1

I'm creating a web project where users can add markers on a map. I'm dealing with a demographic who are used to instant-gratification. I want to keep the threshold for engagement as low as possible so I'm considering making user registration entirely optional. There are a couple of pitfalls I foresee but since I know I'm a bit of a pre-optimizer (which is not always a good thing), I'd like to hear your opinions on the matter.

I think some of the most important risks are:

  • Spam-sensitive: since you are not required to identify yourself, it would be easier to get away with spamming the map with markers. I've seen it happen to a friend's project. He ended up with markers depicting genitals covering most of France.
  • Untrustworthy: because the source for the data can be anonymous, the data could be seen as unreliable.
  • Non-human interaction: one of the things having authentication helps with (usually) is barring machine data entry / spam bots. Leaving the app 'in the open' could attract bots.

Things I have considered to mitigate these problems:

  • Use only a CAPTCHA for anynomous data entry. I feel this is the bare minimum to do something about bots. Preferably use Google's one-click "I am not a robot" to keep the threshold as low as possible.
  • Offer some form of perk when using an account vs. anonymity. For example, earn reputation for your data entries based on votes from other users. However, I doubt most people will care enough.
  • When entering data anonymously, keep the data around for only a temporary period, for example 30 days. The data gets removed unless it is confirmed by a registered user.

The project will mostly be used by data consumers rather than data creators. This is the main reason to make data entry as user-friendly as possible. Because every consumer would automatically qualify as a creator, I just need to convince them it's worth their time. Therefore, I want to waste as little as possible on account registration.

  • You will also be dealing with malicious individuals, who will enter invalid content, perhaps a lot of it, completely ruining the usefulness of the map. – Ben Jul 13 '16 at 2:45
1

What if you could have your cake and eat it too? Here is what I am thinking:

  • Allow new users to drop pins on the map without forced registration.
  • Trigger an optional login/create account request after they've dropped a few pins and am feeling proud about their work.

The request must explain that pins coming from anonymous users are not immediately published to the community and instead are subject to quality review. The other benefit you should focus on is attribution - that they will "get credit" for their submission.

UGC contributors often perceive that other participants enjoy their UGC and trust their input. This boosts the contributor's perceived competence and autonomy [source].

These motivations mean that login/registration is key to getting what they want: their pins published immediately and credit for their work. You've turned the tables on them :)

If they do not login then it would be important to maintain session data alongside of each pin to help in any cleanups of unwanted content. You do not want to be in a situation where an otherwise high quality map is destroyed by someone bombarding it with pins - you would not be able to isolate the good from the bed.

  • I like it. But rather than quarantine anonymous data, I could display it anyways and allow consumers to toggle them to improve signal/noise. – Marc Dingena Jul 13 '16 at 6:26
  • In doing so you remove one of the major motivational factors we worked to create. I would at least begin with a model where content is quarantined - and see how much anonymous content you actually have. The other issue is that providing users with the ability to segment the data implies that the anonymous data is of lower quality - which may or may not be correct. – Tom Griffin Jul 13 '16 at 6:31
  • But if nobody can see the data, how is it ever going to be approved? It's not immediately clear from the OP, but I'd prefer if the data is also user-managed. I don't intend on curating the entire map (which includes the whole world) too strongly by myself. – Marc Dingena Jul 13 '16 at 6:35
  • My underlying hypothesis is that a user who spends time creating high-quality content will almost always create an account in order to expose that content to the rest of the community. They are not going to want to see their work go poof or have it subject to review. Content that is abandoned was likely created under a different set of motivations and is more likely than not junk you wouldn't want anyway. All would need to be tested be refined based on actual observation. – Tom Griffin Jul 13 '16 at 13:03
  • Having the community police itself works when you have a critical mass of active, contributing users. It's common for sites to handle quality control themselves until they reach this level of maturity. You'd never be approving pin by pin - you'd approve at the session level where it will be more obvious if someone is screwing around or not. Finally, there is always the option of faking it: if users are more likely to register in order to publish immediately but you don't want the burden of quality review, enforce an arbitrary delay before automatic release. – Tom Griffin Jul 13 '16 at 13:17
1

Having seen first-hand what un-moderated map pinning results in, there are a couple of ways that users will try to abuse your system:

  • Using pins to draw or write on the map.

This is easily solved by simply limiting the number of pins available to non-registered users. Once they reach their limit (anywhere between 1 and 5 - just below whatever you're expecting average usage to be) you could trigger a message to encourage registration.

  • Using pin labels for profanities

Depending on what you're planning to allow pins for you may either need to obscure these (as in "F**k") or filter them out altogether during form validation when the user clicks the 'save' button on their entry.

  • Using pin labels for spam

This is a little more tricky in that you will need to understand what constitutes spam for your use-case. A few possible patterns to look for may be links or link equivalents (www-dot-website-dot-com for example), email addresses or their equivalents, and telephone numbers. As with profanities, these could be filtered out when validating the input form. Of course, an incentive to register could also be the ability to include links in the label.

As for incentives to register, there are a few above - bypassing the various anti-spam precautions and more trust for content - On top of that you could add more detail for registered users: a more structured for for label content (title, summary, link, phone number, photo...) to encourage them to create better pins. A points system is not a bad idea either - especially if you include points for frequency of acceptable pins, geographical spread of pins, number of pin categories... a bit like a virtual Foursquare - if done carefully this would not only encourage registration but also more useful, accurate, and detailed pins.

  • I was thinking about rate-limiting instead of capping their maximum allowed anonymous pins. But creating a cap is easier to implement and a stronger encouragement for registration. Thanks! – Marc Dingena Jul 13 '16 at 7:26
  • Both are easily circumvented. And if the restrictions are too tight, your ux will suffer even more :-/ I would recommend I'll be starting with the simplest approach and then working your way back. – Tom Griffin Jul 14 '16 at 5:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.