Ask HN: Does passwordless hurt our conversion rates? | Hacker News Reader

Ask HN: Does passwordless hurt our conversion rates?

1 pointcheerioty4 comments

Hi there,

When we started work on our new workflow/tasklist app, going with passwordless felt like a no-brainer. While it felt like a neat thing for new users that adopted the product for creating something, it felt even more appropriate for the people who get assigned/invited to a workflow/tasklist by others.

And while it worked amazingly well for the latter, it seems to hurt our conversion rate for new users/creators. We see about 10% of people coming to our website or landing pages to launch our app (which feels okay-ish), but then another 10%+ drop off at the login/signup. We only ask for their email address or mobile number; that's it.

Is anyone else seeing such a high drop-off rate for their login/signup page? Or is there anything else we might be missing? Should we try and provide an experience without login and only ask them to log in before they want to assign something to somehow else, maybe?

Our app website is https://www.questmate.com, and our login/signup is at https://app.questmate.com.

Any suggestions are greatly appreciated!

Edit:

I should clarify that we send out OTPs, not magic links :)

As someone who uses a password manager, "passwordless" is a major downgrade.

With a password manager, I can click on the site's entry in it and it'll automatically open the page, fill in the login form and submit it. One click and I've opened your site and logged in. If I manually end up on your site I can just hit a keyboard shortcut to auto-fill the login page and submit.

With "passwordless", I have to type my email manually (not all password managers properly support only filling in the email address without also looking for a password field), then wait ~10 seconds for an email to arrive, click on the email link (or copy/paste the OTP) and don't forget to delete the email.

I will not be giving anyone my real mobile number, so that's right out.

If you want to reduce friction for users initially exploring your app, don't require them to input any account details at all. The time to do this is when they want to keep the data.

Many SMTP servers have delayed delivery for untrusted source servers (eg. via greylisting or any other spam checking system).

I'd certainly bail if I had to wait 3-15 minutes for my password to arrive before I can enter it on... where was that tab?

Anyway, this is just a theory, and I am not sure you can easily (dis)prove it due to the nature of SMTP, but perhaps do a small scale experiment where you allow a subset of users to set their password, and see if the drop off rates are identical?

Great idea, we’ll do a such an experiment.

Reply on news.ycombinator.com