How to get to the front page of the Market - page 9

 
abolk:

Everybody knows perfectly well, and meat is absolutely right about that - 99% of the stuff on the market is useless.

Please don't speak for "everyone" and cite figures "from the ceiling".

 
DiPach:

Please don't speak for everyone and quote figures "from the ceiling".

Yes, you're right, the key point is that the 1% understand that 99% of the market is rubbish
 
abolk:

I was a little excited with all the products, can I leave out the trading boards (buy, sell, etc.) and indicators?

Will at least one person on the forum give you an answer - where are the standards for junk?

Hello, here we are - but the risks in the financial markets, it's all a fairy tale :)

No one wants to discuss my suggestion on the construction of the product page? Let us not deviate from the goal of making a market, the more we "bicker" the longer .............

It's about time we had something constructive to say.

 
abolk:
Yes, you're right, the key point here is that the 1% know that the market is 99% junk.

That sentence was also about numbers from the ceiling. Applies to the numbers you mentioned here as well.

 
Urain:

A multi-criteria genetic optimisation algorithm.

Add personal ratings to user profiles (those favourites for products and ratings for ordering in favourites) - this is a purely private section, each profile has its own.

The user will get an ordering of what they are interested in, you will have access to this data and you will be able to assess the pent-up interest in the product.

It makes no sense to build ratings for each user.

This is a wild cost with zero return. For many companies, behavioural data mining leads to zero or negative results and it is often easier to settle on a random strategy or the most primitive "with this they buy this" calculation.

 

server:

Let's not get sidetracked from the goal of making a market, the more we "bicker" the longer ............

It's about time we had some constructive feedback.

Almost at the same time as you published the posts.

So without intent on my part it turned out to "...deflect from the goal..." despite your sensible comment about constructive. Sorry.

 
Renat:

There is no point in building ratings for every user.

It is a wild cost with zero return. For many companies, behavioural data mining leads to zero or negative results, and it is often easier to settle on a random strategy or the most primitive "with this they buy this" calculation.

The "with this they buy this" is a muving with all that it implies, the muving will never show where it will swing, I suggest to show pent-up demand, those are insiders.

The forum has favorites for topics, it's true there is no custom sorting, I don't think it would cost much to set what should be displayed above what is below.

A user that tidies up their profile automatically gives you an insider's view of their intentions, you just have to summarize the information.


Look you cache data that I viewed, what topic visited and whether there are new messages in the topic, while it is individual for each user, and yet the topics on the forum abyss.

And products in a user's favorites will be a mizzero dozen or 100 at most, it's expensive to give them ratings?

Let's do the math: 100 products*200 000 users = 20 000 000*4 bytes =80 000 000 bytes ~80 Mb to process all user rating activity.

 
Urain:

"With this they buy this" is a muving with all its implications, the muving will never show where the swing is, I am suggesting to show pent-up demand, those are the insides.

I'm afraid we don't have a track record for 99% of users.

We are talking about deeply sleeping masses of traders who do not go to the site at all. For them, we just launch mechanisms to buy services and products without registration at all. Otherwise they do not react in any way.

So, the challenge is still the same: how to make the top rotate automatically and intelligently, so that it does not disadvantage the best sellers, but also gives others the opportunity to rise?

Perhaps a good enough option (with zero sorting and analytics costs) would be to rotate 50% of the top randomly at all. Or perhaps leave the top as it is.

 
Renat:

For example, your suggestion "let every seller, without having sold anything yet, pay for placement in the marketplace" is an excellent demonstration of both points: the advice to kill the service and the focus on individual advantage. I'm not even talking about the second step - what happens when the top is flooded with "honest" spammers and cheaters.

OK, then what is your vision of the Market as a service in general? Perhaps we are looking at it from different perspectives. Do you see it as a business? Or as a stage for young talent to perform?

From a business point of view I don't see anything strange in my suggestions. Have you seen a marketplace somewhere that offers outlets for free to all comers? It's scary to imagine what it would be like then. In any business, the entrepreneur "pays before he sells anything", bearing the market risk. What's strange about that?

 
Renat:


Perhaps a good enough option (with zero sorting and analytics costs) would be to rotate 50% of the top randomly at all. Or perhaps leave the top as it is.

Alternate the top 50% by 50% with the rest of the products - by time, or by algorithm (just which one? )
Reason: