect logistics in a form and run the judgement questions separately.

Questions recruiters ask

Can I keep using Google Forms as well?

Yes, and several teams do. Collect the administrative fields in a form, where they belong, and run the questions that need judgement separately. Nothing has to be migrated in one move.

Do candidates need to create an account?

No. They get a link and they type. No account, no install, no recording, no scheduled slot. If they can fill in a form on a phone, they can do this.

What stops a candidate answering with ChatGPT?

Nothing, and pretending otherwise would be dishonest. The questions ask for specifics from the candidate's own work, and the report tells you whether the answers actually named anything a person could check. An assistant writes fluently about a project it has never seen, and fluency without specifics is exactly what shows up.

How many questions should a screen have?

Three to five. Enough to cover the competencies that decide the role, few enough that people finish. A long questionnaire filters for free time rather than for skill.

Is this an applicant tracking system?

No. It covers one step, before the interview. Whatever you use to track applications and schedule people, you keep using, and the results export as CSV.

What happens to candidate data?

It is processed on EU infrastructure, kept only for a period tied to your hiring process, and deleted automatically after that, whether or not a decision was made. A data processing agreement under Art. 28 GDPR is published and forms part of the Terms.

Which one you need

If the form is ugly, get a nicer form. If you have lost track of who is where, get an ATS. If you are reading a column of free text and quietly aware that the bottom half of it is not really being read, that is the problem this page is about, and no form builder solves it.

Pricing, including the free tier, is on the main page. The longer version of the volume argument is on the high-volume roles page, and the same idea applied to protecting senior engineering time is on the engineering hiring page. If you are weighing this against recorded video answers instead, that comparison is on the one-way video page.