Hello everyone. Created this account to learn more about ad arb. I have something I would value your opinions on -
Background
I am about to enter into a partnership where I pay this guy a fee, and he'll sell me a site doing Ad Arb. He'll manage it for about a year and a half and I'll get an inside look into the business.
He has set up the site and is driving traffic to one landing page (single page, not a slideshow). It is profitable so far. We haven't closed on the deal yet. I've paid the deposit and now have access to the site's Clicky analytics.
What I observed:
I noticed an interesting pattern, that seems a bit fishy to me. It's best explained using this video:
https://www.screencast.com/t/LE5NG5gDZma0
Other things that are making me suspicious are:
It seems fishy to me too. The untracked traffic has way higher ad click rate and the consistency is weird too. With a natural loss of the parameters (which is normal to some extent) the untracked visits should go up aswell when the tracked ones grow.
It's hard to say something with 100% certainty, but it does look like some BOT or super high quality clicks being sent from some other source to improve/inflate the overall numbers.
The conversion rate actually looks a bit weird.
The no parameter traffic seems to have a crazy high conversion rate in terms of ad clicks as opposed to the traffic with parameters that looks to have a pretty normal-ish conversion rate.
Thanks for looking into this @matuloo. Yeah I'm trying to hash this out with the guy now. He says the loss of parameters was a tech issue. And doesn't know what to make of the data. He has installed Google Analytics, and we hope that it will provide some illumination.
Hey @jack_l,
Thanks for roping the others in to look at this. Really flattered you see me as someone who knows a ton about monetization, but truth is I only know how to flick a switch to turn ads on in Ezoic. The rep handles the rest haha. So I have a lot to learn from you all.
Ironic thing is, I did find this guy on Flippa. But I didn't want to shell out what he was asking for on the listed site. So I contacted him separately and worked out this deal. He seems legit - 100% positive feedback on Flippa, member since ages ago, $30k+ total transactions. He's Google-able: Got a LinkedIn, Twitter, Resume is online, not a bad review about him... the works.
So I don't know what to make of all this... Maybe I'm just naive about bot traffic levels?
Hey @platinum,
Thanks for sharing your thoughts.
Yeah I mean that it's a single page, no pagination. Which (from what I've heard) is different from how most people do ad arb. I just find it hard to believe that with a single page, and ads at the bottom of the content, that conversions would be so high.
What's a typical conversion rate that you'd expect?
On Clicky, I can only see that nearly 100% of the traffic is categorized as 'direct'. Speaking with the guy, there is no organic traffic right now. It's a brand new site and we're only running paid traffic.
The content that we're running traffic to is brand new too. So I don't think it would be getting its own organic traffic. Is that what you're alluding to when you mentioned running a small test with another article?
His explanation for why the campaign was profitable right out the gate was: he has had experience in this particular niche (sports). So he knows what traffic sources (site ids?) will do well with sports content. And he can pretty much put any sports content in front of this audience and he knows it'll do well. What do you think of that?
Can you get access to the server logs? You could look at patterns in user agent / ip / asn.
If the bot is really sophisticated, it won't be enough, but it is a good start.
Hey @jeremie,
Thanks for the suggestion. I have just requested for those. The site is hosted on his server so I don't have direct access.
Will update if anything.
Hey @jeremie,
Your post pointed me in the right direction.
He was not able to provide access logs, but the Clicky analytics software did capture IP addresses. I ran an analysis on those and found that almost all the outbound clicks were coming off of 14 IP addresses, when there are unique 26,000+ IP addresses. The vast majority of pageviews have no outbound clicks at all.
I have shared my findings with the other guy and we'll see what he says. However, I foresee that I will not proceed with the deal even though I'm $1.4k in the hole.
What do you guys think? Is bot traffic something you look at closely? Especially for Ad Arb where you're not paid on CPA basis.
@jeremie, here's the thing. He actually tried to get them. But after trying, he said that, "...server logs are split across eight servers, so that's unpractical for us to stitch them together." What do you think?
If I took all the duplicate IPs as bot traffic. There'd be virtually no monthly profit.
Anyway, he has gotten back to me about the duplicate IPs:
"Yeah, so our best take on the IP issue is basically they are masked via the adsupply pop up windows (this is the ad we're buying). We've run a program a few time that pulls the actual IPs just to check (we noticed the same issue when first testing this out, obviously) and the IPs we saw were different. This was further confirmed over time as the advertising partners haven't raised any red flags with the traffic, where they definitely would have -- obliviously -- if the same 14 IPs were clicking the ads. An ad account would be shutdown in a few days if something like that was running, so, considering these adblade accounts are running 1+ years with no issues, we're confident in this analysis."
What do you think? Does that make sense? It doesn't make sense to me because, let's say the ad supply is masking IPs. How would they know beforehand who would convert? Why is it that almost 100% of conversions have their IPs masked?
However, what he said about the Ad account not getting banned does make sense. Surely if it were so simple to id bot traffic they would have shut us down by now? I have access to the Adblade account and it's still running (it has been about 14 days).
If that's not practical, he doesn't need to stitch them together. He can just send separate files... Even from 2 servers that would be enough.
26000 IP for non converting / 14 for converting. There is no IP masking, just BS from him.
Not all traffic sources are controlling all their traffic. His bot might be passing through the control from that specific network.
I buy from time to time web assets, and when someone is not able to provide the data, i just pass.
While I don't know squat about the technical side, I do know to run when you get too emotionally invested in a business decision.
Right now it appears you really want to buy this site and are trying to justify to yourself why it is a good decision. If you can't understand why, then it is probably best to pass, whether it is a good purchase or not.
I'm not on the same planet as @jeremie but this:
I'd pass, this smells. I understand you'd love to buy a solid business, but this doesnt look like one, there are too many red flags here.