Sorry, what do you mean by normal influencers? The "famous" people? Would be great anyway to know more about your plus against them
Normal influencers are commonly the ones with over >100'000 Followers
The biggest difference between Mavin and the rest is that we are baseline up.
That means our system is made to scale.
A normal influencer plattform cannot scale down to nano and micro influencers because there is to many work.
You need to know that a normal influencer campaing only has abount 10 influencers working for that campaing.
It's very easy to work with onyl 10 people and check their posts.
But if you want to do it with 1'000'000 people than you need a system which is made for that. And that's MAVIN.
Verification of the quality of the influencers will be managed by an algorithm and human together. You have already selected these people ?
Are people internal of project or external ?
No, thats the best part of the story:
everyone can be a verifier 
if all of us can leave a feedback, can be a verifier.
it's every time the same game...
I leave a real positive feedback
I leave a fake positive feedback
I leave a real negative feedback
I leave a fake negative feedback
how to filter fake feedback???
Because your feedback is also rated agains the consensus of the other verifiers of the same post.
The system will detect your fake ratings because you are not in consensus with the others.
If the system detects has too much discrepancy then he just adds more verifiers for the same post.
That means that sometimes a post needs 5 verifications and sometimes 20 or more, the system will only accept the verification as completed when the numbers are within a certain threshold.
Take this picture as example:
https://scontent-frt3-2.cdninstagram.com/t51.2885-15/s640x640/sh0.08/e35/22794206_184634598768007_5656685964232228864_n.jpgOne of the questions is : "How good is the brand visible in the image"
Verifier 1: [...................x]
Verifier 2: [.................x..]
Verifier 3: [..................x.]
Verifier 4: [...................x]
Verifier 5: [...................x]
Verifier 6: [...............x....]
Verifier 7: [...................x]
Verifier 8: [.............x......]
Verifier 9: [.............x......]
The system will calculate the min,max,avg,median,variance,stdDev and determine the score for this question
In this case the final score would be the following [.................
xx]
and verifiers 6,8,9 would be protocolled for further inspection if this happens to many times and repeatedly
Another example:
Lets take the following result
Verifier 1: [...................x]
Verifier 2: [.................x..]
Verifier 3: [............x.......]
Verifier 4: [.......x............]
Verifier 5: [.x..................]
In this case the system would detect too much deviation
and add more verifiers.
Verifier 6: [.............x......]
Verifier 7: [..x.................]
Verifier 8: [.....x..............]
Verifier 9: [x,..................]
Verifier A: [.....x..............]
Verifier B: [...x................]
Verifier C: [.x..................]
Verifier D: [x...................]
Now the system would have enough information to determine the score in this range [.
xxxx...............] which is low and the post would be rejected if this merged with the other results would be below a threshold value.
). If I can use an eth wallet as MEW I don't see limitations for the monetization part. It seems a good point for the project, in my opinion. 
