Few people asked what I changed that caused this sudden change in incoming leads.
I do not know what exactly worked - but here are few things I changed.
1. It may be the calendar - but my intuition says its not just that.
2. Made my profile more keyword specific instead of lots of text. Uploaded a picture. I also researched a lot on other successful Elance profiles. There were two categories there a) Sounds cool to other freelancers, developers - I switched from this to the one below: b) Well formatted, keyword rich (for humans too) and shows up better in search results and shows up important stuff in the shorter profile views.
3. Removed the hourly rate from my profile. I did not charge much lower later either. But I figured out that fighting for that $1 extra is not worth if I am going to lose a lead and spend more time looking for other work. I think 49 < 50 should be represented as 49 <<<<< 50 (These are not my actual hourly rates). This may have been a biggest problem in getting invites.
4. Changed my profile such that it is more focused to what I want to do and passionate about (data-science/machine learning) rather than what I am good at (web-development). This doesn't really seem to have any impact on the kind of work I get - I get both equally. This has more due to my personal choice - I did not want to do web-development at my day job and at freelancing (it kind of gets boring)
5. Updated my blog with one pending post I had created earlier but had been lazy to upload it.
---
6. Reduced a lot of self-pitching in the proposals. I carefully choose whom I write proposal to. Client history, Previous hiring rate/cost, Client's project award ratio, understanding of technology, understanding of requirements, timeline suggested. I tend to focus on leads who seem to have a lesser understanding of technology but better clarity on requirements.
7. Chances of getting selected when somebody invites you are far greater than competing with 50 or 100 proposals. If you are invited - it seems that you are more likely to get work at your preferred rates.
8. Placing bids when the requirement is fresh and new (or has less proposals). Chances are that in case of lower proposals, client has received many proposals like "I can do in it $XXXX". A good proposal with relevant questions tends to stand out. If something has more than 12-15 proposals with rated freelancers - I back out unless I think I can make a very strong proposal or if this project is a "must-try". There may be tempting high value projects 3 days old with 80 proposals - chances are that the client has started discussing with somebody else in those 3 days. I used to spend a lot of time on such postings, and almost immediately discard those nowadays.