Quantcast
Channel: JohnGrginov.com » Captcha
Viewing all articles
Browse latest Browse all 2

HUGE Spamvilla ReCaptcha OCR Case Study

0
0

Here is a sort of followup on the recent Captcha Breaker vs Captcha Sniper test I did. Besides being a fun way of creating content for my blog, it was something that needed to be done. Both sides were trigger happy with claiming the title of the best captcha ocr. The test mearly confirmed what I was expecting and now stands as a great resource for anyone doing some kind of linkbuilding. It showed that CB did better and that it’s one of the best ways to save money if you’re building lots of links.

However, the test also showed that CB is not as powerful on it’s own when it comes to building links on high quality sites that are used in your first tier of links. Sure you can use it as the only captcha service and you will get a decent number of links but you might miss out on quite a few of the good ones. And the problem gets worse the higher your PR filter is set. Lots of these sites will have Recaptcha and despite what anyone might say, CB can’t solve it reliably. Most of the time, people will deal with this problem by setting CB up so it deals with easy captchas and passes the hard ones which it can’t solve (mainly being Recaptcha) to a service such as DeathByCaptcha (DBC).

This wouldn’t be a problem if it weren’t for the on-going cost of DBC and similar services which only increase as you’re building more and more links. There was a need for something a bit different…

Introducing a ReCaptcha OCR

This is a new service from Spamvilla, one that makes our lives hell of a lot simpler. It’s an OCR that can solve ReCaptcha with a pretty good accuracy (advertised as 50-65%). What’s best about it is that it works with CaptchaBreaker as well as extra cheese on a pizza.

CaptchaBreaker options CaptchaBreaker + Spamvilla Re-captcha OCR integration

Basically, if you send a recaptcha to CB, it won’t try to solve it but instead it will automatically send it to Spamvilla’s OCR. Every other type of captcha will be handled by CaptchaBreaker itself. It’s a simple but powerful feature.

The cheapest package is $30 / month and from what I’ve seen, there’s absolutely no need to buy the bigger ones unless you plan on running it on multiple computers.

But let’s put all that generic info aside and go for something more interesting. Let’s see how this OCR compares to the likes of DBC.

Test #1 a.k.a. “High Quality Pwnage”

The goal here is to see what is the most cost-effective way to go about solving captcha when building tier 1 links. To do this, I’ve done the following:

  1. Setup 2 identical projects in GSA SER: 1 using CB & DBC, the other using CB & Spamvilla OCR
  2. Compiled a list of 45k PR1+ sites that could be used as tier 1 links (article, social bookmark and social network sites)
  3. Generated very similar content for each project using my favorite content generator.

Now all that was left for me to do is to fire up the projects, wait while they finish and take a look at the number of successful submissions for each projects.

Test 1 - Spamvilla vs DeathByCaptcha results

Spamvilla setup resulted in only 682 less submissions than the DBC setup. When you take into account that the initial list consisted of 45,000 sites, we see that our Spamvilla setup is 98% effective as the DBC setup.

Even tho this is quite huge and you could cut all your on-going captcha costs to $30 / month with only a neglible loss in the number of links you are able to build, the catch is kinda obvious.

If your monthly captcha costs are < $30, you can’t really justify the monthly price of the OCR. 

Or so you would think…

Test #2 a.k.a.”Blasting Away”

After completing test #1, I was really stoked. The idea that you can use an all software solution that doesn’t scale it’s cost the more links you build is huge! But my enthusiasm quickly died off.

I’d imagine most people doing linkbuilding never spend more than $30 on DBC per month for tier 1 links. And for them it’s probably cheaper to keep using that vs a monthly service. Only those that had worked on larger number of sites would build enough links to justify this purchase.

This is when it hit me.

Why should you use something like this only for your tier 1 links? Sure it’s not profitable to use DBC for tiers 2 and because the number of links being built here is much higher and your on-going captcha costs would be insane.

But the OCR has a fixed monthly price for unlimited captcha. Why wouldn’t you use it in those tiers as well to help Captcha breaker out? Surely this would increase the number of links you can churn out to your lower tiers since even many of the lower quality sites use re-captcha. The only question was: how much better will it be? Time to find out.

I did the same thing as in test #1:

  1. Setup 2 projects, 1 using just CB, the other one using CB + Spamvilla
  2. Compiled a list of 110,000 sites of every platform imaginable (that GSA SER recognizes) and PR ranging from ? to 9
  3. Created very similar content using WAC

Here’s how they performed:

  1. CB project – 38,363 submitted links
  2. CB + OCR project - 42,180 submitted links

The added bonus of the OCR guaranteed a better run tho that meant the project took longer to complete as they recaptcha takes longer to solve than regular captcha does with CB.

When you should or shouldn’t use Spamvilla’s reCAPTCHA OCR

So we learned that it doesn’t really matter if you’re using DBC or the OCR (in addition to CB) when doing tier 1 submissions. We also learned that by adding the OCR to CB, you will get more links no matter what site list you use. Even so, you might have a hard time deciding whether to comit to something like this. Here are some reasons for and against using this

When you shouldn’t use the OCR 

  • If you don’t have any on-going captcha costs
  • If you’re comfortable with the number of links you’re able to build
  • If you’re posting to sites that don’t use re-captcha
  • if you’re building only a small number of high quality links without higher tiers
  • If you want to nuke as many links as fast as possible (while using it will increase the number of links you get, it will take longer to build them which kinda negates this benefit)

When you should use the OCR 

  • If you’re building loads of tier 1 links
  • If your on-going captcha costs are more than $30 a month
  • If you want to be able to build more links in higher tiers
  • If you’re doing a lot of posting to sites mainly using re-captcha

If you recognize yourself in one of those 3 examples, you should definitely check out this one of a kind service as soon as possible.

That’s about it for this case study! While it wasn’t as professional as I would have liked it to be, I think it’s a fair representation of what this service is capable off and tells you exactly why it might help you in your daily linkbuilding efforts and when it might better to save your money or spend it somewhere else.

If you happen to use this service, please share your experience with us down in the comments.


Viewing all articles
Browse latest Browse all 2

Latest Images

Trending Articles





Latest Images