On the plus side<p>- The convenience factor is compelling, like auto-complete.<p>- As a v1 you can see the true magic in a seamless signup.<p>On the downside:<p>- It does something unusual / unexpected / creepy.<p>- It's too slow - noticeable lag vs. say the auto-suggest in search.<p>- If accuracy is good why bother exposing that to the user? Intercom.io does this all in the backend off of email only registrations and seems to have about a 70% hit rate. The data is useful to business owner but I'd bet the user would feel less enthusiastic knowing that I have their personal fb, twitter, linkedin, etc as soon as they register.<p>I am sure that svbtle can show the A/B data validating this is worthwhile for THEIR audience.<p>But I'm not sure for a broader use case this is going to help due to the factors mentioned above.
Presumably, they have some preexisting database that maps email addresses to personal information. Their Privacy page, however, is not transparent <i>how</i> exactly they got this data:<p>> Broadly speaking, we collect information in three ways: (1) when you provide it directly to us, (2) when we obtain information about you or your company through trusted third parties or indexing systems, and (3) passively through technology such as “cookies”.<p>This is a disconcerting use of "secret sauce," particularly since email addresses have the same weight nowadays as usernames.
Why would you use this instead of login with Facebook/Google/Twitter/etc., which only requires two clicks from most users and actually has better privacy implication since the user knows which data you can access?<p>Or maybe the goal of the service is for web sites to surreptitiously get more data from users who are otherwise unwilling to provide it, but are somehow willing to provide an e-mail address tied to their real identity?
Actually this make me very angry that my private details exposed in such creepy way, even if this is using some public APIs etc.<p>I would never sign up to service with such form.
On the one hand, yes, I can see how this could improve sign-up rates for those who are concerned about those things, and I count myself among that number from time to time.<p>On the other hand… yeah, creepy as hell.
Say what you will about Dustin Curtis, but everything he makes is just so nicely designed and enjoyable to use.<p>I'm curious about the pricing model: $20/mo for 2,000 requests. I wonder if you still pay for the request if it returns a non 200 or if their API can't find a name.
I have a fairly unique last name, which is in my email address.<p>So far two dead relatives that never really used the internet have been "marketing matched" to my email address. So I get spam for my deceased grandmother and uncle. Its a little jarring. I'm not sure what kind of matching they do but clearly its not working.<p>Plus isn't the point of using a service being able to decide to match it with twitter, Facebook or just not link to those things.
<i>Because it helps your users save time, Magic increases conversion rates and makes the web easier to use.</i><p>Honestly, how much time do you really save by auto completing a couple of fields? This seems like a solution in search of a problem.<p>Also, while I don't have the data to speak for other users, I personally don't see myself paying for any service that presents my own personal information to me before I even know what the fuck it is.
I don't get it :-/<p>I couldn't even try it because my browser's autocomplete automatically completed all fields.<p>Doesn't everyone has autocomplete?<p>Also, it's pretty creepy.
I suppose it's a good thing nothing came up when I entered my email?<p>I would be turned off from a website or service that prepopulated a lot of my information without my consent.
I wonder if a new way to do signups would be just an emailto: link that you click and send. From there it automatically logs you in and sends you a password.
I have no problem with this, but it doesn't work very well for me.<p>Tried about eight of my email addresses. School and work ones did nothing, and neither did my Outlook.com account, or Yahoo.com.<p>For Gmail addresses it just pulled the name from my Google+ account for the one that had it, and failed on all the rest.<p>Not super impressive IMO. Especially considering all my email addresses are some variation on firstname.lastname or flastname.
So, I guess it's cool now to overload app/service names? E.g.: <a href="https://getmagicnow.com" rel="nofollow">https://getmagicnow.com</a>.<p>But more than that... "Because it helps your users save time, Magic increases conversion rates and makes the web easier to use."<p>There's one heck of an unfounded/unvalidated assumption.
Probably a way better way to increase "conversions" would be just to not have a pain-in-the-ass login process.<p>oAuth/Persona, and then don't ask for anything else until its actually needed.<p>And now you have (for the logged-into-oauth-provider case) a 2 click login system. It can not get easier than that
Isn't there already a service called "Magic"? <a href="https://getmagicnow.com/" rel="nofollow">https://getmagicnow.com/</a>