$1000 Net a Day (19)
Quick intro: I run a low 6-figure business. We’ve mostly done just SEM traditionally, with only the occasional dabble in small media buys here and there. We’ve been doing business online seriously since 2011.
My main focus at the moment is building our new flagship product, which’ll be a really dense, high polish, 40+ hour video course that rebills over a 10-month period. However, I *really* want to learn media buying, so that when our course is out, we aren’t just relying on our own list and readers + whatever we can scrounge up from affiliates. Would be nice to earn some extra coin in the meantime too, either pushing our own programs or someone else’s.
This follow-along is going to be my process of going from dumb newbie to… wherever I end up. Hopefully somewhere not too terrible.
Starting objective is going to be $1000 net per day.
We can probably throw about $100 / day into testing without losing too much sleep over it, however I’m cutting ineffective campaigns long before that point (and I’ve been doing plenty of cutting) because if it’s clearly not working, no point throwing away money on it.
Onwards!
Chase
My first attempts, about a month back, were on SiteScout. Here’s what I tried:
- Ads for advertorials that then directed prospects toward VSLs from two of our niche’s lead offers (both with $100 payouts)
- Ads that took prospects directly to an industry-leading VSL
- Ads that took prospects to a quiz on our site
Spent a few hundred bucks on these tests, received close to 20,000 impressions, zero conversions. I tested a suite of different contextual groups (separate campaigns for each one), and tested price points everywhere from 10 cents CPM to $1 CPM, didn’t seem to make a difference. I noticed a lot of the larger mainstream sites have $2 and $3 CPMs, but I didn’t want to target them.
Next efforts have been on PopAds, after I saw a recent comment on STM (think it was from Caurmen or CMDeal?) that PopAds is really hot right now and much easier to get working than SiteScout.
Here, I tried:
- Popunders direct to an industry leader’s VSL ($100 payout)
- Popunders to our email signup form on our own site
CPM was around $4 for the VSL, and between $1 to $5 CPM for the email signup, depending on whether we were targeting sports, file sharing, or adult traffic.
We sent around 8000 prospects to the VSL, no conversions, so I pulled the plug.
Sent 6600 prospects to the email signup form, no signups, pulled the plug.
I know our niche is fine, because the big players are all still doing large volumes to their VSLs and landing pages, but I might be trying to reinvent the wheel too much by trying traffic sources they’re not on. Mainly just trying not to step on anybody’s toes, but maybe there’s a reason they aren’t using these traffic sources.
Anyway, back to formula!
Chase
Deciding what I should try next, I've got a couple of options:
- Try a different offer within our vertical
- Try a different traffic source
- Test out a wider variety of landing pages
- Test out offers from completely different (less saturated?) verticals - that'd be the biggest leap / most time intensive
- Manually research and handpick placements, rather than relying on categories provided by traffic sources
I'm on a tight schedule for the next week or so so can't do a huge amount of running researching, so next campaign is going to be another easy-to-set-up SiteScout display ad campaign, this time directing to our email optin form rather than to our quiz or an industry leader's VSL.
I suspect a large amount of bot traffic coming from SiteScout - some of the placements are almost certainly relying on bots (low traffic websites with high numbers of impressions / high clickthroughs). I've already blacklisted lots of those on previous campaigns, and this one will inherit blacklist settings from an older campaign.
To get better data, I'm test driving Caurmen's guide on finding and destroying bots, here:
How To SEEK AND DESTROY Bot Traffic To Your Campaigns
Can't use his method #1 since we're using display ads and not direct landing page loads. So using Caurmen's method #2, with Bit.Ly. Set this up, tested it out, and it tracks my click in Bit.Ly, so it's working perfectly.
I'm doing tracking with
- Email signup confirmation page (payout = $0)
- Content subscription completed order page (payout = $111)
- Product sale page (various products use the same thank you page, so set average payout = $80 for now until)
This way, I'll be collecting conversion data for both a.) newsletter signups, and b.) actual sales, subscriptions, and revenue.
To start the campaign, I created 29 300x250 sidebar ads, everything identical but the headline. If there aren't too much bots and the traffic is legit, we'll get data on which headlines are performing best. (if lots of bot traffic, this'll just be useless data)
Starting out with a $0.10 per CPM bid on SiteScout, which means $0.20 / CPM after SiteScout's fee. Let's see how this one goes.
Chase
Verdict on SiteScout's contextual targeting: mostly bots (at least, the contextual filters I tested). Only ran to 2,700 clicks (~34,000 impressions), then shut this experiment off.
96.5% of all traffic did not load the transparent 1 pixel GIF, which means 96.5% of clickthroughs were bots. That's pretty terrible...
Of those 2700 clicks, 95 did load the pixel, but not a single one opted into our email list to get the freebie they clicked on the ad for. I'm not sure what that means, since opt-in squeeze pages like this typically convert in the 40% to 60% range, at least that's what I've seen with direct buys and what I've seen friends do with Facebook, etc. traffic.
It's odd that 95 people would click on an ad saying, "Get X report", then all 95 of 'em wouldn't bother to plug their email in to get X report on the following screen. Some kind of bot that loads the page? Or site owners clicking on ads manually?
An email opt-in is pretty low friction, so I'm guessing (hoping?) the problem is the traffic source. Next stop will be trying the same offer (email opt-in), different traffic source... likely one I already know works for our niche.
Chase
It is probably worth starting with the basics, like what is the course about, what is your target market, and what motivation drives the purchase? That will drive a lot of your strategic decisions on the media buying.
Thanks for the reply, cmdeal. That'd make sense; maybe targeting just wasn't tight enough on that campaign.
My demographic targeting may be too broad. I'm going in saying, "This group fits our demographics, so this ought to work," but maybe it needs to be more like, "This group fits our demographics AND is almost certainly looking for what we're offering, so this ought to work."
Going to check out some other traffic sources and see if I can find some that offer tighter demographic targeting than SiteScout or PopAds do. Or it might be time to use the direct buy list I've put together and see if SiteScout can negotiate placements with some of these sites.
Chase
Updates: tried some direct buys last week from some very closely related sites to ours and got snubbed:
I've seen that before with other sites; we're in a competitive and reputation-sensitive vertical, and nobody really likes advertising for anybody else.
This week's goal will be to kick off 5 new campaigns, one per weekday.
I started with Mr. Green's post here:
Cool Media Buying on a $500 [Gigantic Guide]
... since we've been unsuccessful with SiteScout so far, I opted to try AdBuyer.com. Only to find they no longer exist... so this time I went with AdBlade.
Today's (Monday's) was another email optin campaign. I've seen other guys in our vertical doing well with AdBlade, so I'm assuming there's some opportunity there. We'll start with the email optin to test for bot traffic, and if it's low, even if email doesn't convert, we'll try an affiliate offer after that (since all our VSLs suck and our new course won't come online until spring... and I'm not going to write a new VSL for one of our existing offers with limited lifetime customer value).
AdBlade offers both blog traffic and premium traffic, apparently, but I couldn't seem to figure out how to get a campaign set up for the less-pricey non-premium traffic.
No matter though; we'll wait for approval, and then run the campaign, pay a little extra, and if it works on premium I'll sit down and figure out how to set up the non-premium network.
One thing I'm realizing is how few sites let you split test ads like SiteScout does. That's a major disadvantage for figuring out the most profitable headlines, body copy, images, and design. If we're able to get a campaign profitable on one of these non-split-testing sites like AdBlade, we'll probably have to move it to another platform to figure out what angles actually perform best.
Chase
Missed a day due to getting my clock cleaned by a head cold I've been battling, so set up two campaigns today.
To save time, both are SiteScout, using Audience targeting (previously, I've only tried Contextual, which either provided negligible traffic to our offers or heavy bot traffic). For each of these campaigns, I'm testing 29 different display ads (everything identical but the headline).
To run these campaigns, I've set up 5 identical landing pages with slightly different URLs, each tracking a different transparent pixel via Bit.Ly so I can determine which (if any) get legitimate non-bot traffic.
Our AdBlade ad is still awaiting approval; if and when that one gets approval, I'll fund it and we'll be off. SiteScout's already funded, so both of those campaigns will begin as soon as they're approved - I don't see any issues there since we've had the same offer approved on SiteScout before.
Also, I'm all booked up for Affiliate World Asia; not a huge fan of Bangkok, especially due to the military coup and police shakedowns of foreigners they've been dealing with of late, but I've heard good things about some of the speakers and I'm sure this event will be epic. Hopefully I can get enough test runs in of various media buys before the conference that I get the most out of the various presentations.
Chase
Hey there. It is a bit hard to give any really actionable advice without an understanding of what exactly you are seeking to promote. I would probably suggest that you consider paying a link to your product if you feel comfortable with that. I think you will get a lot more and better quality, more specific feedback that way.
Fair point, cmdeal! I saw guys on here complaining that STM lurkers had sniffed their campaigns and were copying them, but these were all guys doing follow alongs where they were successful and making plenty of money, and that's probably a ways off for me yet.
Here's the landing page we're testing with right now, in the men's dating advice / pickup vertical:
We've also tried several of the top offers in the niche. So far, haven't been able to get a single conversion, either to our email list or a top offer.
In light of our latest campaigns, I have some thoughts about why.
UPDATES
I missed last week's "5 days, 5 campaigns" goal after being laid out in bed much of last week with my worst cold in years. I'm completely back now, but not ready to dive into another goal just yet... we had a site upgrade on Monday and a lot of cleanup / catching up on technical and other stuff I need to handle first.
AdBlade
Still awaiting approval. I've written to ask them if 10-day approval periods are normal for them or if we've slipped through the cracks somehow.
SiteScout
Here's the interesting one. Neither of the audiences I picked were large enough to scale (even if I doubled our bid, and I already started it out high), but one of them received around 4500 impressions per day. So we could at least get some data from it.
This one had fewer bots... more like half and half. Half legitimate traffic, half bot traffic.
But of the legitimate traffic that clicked through, again, none of them signed up for ours newseltter.
I checked out the sites showing up in the audience we'd selected, "Urban Males". What I discovered:
Of the top 5 sites, 2 were about crocheting afghans.
One was a cooking site.
Another was just a garbage reseller site.
Unless urban males have taken up crocheting afghans and cooking as big new hobbies (hipsters?), the targeting on this audience is way off.
So, at least now we know that SiteScout contextual and audience targeting is a.) bot-heavy, and b.) not well targeted at all, so not to use these.
I'll try networks we know work for our vertical next.
Chase
Okay, spent a few hours today doing the hard work of campaign planning for the week. I've planned out 5 campaigns on 5 different traffic sources (all new) for this Monday through Friday.
I'll be using the same email opt-in for all of these, if only to test a.) if we can get opt-ins and b.) how much of the traffic we get is bot traffic. However, I've modified the opt-in form to make it more mainstream friendly, removing the "And Into Bed" portion.
I selected the traffic sources we'll try out this week from g4offers_mark's post on traffic sources. First one is LeadImpact, which requires a $1000 minimum deposit (blech). Here's hoping we can get something going, and that the traffic is indeed quality...
Our keyword selection should be sound, since I just pulled those straight out of Google Analytics, 3 years' worth of keywords from 14 million organic search visitors. Just picked the top 1000, edited out repeats of our URLs or searches for site-specific content, and pasted them in.
I wrote AdBlade and Allison wrote me back to let me know they do not review campaigns until your account is funded... so I funded the account. Still awaiting approval, however. Also took the time to defund PopAds, since I couldn't get anything working there (not our offers or known high performers in our vertical), after trying all the targeting options that seemed like they might fit.
Chase
I am not involved in this vertical at all, but I think johna5150 is. I believe he focuses on email and list building. As for low quality traffic sources like Site Scout you will need to spend a lot of time blacklisting sites. Lead impact is also pretty pretty poor.
Thanks for the heads up on SiteScout/LeadImpact and the note on johna, CMDeal. Looking like these are pretty crummy traffic sources, yeah. And johna mentions being in this vertical in some of his posts from what I can see.
SiteScout is flooded with bots. LeadImpact has sent over 5 impressions in the 3 days it's been active... at that rate, I'll need just shy of 103 years to use up the $1000 I funded that account with. Even with advances in medical technology and clean healthy living I may not outlive that campaign. Maybe g4offers_mark's traffic sources post is a little out-of-date?
Guess I'll avoid the minimum-$1000+ sources until I have a working offer running.
Seems like figuring out which traffic sources are good or not good takes a lot of trial and error (mostly error).
I'm reading through johna5150's posts right now. Will ping him to connect / bug him with a few questions once I finish browsing what he's shared already.
EDIT: John's got some wonderful posts (like this one and this one). Much appreciated for that terrific recommendation, CMDeal.
Chase
Quick update:
Finally got our first conversion in December! 1 email signup. Worth probably about a buck or two on our list. Cost $2000 to find him, everything we've spent on media buying considered (across various campaigns for email signups, affiliate product offers, our own in-house offers, etc.).
So, safe to say nothing we've tried has panned out. However, I've learned a lot, and the money and time's definitely been well-spent.
I came away from Affiliate World Asia with a wealth of ideas and have started implementing some of them.
Since AWA, I've been knuckled down spending disposable time building our massive continuity course so we'll have everything in place to be able to film in March/April. Media buying's moved to the sideline momentarily.
We're designing the course for cold traffic and will test relentlessly until we've got it profitable there once it's ready, so I'll be circling back to this project once we get nearer to our launch date.
Chase
So, was the ultimate strategy here to test related products, or products within the same vertical such that you were advertising to the demographic of your eventual product?
I think a lot of the success here will come down to the pairing of ads, traffic and offers.
With the offers, did you have any guidance for which to advertise? Picking a few "industry leading" offers sounds great in principle, but in affiliate marketing - at least when starting out - you really want insider info of some sort to reduce time wasted on products that don't convert that well (at least vs other offers you could be promoting).
With the traffic... SiteScout is a beast. The platform is awesome and all, but this kind of traffic is very difficult to tackle vs say Facebook, Adwords etc. just because of how many placements, ad sizes, etc. there are and the generally poor performance.
Media buying is a pretty generic term but generally gets associated with desktop/mobile banner traffic. Banners are... the past, really. I would focus on more interesting advertising media that works well for capturing prospects attention, e.g. Facebook, Adwords, native desktop traffic, native mobile, solo ads, etc.
Originally Posted by chase_
You are obviously doing something massively wrong.
I really think you would be better off transparently posting EVERYTHING you are doing, so that folks on STM can give you some actionable advice.
I would not be afraid of anyone trying to "steal" your campaigns at this stage. Really, no one wants to steal a campaign that generates 1 email signup for $2000.
Ouch $2000 I would of stopped at $200. Like cmdeal said you must of really been targeting the wrong crowd. I hope you keep getting more success but at a lower cost!
The purpose of a follow along is to be transparent so people can comment on what you do. You don't have to hide anything because you have nothing. The reason many follow along die after a while is because.... they got profitable. THEN they start to hide. Not hide nothing right from the start.
But based on your lander, you seem to want to do dating stuff. And dating works best on -> Adult. So you might want to consider using adult sources instead of the sources you currently are using.
Another thing you can start with is spying. Either with a spy tool or manually spying. So at least you can get the best creatives(if they can pay for first impressions it should mean that creative is working right).
Spy for your creatives and match your traffic to the niche you want to do. That would be the advice anyone will give you with no information to go on.
APM Jeremy
Originally Posted by acepowermarketing
Home > Display / Media Buying >