Manifold is thinking about letting users pay to promote their markets at the top of Browse. See this proposal for a rough sketch of what we're thinking about.
Thirty days after we launch some version of this idea, I will start the clock for another 30 day period. I will then the sum of all boosts sold in this period to resolve this market. If a boost costs $100 and a user pays in mana (Ṁ10,000), I will count that as a $100 transaction. If we don't launch a feature like this in 6 months, this market will be N/A'd.
1,000@SG the first day on https://manifold.markets/stats with boosts being sold is Feb 14. Does that mean that the clock for "month 2" starts on Mar 16?
@SG starting the clock on march 16th, we should now be at $2000, with 5 days left out of 30 (according to manifold.stats).
IIUC, there are some boosts that were given by the site (at least, I didn't see a transaction to the manifold account, unless I missed something), but I'd assume those are excluded from this market. and i'd assume everything on the manifold.stats page is fair game
@ScipioFabius 25000 mana is definitely too high. Ideal price point should have at least 2 boosted markets at any particular time, and occasionally have all (5?) slots sold out.
Any time with nothing boosted fails not just on opportunity cost but also on the feature not advertising itself.
My guess is around 9000 mana is the sweet spot.
@cthor auction the spots, if there are 5 per day you could auction one every 6 hours and have the final slot be available for purchase at a rate higher than all recent auctions.
@Eliza Auctioning the spots could be a little frustrating due to all the reasons any auction is frustrating, e.g. ebay, having to babysit the auction last second. If a person really wants a market boosted right now it could be frustrating having been outbid. But I agree there definetely should be some dynamic pricing mechanism for these slots. Maybe a good start would be a flat scaling for slots, for example 5000 -> 10000 -> 15000 ... guaranteeing that at least one slot is advertising the feature itself like @cthor mentioned.
@Eliza Auctioning would probably drive the price down, since the slots are mostly fungible, so people would be happy to wait out until they can get the best price. You'd need to set a minimum price to address bid shading, and if that were picked right it would probably end up as the clearing price anyway (unless it's set quite low), mostly defeating the point of an auction. And an auction adds friction to the sales process, which is never good. It could be fun to try various auctioning mechanisms to address shading, but that's also a bunch of work to organize.
@ScipioFabius the ebay problem you mention could at least be addressed with some kind of sealed-bid auction.
I suspect with such a scaling price people would also wait it out for the lowest price, since although they're mostly fungible, a boost on its own is probably worth slightly more than a boost amongst others. Selling the lower value product at a higher price probably isn't an ideal price strategy.
It could work if most of the time there are 0 slots for sale, and people who want to boost get FOMO thinking the there might be 0 when they need it, so they overpay a bit. For that, the price of the first 4 slots would need to be quite low, and then the 5th only a little above the clearing price. So you still need to know the clearing price, and if you know that why not just sell at that?
I boosted my market with mana and, based on the results, would have been happy to pay $250 real dollars (as a business expense).
@Ziddletwix Out of the 5 there today, were they all paid for by the user or by someone else? Just curious, that's all.
@10thOfficial ah got it. hmm glancing at some balance logs, it looks like these were mostly purchased by users themselves (staff have boosted some of their own markets & ian paid for the boost to the openai subscription? but my approach isn't comprehensive e.g. at a glance i'm not sure who boosted the zvi market)
So far the Stats page shows $1600 in sales on just 5 days of usage.
If we can just keep that up starting in 25 days.....
@Eliza According to SG the initial target is to sell 5 slots per day for the most possible revenue. Priced at $100 each, that would be $15,000 over 30 days. Maybe it will end up a little lower by then. We'll see.
@Eliza Now with 9 days in the stats feed, $2700-equivalent revenue. If that pattern continues into the 2nd month it would take 7 days to reach the $2000 threshold and after 30 days it would be just shy of the $10,000 threshold.
@10thOfficial Thanks for sharing. This is a bit lower than I would have hoped, but also better than zero.
For the record, I was one of the new traders but I literally only did it to see if it would come through in the statistics somewhere (which it did).....so, take that into account.
@SG how many people have “for you” selected all the time? I use that toggle very actively (it’s super useful!) but a lot of users don’t curate topics and so they just leave FY always on/off. I think it might make sense for boosted questions to be on top for FY as well?
@Ziddletwix Boosted markets will appear at the top of For You if users are following the relevant topics. Boosting changes a market's importance score across the entire site and also affects visibility on the Explore feed and related markets too.
@SG I think that’s sensible enough but if the # of views are lower than expected I’d check to see if many have for you turned on (without following many topics) and tbh I think it’d be fine if it got boosted even for those using For You, I mean it’s functionally advertising fine for it to be pushed
@Eliza yeah I do actively toggle it but I have a bunch of followed topics AND often bet outside that. Many users haven’t followed many topics so they aren’t using the feature (or just leave it always on)
@Ziddletwix Another anecdote: I try to leave it off most of the time but it keeps turning itself back on. I don't find it helpful, it just seems to hide markets from me.