I'm wondering how you can review an account about the respect of the policy if you are not suppose to collect any search data or activity data about members?
If privacy is your main concern, what data are you collecting to analyse?
How can you determine that at the withdrawal moment a report with no data??
I had the perfect answer about this question on Telegram from an admin: "The review is based on search pattern and not what you search."
So I still believe in their privacy! Gonna HODL!!
Thanks for your support, CryptoTokenAnalyst.
We do not store or view search terms.
It is really based on your patterns, and requires a fair bit of time to identify abuse.
That is why the token withdrawal threshold (1,000 eligible tokens) is on the high side.
Time provides data to make a real-time determination of how similar your behavior is to behavior it has been trained by a human (often myself and others like Lindsay, Jeff and Thomas on our team) to identify as skewing toward 'mining' tokens or using bots.
It communicates with the Member by providing them with a 'level' that scales from 0 to 10, with 10 being the least likely to be trying to collect tokens by skewing their search intent (ex. searching more than needed for things that you aren't truly looking for information on - for example: 'aaabbbccc' or 'random word').
One of the biggest challenges with incentivized search is that if you lose true search intent, the quality of the query decreases significantly. Poor quality equals a lower signal to noise ratio, making the overall stream of queries less valuable to those who may want to target their sponsorships to the keyword being searched, which significantly improves ROI while providing searchers with a more relevant, less distracting message (which they are compensated with PRE tokens to be exposed to).
This is the real underlying magic to Google's Adwords system, which is the most successful business model in history from an ROI standpoint (for both Google and marketers).
Part of Presearch's underlying magic is the token reward system, and the fact that the action and reward are so directly related. Hit search, receive a reward visualized as a token.
To protect the magic of both models is critical to the success of the project's token economics. It is possible to run alternative monetization models, but the token is the value storage unit that aligns both Members and Sponsors. To do it similar to BAT and others are doing it does break the direct alignment, whereby a token really isn't the actual value storage unit and is an abstracted layer primarily designed for market trading and project funding.
Nothing wrong with that, but just a different level than the Presearch approach.
So, all that being said, protecting the project from abuse is extremely important. It is also extremely complex. We think we have developed a pretty solid approach, but it is one that is in the early days and will certainly evolve.
We must continue to improve our communication with Members, provide value to Sponsors, and also improve our transparency with regard to this process and others.
We are moving in that direction as we put our initial efforts out into the public for trial and feedback.
We appreciate you and others taking the time to help us learn and improve. Please keep the feedback, good or bad, coming.
Thanks for your interest in Presearch!
Colin
