So, we found two important factors:
There is the sunk cost effect, aiming to coerce the user into to clicking "Like" and finish registration.
And there is the increased risk for loosing a potential customer, by disgusting him before the first login.
The first is increasing the count of full registrations, by pushing users over the forced Like;
The second is decreasing the same count, by provoking rejection when pushing him.
There will be some balance between the effects.
When looking at persons that clicked an initial signup link as potential customers, we could empirically find the increase in Likes, and decrease in signups, when disabling or activating the forced Like feature.
Simplifying a little by assuming everything behaves linear, that gives a nice compact number, comparing change in Likes to change in signups, say
1000 likes / -10 signups
which means loosing a signup for every 100 likes.1 That could be good, or bad - hard to tell.
But there is more to it.
There is an effect of the "coerced Like" feature to the users that give in and finish registration. And I think this effect can be strong enough to be relevant.
Assume we have measured that the feature actually does reduce signup count.
So there is some psychological effect that leads to the rejection; let's call it "irritated anger".
Using the assumption that this effect is somehow continuous, causing rejection at some threshold, we see that there are users that have signed up, but were influenced by the effect.
Some users will be not influenced at all. Some other users will be influenced. With the strength of influence from weak to strong, up to the thresold of rejection will Some of them strong, up to the threshold for canceling the signup.
Now, there are two interesting questions:
- What follows from the influence in the behaviour of a user - in units of business value.
- And how "strong" is the influence on how many users.
These are not as easily accessible to empirical methods.
The point I'm trying to make is that they are relevant.
We're talking about a substantial count of customers on the site being irritated and angry about the site to some degree when first visiting it. I think this will make them sceptical about the site, it not to the company. That also directs the users attention to other features that may be intended to exploit him somehow - so he will notice these features when they are borderline acceptable, when he had ignored them otherwise. Also, when noticing negative aspects, he may remember that he had recently made a public and permanent statement that he agrees with how the site is run. And may not like that thought.
In summary, I expect that the users that signed up for the site - remember, each of them was coerced into some action - will see it in a much more negative light, in part based on increasing his attention to questionable aspects, which would normally be tolerated, or not noticed.
This obviously would affect the rate of participation, rate of users not showing up again, etc.
So, no, "Like" on a signup form should not be a required field;
Because - ignoring so many othervalid reasons - it does not pay.
The tricky part is that the total negative fallout is hard to quantify - empirically or otherwise.
1Someone else has to find the increase in "potential customers" caused by the increase in Likes.