Tell me more ×
Answers OnStartups is a question and answer site for entrepreneurs looking to start or run a new business. It's 100% free, no registration required.
up vote 6 down vote favorite
3

I'm experimenting Facebook Ads since a week or two as I got very disappointed by AdWords recently. But Facebook is acting exactly like AdWords.

I suspect that most of leads that come from them (I paid for each of them) are generated by an apparently non human system.

I spent a week monitoring each lead that came from them using a special live stats software. I observed their behavior on the website and I noticed some patterns. I compared them to what I call legitimate users which also have a specific behavior pattern that is specific to how the website is designed (where the buttons are placed on the screen etc).

However those I suspect to be robots doesn't see buttons. They see text links. The possible automatic system is pretty well designed. When you set "goals", it is able to detect them and systematically go trough them, without any coherent behavior. The only objective seems to generate a click and if possible, match one of the goal you defined in AdWords, then leave the site.

I also measured the difference by comparing download page hits (it's a goal) and actual welcome screen display (when the software is installed, the user is redirected to that page). None of the suspected automatic leads access that page.

Have you experienced the same behavior? How to avoid those fake (and paid) leads?

share|improve this question
I experienced the same with FB ads and suspect a bot. Must be a fairly well thought out scam because the IP indicates its from targeted city. However, it does nothing after landing on our site, unlike other legit visitors coming from our backlinks. The behavior is entirely different from visitors from other sources. – user12570 Aug 9 '11 at 8:50

3 Answers

up vote 7 down vote accepted

Ok I answer my own question. I hope it will be useful to others.

My suspicion was correct. Here is how they do it:

  1. They create bot-nets by infecting thousands of machine world wide.
  2. They create page full of AdSense contextual Ads.
  3. They control their bot-nets remotely to access their pages and "click" on the ads.

Google is working on the problem since a long time. Here is a page found on their own website that explains the process. It was published 4 years ago. The process is described in details in this document.

Has Google found a solution to shut them down and therefore reduce dramatically their own profits? Nope.

Here the solution I found from different sources.

  1. Click on Campaigns
  2. Click on Network tab
  3. Scroll down and expand Exclusions link
  4. Add domains to exclude in the box.
  5. If you want to exclude IP Addresses, click on Manage IP Addresse Exclusions

Now how to determine which leads are coming from bot-nets?

Use one of the following methods:

Method 1 (Manual)

  1. Download and install a live website tracking software. I use ProvideSupport.com
  2. Turn on sounds on "new visitor" event
  3. Each time a new visitor come in, watch it's behavior. With practice, you will identify fake users pretty fast. They also have something in common: they all come from few similar websites.
  4. Exclude those websites in your AdWords account

Method 2 (Semi-Automatic)

If you want to automatize the process, install some server-side tracking tool that will record referrer from Google and IP Address (it's in the Referrer variable). Check which IP Address subscribed to your newsletter, website, software or anything. Check which referrer generated the lowest conversion rate (I never saw any generating more than 0%).

I prefer to use the

Take this seriously, for each $1000 I invested, almost $950 was junk.

share|improve this answer
How could Facebook suffer from the same problem? They don't host their ads on third-party sites (unlike AdSense). I'm very surprised that you report such high ratio of fake users. Is there something special about your ads, like very high CPC? – Alain Raynaud Jan 3 '11 at 1:49
@Alain: I didn't found any information on Facebook yet. I'll check with the person that maintain campaigns here and hopefully come with a solution too. – user3997 Jan 3 '11 at 6:48
2  
@Pierre: You're right, and so far, you should accept your own answer on the question. What you are describing is called "click fraud", en.wikipedia.org/wiki/Click_fraud . Preventing it is hard. The most common solution seems to be to narrowly target your ads on geography (countries), and to monitor log files and complain to Google when fraud happens. If you find better solutions, I'm sure many here would like to know. :-) – Jesper Mortensen♦ Jan 3 '11 at 14:37
This is the biggest time waste of time since Angry Birds. Believe me, there are more productive things you can do with your time that watch each visitor's IP address. If you find the ads that worthless, just drop Google Content Network and stick with search-engine only ads. – Alex Papadimoulis Jan 3 '11 at 14:39
@Jesper: thanks for the link! I did some research for Facebook, and apparently they mask the referrer page, so there is no way to block anything. I think blocking feature is not yet implemented. So we must evaluate total cost in comparison with total gains to see if there is still value in Facebook scheme (I'm currently evaluating it). – user3997 Jan 3 '11 at 14:39
show 4 more comments
up vote 2 down vote

First and foremost, what you are paying for is visitors, not leads. This is a critical distinction.

Visitors are the people who walk into your store. How many times have you walked into a store, looked around real quick, and then walked out. Maybe it's the store's fault (they words "crackers with your purchase" were not visible between FREE and BEER). Or maybe it's your fault (you wanted to window shop).

Leads are identifiable visitors who've actually expressed interest. Maybe they tried on some clothes. Maybe they talked to a salesperson. Perhaps they grabbed a brochure. Only some leads turn into sales... again, it could be their fault (no money) or your fault (poor sales ability).

That said, you are asking the wrong question. Having done a whole lot of advertising on behalf of software companies, I can say with 99% certainty that there are no robots who are clicking around your site from ads.

But let's say there are. So what? In the end It's not about ensuring all visitors behave a certain way, but a reasonable percentage. Determining what the % is requires crunching a whole lot of variables like targeting, copy, goals, design, etc.

In the end, the question you should be asking is, how can I get more visitors to do this... And to begin answering that, you need to identify a whole lot of those variables.

share|improve this answer
1  
No, the question was how to reduce the amount of fake visitors. You just gave me your definition of lead and tell me your opinion on fake visitors promoting your own company at the same time. A comment would have been more appropriated. – user3997 Jan 2 '11 at 22:05
Just got the answer, check this out, this may be useful for your business. – user3997 Jan 2 '11 at 23:27
up vote 1 down vote

There are hot areas on web pages: near scroll bar, near to tabs bar, near to embedded flash games, etc. Users can click on ads in these areas accidentally. Page will be valid for Google, but will generate suspicious traffic anyway.

I recommend to visit referred pages and to look at position of banners. If they are in hot areas, then you should ban them, because they will generate lot of accidental visitors.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.