Home > General > Affiliate Marketing Forum

---

A + b + c =$0 (17)

---

For the last couple of weeks I've been playing around with SiteScout.

I started a campaign in a rebill niche, on two sites that I thought fit the demographic reasonably well.

I started getting a pretty good CTR (around 0.6%) and a good landing page CTR to the offer, as well (40-50%). All awesome, right? Except I didn't get a single conversion. I spent around $200 testing (at the CTRs I was getting this was hundreds of clicks to the offer) the offer on two different networks, and the offer was a well-known converter.

Oh well.

Then I did a bit more research into the demographics, and realized I might have had the age off by around 10 years, and decided to give it another go with an offer I figured might suit this demographic a bit better.

At first, it looked awesome. I've got a 1.5% CTR, which is pretty awesome, and a 60-70% CTR from the landing page to the offer. In the past few hours I've sent literally hundreds of clicks to the offer, and again zero conversions on an offer I know converts because I'm running it elsewhere. At least it's only cost me like $7 so far, we'll see if I've got a conversion by the time I've spent $100.

I'm mostly just complaining about how I can get every variable to work except the most important one, but if anyone has advice I'd be happy to hear it!

---

I'm having similar issues over at Adbuyer, except I've sent about 6k clicks to my lander which has converted with other traffic sources, about 1k to the actual offer and .. well, conversions are dismal (1 or 2 I THINK came from that source)


This is for an international CPS offer .. I've ported the campaign to SiteScout to see if the traffic quality is any different for news/ref sites but my CTR is low, CPC is around $1.50 and 0 leads so far with that also.

MediaBuying is a whole new ball game, I think working backwards might be a good idea..

Pick a few high volume sites.. get demographic data... find a broad offer that might appeal the demo.. throw up some tests and see what happens..

I'm going to try this approach with my next tests

---

Check your Sitescout stats... if you get your clicks from a site with a very high CTR which doesn't have a good track of record (check Alexa), you might be sent bot traffic.
I had this for some of my campaigns where single targets constantly had 1%+ CTR but it didn't back out after all. Did you ask your rep at Sitescout about the issue?

---

Yeah that was my issue with my Penny Auction (follow along) campaign. I had ads running that had >2% CTR, landers with >90% CTR, sent thousands of clicks with zero conversions, to several offers, over several networks.

Then I switched tactic and did as Andy is suggesting, picked the traffic first, figured out what they wanted, and gave it to them (market first). Profitable off the bat.

Pretty sure some sites there are sending bot traffic though. There are certain targets where no matter what you put up as an ad, you get thousands of clicks and nothing backs out.

---

^^^^^ I had the same scenario with AdOn Network. A shit ton on clicks that had my tracker sweating like a fat lass. Zero conversions - bots anyone?

---

Setting up bot traps on my LPs is the first thing I do when testing a new traffic source.

---

What script(s) are you using for this @leadpiss?

---

@ppchound: You could use a conversations tracking tool if these bots click on links to simulate real visitors (what they seem to do). Just place a hidden link on your landing page and define the linked page as conversation page in your conversations tracking tool.

If someone visits this page it has to be a bot because real visitors can't see the link. Now you could look up from where the bot came and remove this website.

By the way, use a dot or a white 1x1 pixel image as link text or link image in case a visitor uses a too old browser that doesn't supports the technique you use to hide the link.

Here's an example for a hidden link: <a href="http://xy.com/bot-trap.html" style="visibility:hidden">.</a>

---

Yeah, the second campaign was definitely chosen based on the demographics of the two sites (I say two sites, but one of them gets 95% of the traffic so it's really only one site) and then I chose the offer around that. I think you're right Andy and that's the way to go as well.

obolus, thanks for that tip. I'm definitely going to be implementing that!

Unless bots have gotten to be a lot more sophisticated, my P202 stats (I don't usually use P202, but I'm using this same LP elsewhere so still getting the info) don't really look all that suspicious:

---

I've done this, and continue to get hits to my LP at the same rate as before. Time to hit up my rep about the fact that I'm apparently getting a ton of bot traffic...

Thanks for the info about doing this obolus!

---

After thinking about it again, I realized that a conversation tracking tool is not the best solution. The problem is, most of these tools need JavaScript and cookies enabled to work. So if I disable one of the two I'm invisible for such a trap. I think most bots don't have JavaScript enabled.

A better solution is this:

You need:

* a hidden link to a bot trap page
* a web stats tool that runs on the server (for example a tool that analyses the log file of your webserver)

What you do is this:

1.) Look at the IP addresses that request the bot trap page
2.) Search the first request that came from such a IP
3.) Now either look at the Referer header of this request (the page from where the bot came) or -- because the client provides this header and it's therefore easy to spoof -- look at the requested URL. Of course you would need to add something to the URL to be able to identifier the traffic source

Let me know if something is unclear.

@Hannah: Basically, you stats say nothing (nothing good and nothing bad).

It's very easy to spoof the user agent (the browser and OS type). The client (browser/bot) sends this information to the server.

It's also easy to obscure the IP address. For example with a proxy server. Another easy way would be the following: If I reconnect to the internet I get a new IP. I pay around $10 for my mobile internet flat-rate. Reconnecting takes probably around 10 seconds so if I invest $1000 I have 100 new IPs every 10 seconds (of course, it's possible to get the same IP assigned, but the IP pool of an internet provider is very big so it's not very likely).

What I want to say is: It's not possible to look at your stats and see whether a visitor is a bot or not. You only find the most stupid people this way.

What I would do is to check things like the following:

* How many people have JavaScript enabled (low percentage means your visitors are most likely bots)

* How long are visitors on your website or before they click on a link? Is it possible to read the text on your landing page this fast?

* Do visitors request images and other external resources (CSS and JavaScript files)

Please note, that it's rather easy to program a bot that's not detectable with the above. Basically a good bot is not to differentiate from a browser, because a good bot IS a browser. And a good bot uses the same IP ranges as regular people (see "mobile internet flat-rate" above). With the above you can only catch stupid and lazy people. I think the above is still useful, because many people who do such crap ARE probably stupid and lazy. Otherwise they would do something else -- but that's only my opinion.

I'm not sure what you mean. Such a trap shows you traffic sources that send bot traffic so you can remove them. It doesn't stop bot traffic.

---

I had similar experience last week.
There were some new sites on SS with huge traffic so I tried few sites, I was getting great ctr(one of them was 8%) on my banners but after I sent over thousand clicks to few different offers from few sites there, I just had to give up.
The sites I tried weren't getting huge traffic up until Nov last year, they were suddenly getting heavy traffic last couple of months.
I am sure that all those sites are sending something they are not suppose to...

However, I don't understand about this thread why we have to find out our own if we are getting bot traffic or not.
We are the customers and SS are selling us traffic, isn't it their job to sell us real traffic?
We shouldn't have to go through all of this if they are selling real traffic.
Just my 2 cents

---

What I mean is that I set all of my links on my landing page to hidden. My campaign had hit its budget cap so wasn't running at this point. I double checked and sure enough, every single link on the page was now hidden. So I added another $10 to the campaign daily budget and continued to get the exact same LP CTR, which I take to mean it must have been bot traffic, since normal visitors wouldn't have been able to see any of the links to click on. It definitely hasn't stopped the bot traffic at all, but I'm pretty sure now that something is definitely up.

I've emailed my rep about it, but haven't gotten a reply as of yet.

---

Ah, okay. I thought "hits to my LP" means clicks from the ad to the landing page with the hidden links. Anyhow, great that I could help. Maybe I should write a tutorial and post it as it's own thread so other people can use it too.

---

Yeah, you're right I did write that really badly. But yeah, I definitely would say start a thread. The info definitely helped me, can probably help others too.

---

Hi Hannah,

Did you ever get a reply from your rep? I'm interested in finding out how things turned out.

---

It turned out that for whatever reason my emails to Steve weren't going through at all, which is why I never got a reply. We only found out that this was the problem like 2 days ago (I asked Dr. SiteScout about it and he confirmed that my emails definitely weren't making their way into Steve's inbox) and I'd forgotten I meant to ask about this.

---

Home > General > Affiliate Marketing Forum