Bitcoin Faucet - One Month In
The 15th of November marked the one month anniversary of my bitcoin faucet, So I'll once again share some stats and thoughts about what the month has been like.
I wanted to write this earlier, but life keeps me pretty damn busy. ;)
Stats
FaucetBOX.com List Ranking
My best rank for the month was 10th place, on Nov. 9, and since that day, my faucet has remained in the top 25, which is well in line with my prior expectations. I've also noticed a tendency of some sites to falsely elevate their positions in the list hoping that it will lead to more traffic. You'll find my thoughts on that below, in the section about traffic.
Average Hourly Payout per Day
The demand on the faucet is still rising, and it's grown to the point of paying out around 0.5 BTC per day. On the 15th, I rebalanced the faucet so that it pays out 1750 satoshi every 2 hours. My hope was to lower the average hourly payout by around 40% to bring my spend more in line with my revenue. However, this didn't change the payout much, but my pageviews went down as expected. Thus, I'm going to rebalance it to 960 per hour, and see if that gets it closer to where it needs to be.
Traffic
In the latter two weeks of the month, you can see clear growth in traffic, mainly due to an increase in referral traffic, but also in part due to attacks against the faucet.
Let's see how the referral traffic breaks down:
| --------------- | ---------- | | Source | Referral % | | --------------- | ---------- | | ifaucet.net | 48.99% | | Other | 34.85% | | Advertisements | 6.48% | | bestrotator.com | 4.86% | | makejar.com | 4.33% | | Faucetbox | 0.37% | | Scanthebox | 0.13% | | --------------- | ---------- |
Unsurprisingly, ifaucet.net has continued to be the top referral source, and the traffic from FaucetBOX has increased slightly, but, so far my data doesn't support the idea that the ranking in the FaucetBOX list is really doing anything directly to deliver more traffic.
On Bots and Anti-Bot Measures
During this initial 30 day period, I've seen my fair share of attacks. Some from bots, and some from humans. Unsurprisingly, the bulk of IPs that my system has auto-banned come from ISP broadband pools. Now, whether this is due to people being paid to rip-off the faucet, or due to zombies sitting on those networks, I can't say for sure. But, I can say that in any case, this is the hardest threat to stop.
With bots that operate from datacenter networks, it's fairly easy. You just make a list and ban them. But, when you're dealing with ISP nets, it's not so simple. What I've noticed with regard to bot behavior, is that the botnets I've seen tend to have nodes is datacenters all over the world. They do their best to fly under the radar, and try to evade my auto-banning logic. They will rotate bitcoin and referral addresses, usually operating in a rotation of different faucets.
The CAPTCHA does help in stopping bots, but it's practically useless against mass-human attacks. The only thing that really seems to help at all in this regard is rate limiting. The biggest problem there is with the "Turbo Mode" feature of Opera and YaBrowser, since these are proxied through hosts in various nets. Occasionally I notice that some of these proxies winds up getting rate-banned. My apologies to those who are using such features, but I haven't yet identified all the nets being used.
By the way, I was thinking of throwing together a faucet script based on my experiences dealing with these bots, etc. If anyone's interested in using it, leave a comment in the form below the article.
On Revenue
Here's what my bitcoin revenue looked like over these 30 days:
Anonymous Ads started out being a fairly strong source of revenue, but I suppose I just got in at the right time. You can see that over time, I've earned less and less from that particular channel. This lead to me adding a second AdSense block, and a Mellow Ads block. The Mellow Ads block has performed far better as you can see.
The FaucetBOX source is referral commission from other faucets which are linked on the faucet page, and has been performing well.
Now, I'll talk about the channel that's really on your mind, AdSense. AdSense has performed about as well as could be expected. Unfortunately, the majority of my traffic carries a RPM that's a little on the low side. All I'll say with regard to hard numbers, is that the faucet pulls in around $60/day as a channel, with an average RPM of under $1.00.
The greatest pain in the ass in this regard, is when my AdSense revenue gets docked for "Invalid Traffic." For the month of October, I lost around 18% of my AdSense revenue for this reason, and naturally it has to do with the bot/mass-human attacks.
Conclusion
The faucet is not yet self-sufficient, but I'm working to get it there. I've just rebalanced the faucet again. It's not an easy decision to make, and I predict it will have some consequences in terms of traffic. But, it's necessary to keep the faucet at a runnable level. If anyone thinks I'm just being a cheap-ass, keep in mind, all that satoshi flying out of the faucet has been largely funded from my own pocket.
I started the faucet at 1500 satoshi / hour, then went to 1750 / 2 hours, hoping to lower the payout a bit, and it had no effect. Now, I'm taking it to 960 / hour in an effort to better balance what's going out with what's coming in.
The burning question on my mind is, how much revenue will I have to forefeit this month because of bots, etc.? I'll be sure to write about it. ;)
Ultimately, my goal is to provide a fair and decent quality service for my users while making a little revenue out of it. This is precisely why my anti-bot measures must be as brutal as they are, and unfortunately, this will upset some of my users. But, that's a risk I have to take.
I also mentioned possibly creating a faucet script, which I may do if there's enough interest. What would you really like to see as a faucet owner? Better anti-bot measures, more analytics, or something else entirely? Let me know in the comment box below!