Journey to Reaching $100 profit a day with Native Ads

[Flaxy]

Previously I was on a journey with popunders (here). At the same time I was also testing out native ads for a while(2 weeks or so) and while I wanted to focus on popunders(hence the journey), it seems like native ads is what has struck gold first.

So I will stick to what the data is telling me and continue focusing my efforts on Native Ads instead as I am already 402% in profits today. I know this could change but at the time of writing I am ecstatic!

Screenshot attached in the post.

Preparation yesterday
Visits: 599
Clicks: 144
Conversions: 3
Revenue: $76
Cost: $78
Profit: -$2
ROI: -2.63%

Yesterday
Visits: 614
Clicks: 29
Conversions: 0
Revenue: $0
Cost: $7.30
Profit: -$7.30
ROI: -100%


Today
Visits: 43
Clicks: 8
Conversions: 2
Revenue: $51
Cost: $13.16
Profit: $40.84
ROI: 402.07%

My work method is: Ads -> Prelander -> Offer lander

I have optimized the campaign by removing low performing ad creatives based on ctr%.

Same goes for website ID that provides low ctr%.

PS. Does anyone know what utmsandbox means in Bemob? Seems like I got 2 conversions there.

 

[Graybeard]

Not bad at all

1. Can you do it consistently? <<<every day
2. Can you scale the success to $500 or $1,000 net per week?

I think you need to know that ACTUAL domains that the ads are on so you can see that domains ratio of new:: old [returning] visitors.

Why? Because you want to have new people constantly viewing your creative for it to continue to work (well[CTR]).

This is why buying traffic only known by zone ID is of very limited usefulness.

 

[Flaxy]

Not bad at all

1. Can you do it consistently? <<<every day
2. Can you scale the success to $500 or $1,000 net per week?

I think you need to know that ACTUAL domains that the ads are on so you can see that domains ratio of new:: old [returning] visitors.

Why? Because you want to have new people constantly viewing you creative for it to continue to work (well[CTR]).

This is why buying traffic only known by zone ID is of very limited usefulness.

Thanks! The roi% is already dropping as we speak to about 140%.

The [returning] tag only needs to be edited in my tracker right? Or do I need to adjust it on my traffic network as well? I think that requires a new approval from the network..

Would you duplicate the campaign in that case?

1. I sure hope so.. I would be happy if I could reach $50 a day in profits.
2. $500-1000 a week would be great.. With my limited experiences (about 2 weeks) it will take while I think. First $500 per month.

 

[Flaxy]

no the returning tag in your tracker means: that the IP/user has clicked your ad again or has been seen before by the tracker.

If you put Google-Analytics on your landing pages the user will be cookie tracked.regardless of the IP -- this is browser dependent (stored data -- cookies) However, cookies can be deleted by the user or accepted 'session only' in private/incognito modes. Cookie are maybe 50% to 70% accurate for these reasons.

SEO warning: the high bounce rate of most landing pages may ruin a domains SEO standing.

What I meant by return rates is: the stats of the domains that your display ads appear on.

Thank you for explaining it, in short you think it is important to track with google analytics as well?

 

[Graybeard]

'utmsandbox'

sandbox in a software definition usually means a isolated (or seggregated) testing environment

 

[Flaxy]

I get it, do you use Google analytics yourself?

 

[Graybeard]

no the returning tag in your tracker means: that the IP/user has clicked your ad again or has been seen before by the tracker.

If you put Google-Analytics on your landing pages the user will be cookie tracked.regardless of the IP -- this is browser dependent (stored data -- cookies) However, cookies can be deleted by the user or accepted 'session only' in private/incognito modes. Cookie are maybe 50% to 70% accurate for these reasons.

SEO warning: the high bounce rate of most landing pages may ruin a domains SEO standing.

What I meant by return rates is: the stats of the domains that your display ads appear on.

 

[Flaxy]

Usually. If you create a NOINDEX domain you can use it if you don't care about any SEO factors. Like I said above^^

I used G-A commectial self hosted Urchin -- that is the program that G-A was based upon. It cost $10,000 and Google screwed us and discontinued development and support of the Urchin self hosted software. For the reason it was self-hosted elGoog could not get at any of our data --do the math!

I don't care about SEO because its mostly paid traffic anyway..

That really sucks that they screwed you over... but its good to have all the data yourself now.

As for the journey:

End of Day 2
Visits: 161
Clicks: 42
Conversions: 2
Revenue: $51
Cost: $41.81
Profit: $9.19
ROI: 21.97%

Seems more in line with what I would expect the day to be.. but because its a cod offer I'm not sure if the conversions get updated once a day on their side or how it works...

Yesterday I had no conversions as well, but I saw them at the start of today so lets see tomorrow..

 

[Flaxy]

'utmsandbox'

sandbox in a software definition usually means a isolated (or seggregated) testing environment

1. Are you doing a 302 for the click out of your landing page's CTA (call to action button/link)?
   --and logging that 302?
   --what % are following the funnel path to the offer?
   --have you defined your funnel (conversion path)?
   
2. Is the offer working or are they shaving you?
   Is the affiliate network with the offer shaving you?
3. Is the traffic inbound from the ads the same patterns?
   Or, is the traffic lesser quality?
4. It can be simply now an old ad that the same people are seeing over and over again. <<<MOST LIKELY ANSWER

A 'good website' will have 20% to 40% new traffic daily.
Best test is to add some new websites for your ads --- and see if there is a change ...

0. Is it good/a problem if I have conversions from the utmsandbox?

1. Bemob uses a 302 yes. I think its used with their click url. 160 visits -> 25% ctr to offer page

My funnel path is: Ads -> prelander -> offer page - > gets a call (cod offer). Why?

2. The offer is working as far as I know, how do I know they are shaving from me? I had 6 conversions with 5 approved and 1 rejected.

3. How do I check if its the same pattern or less quality?

4. Impact is ad blindness you mean? Because I do notice ad ctr% is going down slightly for some ads.

Thank you for your feedback, it is appreciated @Graybeard.

 

[Graybeard]

The advantage of using G-A is the depth of the user tracking ...
time on site before the click through and the bounce rates (page abandonment)
you can do a<=>b landing page tests with the GEO; Time on site; time of day;
G-A is 'free' however the price is that Google gets access to more site data that is publicly available (double edged sword)

 

[tsc]

Good thread.
Good luck.
Make sure you keep it updated - stuff like this is invaluable to folk like me who are in the process of sorting out which direction to head in

 

[cpagoose]

oh nice so inteterst

 

[Graybeard]

Usually. If you create a NOINDEX domain you can use it if you don't care about any SEO factors. Like I said above^^

I used G-A commectial self hosted Urchin -- that is the program that G-A was based upon. It cost $10,000 and Google screwed us and discontinued development and support of the Urchin self hosted software. For the reason it was self-hosted elGoog could not get at any of our data --do the math!

 

[Faelmaran]

Could you please let us all know which offer you really pushing or not ? Cause it's really matters the most I would say ? How does that sounds ? I do see from payout and number of conversions it really pays well for that matter, correct ? Thanks!

 

[Flaxy]

Good thread.
Good luck.
Make sure you keep it updated - stuff like this is invaluable to folk like me who are in the process of sorting out which direction to head in

Updating it now for yesterday!

• You need a count of the landing page loads and a count of the CTA link.
• You can fire a tracking pixel for the landing page to get a count of the raw CTR then a 302 to your tracker.
• 302CTA/raw pixel count = the 2nd step CTA or the click through to the offer's landing page
• G-A will give you the page load (in place of a tracking pixel)
  
• The user's browser will effect the count either way. Most bots will not be counted either way (image or JavaScript [G-A code])... so there is another factor to consider ...

I'm still debating over using G-A. Will update if I do.

Could you please let us all know which offer you really pushing or not ? Cause it's really matters the most I would say ? How does that sounds ? I do see from payout and number of conversions it really pays well for that matter, correct ? Thanks!

Sorry, I can't show my offer as there is a chance someone would copy it.

Update #3

Yesterday was a terrible day for me. My inexperience showed and I made a HUGE loss.

Visits: 164
Clicks: 42
Conversions: 0
Revenue: $0
Cost: $47.08
Profit: -$47.08
ROI: -100%

My biggest takeaways after discussing it with someone who I consider my mentor is that I have messed it up by opting to increase my bids on ads to gain more traffic.

I thought increasing my bids to gain more traffic would be good because I got my head in the clouds with the campaign in profits.. Funny how that goes sometimes.

For now, I have decreased my bids back to where they were when I got my first conversions and so far my cost has been reduced by 70% while my traffic only went down by 50%.

Plan for today is to create more ads and split test them. Also starting a new campaign if it gets approved by my traffic network.

 

[Graybeard]

1. Are you doing a 302 for the click out of your landing page's CTA (call to action button/link)?
   --and logging that 302?
   --what % are following the funnel path to the offer?
   --have you defined your funnel (conversion path)?
   
2. Is the offer working or are they shaving you?
   Is the affiliate network with the offer shaving you?
3. Is the traffic inbound from the ads the same patterns?
   Or, is the traffic lesser quality?
4. It can be simply now an old ad that the same people are seeing over and over again. <<<MOST LIKELY ANSWER

A 'good website' will have 20% to 40% new traffic daily.
Best test is to add some new websites for your ads --- and see if there is a change ...

 

[Flaxy]

Who are the clicks and why are they not buying?
Until you understand that you won't know the real reasons.

1. you may have had a flash in the pan because of the newness of the ad too.
2. you may be showing that ad on the same websites to a returning visitor possibly.
3. try making a second (ad creative clone) campaign with a different set of website parameters.

How do you find out why they are not buying?

1. This could be true.
2. Possible yes
3. What do you mean with this? A second campaign that has different widget ID that I target or different ads?

 

[Flaxy]

Update #4:

Yesterday the cost has been greatly decreased with the amount of clicks increasing. With this kind of cost I can run ads for 2 days and breakeven with only 1 conversion instead of needing 2 per day to breakeven.

Visits: 131(-34)
Clicks: 55(+13)
Conversions: 0
Revenue: $0
Cost: $11.81(-$35.27)
Profit: -$11.81
ROI: -100%

I have also created a 2nd campaign and target another geo with campaign 1.

Surely the next conversion is about to happen today.

 

[TheDanmit]

Update #4:

Yesterday the cost has been greatly decreased with the amount of clicks increasing. With this kind of cost I can run ads for 2 days and breakeven with only 1 conversion instead of needing 2 per day to breakeven.

Visits: 131(-34)
Clicks: 55(+13)
Conversions: 0
Revenue: $0
Cost: $11.81(-$35.27)
Profit: -$11.81
ROI: -100%

I have also created a 2nd campaign and target another geo with campaign 1.

Surely the next conversion is about to happen today.

Well, keep us updated on results of that second campaign. So far it seems that it surely has potential, did you cut back CPC on the first campaign?

 

[Graybeard]

• You need a count of the landing page loads and a count of the CTA link.
• You can fire a tracking pixel for the landing page to get a count of the raw CTR then a 302 to your tracker.
• 302CTA/raw pixel count = the 2nd step CTA or the click through to the offer's landing page
• G-A will give you the page load (in place of a tracking pixel)
  
• The user's browser will effect the count either way. Most bots will not be counted either way (image or JavaScript [G-A code])... so there is another factor to consider ...

 

[Flaxy]

Well, keep us updated on results of that second campaign. So far it seems that it surely has potential, did you cut back CPC on the first campaign?

Sure did! Attached the image with this update.

• On US traffic I can do financial profiling on valid ISP traffic by location (city, state, zipcode) (70%+- accuracy) from public records.
  
• On Global traffic I can detect much known non ISP traffic in real time
• I can heuristically detect some non-human (bot) traffic in real time and forensically detect additional bot traffic with forged headers (after the fact)

Before you ask; this is not by using some magic 'tool", free or paid for
It's done with hard earned knowledge and data accumulated by myself and using data accumulated by others who have sourced their data in the public domain.

I can buy public domain sourced data based on the information that a user gives me -- email, location address, telephone number etc.
---------------------------------------------------------------------------
The real problem is in either your traffic or your funnel -- possibly both.
You can only control your end; the offer that you hand-off the referral to basically only supplies you with binary data (yes|no) and no possible reasons for the why part -- only perhaps subjective or assumed reasons on your part.

On your own landing pages you can use beacons and statistics -- what I mean by this are image types served, form inputs, setting your own cookies and tracking them. Then by counting and studying your own redirects that are logged on your server.

Do your own logs agree with the traffic network's counts?
If they disagree what is the under count?
Is their count unreasonably inaccurate?
You tracker may show this if the traffic is recorded more than once on your landing page

1. the first hit in -- is that being recorded the tracker?
   
2. the 302 exit to the tracker (the CTR of the landing page CTA -- the click thru ->again)
   
3. and the redirect to the destination (the offer usually?)

Learn to read and use your own webserver's access logs -- don't just depend on some tracking tool.

The way I see it: people spend money on web design and traffic without any real effort to incorporate accountability. You cannot prevent most problems before they occur but you should be able to discern what has occurred -- and take actions more based on some known facts rather than blind assumptions.

Seems like experience is everything... You are doing this full time by now I assume?

Your post is a lot to take in but I try to answer what I understand and reread the rest.

My tracker differs slightly with the ad network as in I have around $220 spend and the difference between the networks is about $6. Would you say this is significant?

1. My tracker url is the campaign url on the traffic network.
2. What do you mean here?
3. And this one as well?

Currently researching access logs, what they do and how to use them.

Update #5

This weekend was a success! At least for campaign 1. I am going to stop campaign 2 now to focus on campaign 1.

Campaign #1

Visits: 289
Clicks: 72
Conversions: 1
Revenue: $25.5
Cost: $22.55
Profit: $2.95
ROI: 13.06%

Campaign #2

Visits: 156
Clicks: 14
Conversions: 0
Revenue: $0
Cost: $14.11
Profit: -$14.11
ROI: -100%

Plan: Open up widget ID and some other adverts in order to get more data before blocking things.

 

[clairedivas]

Native ads have caught my attention lately and that's why I have been experimenting with them for around 2-3 weeks now With this journey I am going to learn more about native ads and gain some knowledge from more experienced members while inspiring beginners as well.

 

[Flaxy]

Native ads have caught my attention lately and that's why I have been experimenting with them for around 2-3 weeks now With this journey I am going to learn more about native ads and gain some knowledge from more experienced members while inspiring beginners as well.

Why are you quoting me?

No, I meant

• the gross number of ads served
• if you do not host the served content you are working blind

=In other words; if you upload your ad image to a remote server (you do not control) you cannot count your ad image independently.

The image would be status 200 or 304 in YOUR server log or 403/444 possibly

Otherwise, a remote server could tell you anything and there is no verification.

If I read this correctly, they can give you say you received clicks and impressions while not true? Is that what you mean?

Update #6

Campaign #1

Visits: 910 (798 unique) <<<< Why the 100 difference??
Clicks: 173 (157 unique)
Conversions: 5 (2 confirmed, 2 on hold, 1 rejected)
Revenue: $51
Cost: $74.52
Profit: -$23.52
ROI: -31.56%

Today I have unblocked some of my widgets and ads in order to receive more data and as well target more devices which in return increased my budget spend by a LOT as you can see.

Waiting to see if I can get the 2 orders on hold to approve by tomorrow which would push me into the green today.

Anyone know why I got around 100 duplicate clicks and where I can see where they come from?

 

[Flaxy]

• I hate to say so --but padding and shaving is very common when you do not have database control, or the opportunity to audit the database and the books, in this man-in-the-middle AM industry.
• That said; some businesses deal honestly --but it's too damn easy to shave off the top or pad some amount and get away with it.

When I was a field or inside salesman years ago; we took an order and knew it existed.
The customer said yes to our ear or signed on the dotted line then we turned in the order to our employer.
If something went wrong we had the data to either talk with the customer on the phone or visit his home or place of business to confirm the status of the sale --was it delivered or not --hard evidence </case closed>

I consider everyone suspect doing business on the internet. False claims and erroneous data are not that uncommon. If things don't look right I just walk away. I walk away a lot and sometimes can miss an opportunity. It's really a subjective judgement.

I can't say why something may work or may not work based on summarized data.
---------------
$6/$220 is only 3% I think that seems reasonable. But what is that error actually?

total : unique IPs are IPs(traffic) that has requested you ad more than once in a 24hr period.
that can occur when a person sees the same as twice and clicks and possibly if he refreshes the ad?

visits to what? your prelanding (funnel) page and the clicks mean the CTR FROM that page of yours?

$74.52/157 =$0.47 that is what a potential customer cost you (that day).
$74.52/4 =$18.63 is your cost per customer acquired in the ad campaign (that day) less any other expenses [server, time, misc biz expense] assuming the 'pending' sales complete.

1. No idea where that error comes from.. Is it smart to show the same ad to the user more than once a day? If they are interested maybe the chance to convert is bigger?

2. The visits mean they clicked on the ad and go to my prelander, clicks mean they go from prelander to the offer page. Pending sales take a long time.. any idea why with cod offers? They should get completed the same day no?

Update #5

The last 3 days I have been experimenting with targeting more devices like smartphones and the results are.. not so good. Not sure if mobile just performs worse than desktop but it seems so..

I removed 1 OS version for now and if mobile won't perform again I will stop them altogether. Also reduced the amount of ads from 24 to the best performing 6 to see how that impacts ROI.

3 day stats:
Visits: 1856
Clicks: 420
Conversions: 0 (3 on hold, 3 declined)
Revenue: $0
Cost: $151.66
Profit: -$151.66
ROI: -100%

Overall very discouraging... hopefully today will be better.

 

[Graybeard]

Who are the clicks and why are they not buying?
Until you understand that you won't know the real reasons.

1. you may have had a flash in the pan because of the newness of the ad too.
2. you may be showing that ad on the same websites to a returning visitor possibly.
3. try making a second (ad creative clone) campaign with a different set of website parameters.

 

[Flaxy]

I have used 2 views in a 24/hr period. There is the argument of redundancy in advertising in general. Even intrusive ads and obnoxious ads get remembered (for better or worse) \_(‘ < ‘ )_/

Unless, you can identify some device to EXCLUDE due to some design shortcoming --what is the point?
Targeting desktop or mobile makes sense.
Devise targeting is easy because that is given in the reply headers by the browser.

If you want more unsophisticated users: Chrome and Edge are your buddies

Chrome is performing decent with a -11% roi over the last period (30 days chosen)
Firefox suprised me with a 50% roi right now at 224 visits total with a ctr of 110
Safari also doing well with a 17% roi (1 conversion)

Chrome mobile is terrible for me with a -73% roi.

Mobile in general was pretty awful the last 3 days where I started it.. Mostly android eating up my budget.

I'm starting to think weightloss should be sold on desktop mostly.

 

[Flaxy]

Mobile is lots of cheap traffic that is hard to find the right offer for that converts.
Are you trying to use an offer that is COD with a Tier 2 or 3 GEO?

I believe it is a Tier 1 GEO, maybe Tier 2 depending on your criteria.

UPDATE#6

As you might have noticed my updates are not daily anymore. This is mostly because I am learning and optimizing the campaign and having more data to update is better in my opinion.

6 day stats:
Visits: 4214
Clicks: 981
Conversions: 4 (19 on hold, 6 declined)
Revenue: $102
Cost: $354.44
Profit -71.22%

1. As you can see on the surface level a -71% ROI might seem bad but I am gaining a lot of new conversions who need to be approved. I believe that even if I get only a 50% approval rate on those 19 (I think I will get more) I will be very close to a green campaign overall. I JUST FEEL IT, I AM SO CLOSE TO A BREAKTHROUGH.

2. The number 1 big thing that changed tripled the amount of conversions I got was..

.. to turn off day parting. As a beginner promoting cod offers I was turning off my ads every single time the call center in the country would close as well. WHAT A BIG MISTAKE THAT WAS.

People come home after work and browse their phones and pc all night. Why I ever thought it was a good idea is beyond me. The people at the call center will just call them back when the next work day start. This realization alone tripled my conversions the last 3 days from 0-2 a day to around 6 a day. INSANE.

3. In the last update I thought mobile was terrible but it seems like I judged too quickly. Removing a certain OS made it a lot better. I am focusing 90% on MOBILE right now.

Also I am starting to see positive roi appear in my tracker, the data is starting to show which ads are profitable, which widgets are profitable and same for browses and OS.

4. I fixed the problem with utmsandbox which caused me not to get data from widget id so I can start to remove budget eaters there as well.

The Plan
5. At the moment I have 2 landing pages and 4 offers, I am going to focus on only the best one and try to optimize the landing page to turn this campaign GREEN.

This will be done using the heat map and recordings from Hot Jar. I have already collected some data and plan on using it today.

6. Furthermore I will keep reducing budget eating ads and widgets/browsers/os.

That's it! Let me know what you guys think and if this is helpful to follow along.

 

[tsc]

Things are starting to happen for you mate. Good stuff.
The only way is UP.