I am a Pro Tem moderator on Quant Finance SE. We are in a struggle at the moment to get sufficient questions to make it out of beta. One particular issue is that we narrow the question scope to a professional level, which for the quant world is (1) very niche and (2) very high-level. This seems to have limited participation dramatically.

The closest analogy I can think of to Quant SE's ideals is TCS SE. How did this site get the traction it currently has? How was it able to maintain participation while requiring PhD-level material?

up vote 11 down vote accepted

My guess is that a lot of it has to do with that computer scientists also form a real-life community. We read each other's papers, see each other at conferences, try to collaborate. I think seeing people you know (or at least know of) in real life participate in the site helps new members want to join. I don't know if the same thing is possible for quants as, from the little of what I know, strategies, etc., are closely guarded company secrets.

But I'm sure there are other reasons why TCS-SE was successful, including our efforts in promotion, us using real names, and being a genuinely useful tool for researchers.

EDIT below:

To explain some of our publicity efforts. We 1) have promoted TCS-SE at conferences, 2) blogged and tweeted about the site, 3) wrote articles in professional magazines announcing the site, 4) tried to formally connect with SIGACT (a TCS professional group), 5) told our many peers about it.

Perhaps you can do the equivalent of some of these steps?

  • 10
    We did a lot of promotion. I feel we need to do even more. – Suresh Venkat Mar 10 '11 at 6:37
  • @Suresh, I agree :), it has been sometime from our last promotion meta discussion, would you like to start a new one on how to promote the site further? – Kaveh Mar 10 '11 at 7:43
  • @Kaveh: Good point; FCRC is in less than 3 months from now, and it would be an excellent opportunity for doing some promotion. – Jukka Suomela Mar 10 '11 at 16:48
  • 2
    done. look at the system message – Suresh Venkat Mar 10 '11 at 17:49
  • @Suresh, that is nice. :) – Kaveh Mar 10 '11 at 18:04
  • 2
    @Lev, nice summary of our efforts thus far. – Suresh Venkat Mar 10 '11 at 23:46

I hesitated to answer this question, because I don't know if we are a success yet. However, two things definitely helped:

  • Mathoverflow showed that such a site could work, and could attract the best researchers in the field.
  • The active participation of theoretical computer scientists in mathoverflow made us realize that we needed a site of our own.

So in our case, the community was already there.

  • Personally, I was very skeptical about CS leaving MO. I'm still not entirely sure it was the right choice (there are CS-relevant logicians on MO who don't seem to follow CStheory), though things are certainly developing along a better path than I feared. – Neel Krishnaswami Mar 19 '11 at 11:43
  • I will admit being very ambivalent about this early on. My main concern then was the same concern I have now with creating a "super-CS" site: having each community be drowned out. – Suresh Venkat Mar 19 '11 at 15:39

One key question you want to ask yourself is this:

Where are the professional quantitative finance communities online today?

Identifying those communities, and approaching key members of those communities to get their thoughts on what it takes to build a great community within that topic space, is always a good idea.

If there is no such existing community online, that will affect how you proceed, since you'll be bootstrapping from scratch.

Echoing Suresh, CSTheory is still some way from being successful, as it only attracts a small fraction of the TCS community.

However, it is important that early users set a high standard, and it might be worth highlighting the especially thoughtful questions and perceptive answers at quant.SE to date. Perhaps blog posts, an overview article on the Meta site, or an article in a quant publication such as Wilmott magazine could all assist. Such initiatives have all worked for CSTheory as mentioned by other answers. The list at https://quant.stackexchange.com/questions?sort=votes seems to be a great starting point, and I have found voting feedback at CSTheory useful in choosing which questions to highlight, for instance when adding entries to Radu Grigore's list at Broadening the scope of questions and for Nominations for publicity poster at FOCS ?

Listing some of the "greatest hits" allows potential contributors to quickly find content that they are likely to find compelling, and also more firmly establishes the standard that contributors are going to aspire to.

  • 1
    This is fine, but involves the risk of amplifying existing biases in terms of topics. In other words, showcasing popular existing questions will attract more of the same, but not spread very much. That might have happened with early cstheory. So promoters should maybe explicitly dig up interesting questions from areas other than their own, post them and showcase them as well. – Raphael Mar 16 '11 at 13:09
  • 2
    @Raphael: Radu's list was precisely an attempt to do this, to bolster Theory B. – András Salamon Mar 16 '11 at 13:12

I'm not a member of the real-life professional CStheory community...however, I think one major reason why CStheory.stackexchange may have succeeded is that there's no other viable online location to discuss CStheory subjects. There also used to be...comp.theory was pretty good before it got overrun by spammers and other off-topic posts, just as sci.math used to be pretty good before it got taken over by people who don't believe diagonalization proofs are valid, etc.

It's true that there are blogs...but there's no other open forum where you can really get a question answered. As a quant finance person, you'll have to compete with places like Wilmott (although IMO, a lot of those people are a little over-political on that forum, and that scared away some people (e.g., me)).

So, you have two concerns: supply and demand. I.e., is there a demand for a quant finance Q&A site (probably, yes); and is there a lack of supply of other quant finance sites out there (maybe not). I think the only thing you can really do is try to build a niche that addresses questions that Wilmott and other forums don't/won't/can't cover. Or, build a community that specifically attracts people who don't like the other sites.

  • 1
    We've discussed the competition before. Essentially, SE in general is for Q&A, not news or discussions. Plus, it's limited to professionals, not curious on-lookers. And finally, the voting aspect means that higher quality material will be easier to find. It's the same as ask how StackOverflow competes with Reddit Programming; they're very different forums in the end. – chrisaycock Mar 17 '11 at 16:21
  • It's true. However, with StackOverflow and programming in general, you're dealing with a subject that many more people care about and have questions about than TCS or math or quant finance. It's true that your forum is different...but given a finite number of people who care about quant finance and their finite amounts of time, it's likely that some people will end up choosing your competition and not you. I think for you, it'll be a question of tailoring the style to suit the most underrepresented target audience that you can. Maybe limiting to professionals is the problem? – Philip White Mar 17 '11 at 16:31
  • 1
    ip Maybe limiting to professionals is the problem? Suresh already answered your question about this: "a site for experts will draw non-experts, but not the converse". – chrisaycock Mar 17 '11 at 17:18
  • 1
    I agree that a site for experts will draw non-experts, but not the converse. What CStheory does not do is limit to professionals. The "curious onlookers" you describe are welcome, provided they can meet the criteria for questions and answers on the site. Indeed, if this were not the case, I would not be here. Of course, the site is geared towards experts, but it's not intended to be exclusive. – Philip White Mar 17 '11 at 17:21

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .