Home > Mobile > Follow-along Campaigns

---

Goal to learn AM as fast as possible (10)

---

Intro:
So with the advice of manu_adefy and Mr Payne, I'm starting a follow along here to document my AM journey and hopefully get some feedback from the community with where I'm making mistakes. I have small background and experience in Facebook, SEO, and copywriting but decided to make AM a full-time career because of the benefits the nature of the work provides (location and financial freedom), my love for numbers, making logical decisions, being creative with ad angles, and being able to solve people's problems at a mass scale from 'behind the scenes'. My weakness is in reading/analyzing too much and not taking enough action due to fear, but hopefully this follow along will help keep me accountable and sane throughout the AM journey.


The goal:
My ultimate goal is to become an expert at this, learn as much as I can, as fast as I can and take everything I learn into action. I've been following along with Andrew Payne's first follow along and other follow alongs here, and realized how supportive the STM community is (which is the main reason for me coming out with this.) I can also relate with Mr Payne's reason why behind journeying in AM and hope to master the mindset and processes needed to make this happen.



However, everything needs a goal with a number, so here's a preliminary goal to start things off:
-> Achieve $xxx/day revenue within 30 days using low payout offer(s) (<$1) in tier 2/3 GEO


The campaigns:
So far, I've spent about $200-300 testing a few offers and landers, mainly antivirus in India, but nothing too promising.



I came across a currently popular offer and have gotten a couple of conversions with an ROI of about -75%, but my thing with that is, it's a very popular campaign according to adplexity, and I could see by the trends that some affiliates are already in the process of scaling it now. The network that I found the offer on, is pulling the offer in one GEO in a couple days, which is leading me to believe that the offer might be pulled in the other GEOs as well soon. So I'm hesitant to start working on optimizing this (the offer in the GEO i got conversions in hasn't been pulled yet), for fear of the offer getting pulled or too crowded by the time I make it profitable.

My alternative strategy was to test other offers and see if I can find something else profitable or less competitive before anyone else has gotten to mass scaling it.

Is this a wrong approach? OR should I focus on trying to optimize the offer + geo + lander combo I found to convert?

The best campaign I've found so far, happened to be a smartlink offer from one aff network that happened to convert from a segment of traffic from another offer (I used the smartlink as a backup offer for untargeted traffic that leaked in thru the campaign). Hard to scale this, as the targeting settings used was just for a range of IPs, and the aff network wasn't able to see which specific offers from the smartlink were converting, so I couldn't really setup any landers for testing. I've since abandoned this campaign.

Next steps:
Either continue optimizing the offer + geo + lander I found converting at -75% ROI with the risk of the offer being pulled too early, or go find and test other offers.

My other strategies to find and test offers are:
- Mass test offers direct link in the same geo (regardless of vertical) to see which ones convert
- Mass test offers direct link in the same vertical to see which ones convert

Also question: are there specific verticals in which direct linking won't convert, and it's almost absolutely necessary to have a lander? If so, what are they?

Thanks in advance,
Grace

---

Cool follow along.

Are you making your own landers or are you copying what you see on Adplexity?

---

Thanks for starting this follow-along!

India is a big geo - it would take considerable time for an offer to become saturated. I don't think that should be a concern.

Are you running carrier or wifi traffic?

You've spent quite a bit of money running in this geo already. Have you analyzed your placement stats to blacklist the bad converters? If you could start your next campaign with the bad placements already blacklisted, that would give you a good headstart.


That's one drawback about certain smartlinks - they don't tell you details on which offers have converted.

Some networks though WILL show you details on which offers have converted - Kimia for example does this. However, some of their offers are off-market which means you can't run them directly (due to limits on cap and/or other reasons), but sometimes you'll see an offer that converts well that you CAN run directly and be in profit right away.


If you're targeting wifi in India, you can likely cut your way from -75% to green and STILL have enough traffic to profit from. However, before you cut too heavily, make sure you've tested enough landers and offers. How many landers have you tested already? For a big geo like India, I would suggest testing at least 10 landers. Nail down a good lander, then use it to mass-test offers. Do your reasonable best to improve your funnel (lander+offer) first, and THEN cut anything that's not profitable.

In short, try to make enough of the traffic profitable first by testing landers and offers, and THEN cut the unprofitable parts. If you skimp on testing offers and landers, but instead focus on getting green through cutting (placements, OSs, carriers, devices, whatever else), this would be the worse approach. Not only would it be expensive to identify what to cut, but you also won't have a lot of traffic left after the massive cutting. Another consideration is when you get around to scaling the camp to other traffic networks - you'd need to repeat that heavy cutting on every network, which could get expensive and annoying.


Whether this would be a good approach or not, would depend on what type of offers you're dealing with. Are we still talking about antivirus offers here? If so, antivirus offers don't tend to convert very well without landers (and aggressive ones at that).


LOL! Nice to see this question after what I've just said above.

Some offers need landers to presell visitors in order to convert well. Offers that can work with direct-linking are either 1)ones that require little to no pre-selling, or 2)already have landers built into the offer (aff networks and advertisers call them "pre-landers" whereas we simply call them "landers"). Examples of offers that require little/no preselling are carrier-billing offers that have very simple conversion flows (1/2-click). So if you're wanting to give direct-linking a try, ask your AMs for 1/2-click offers and/or offers that have pre-landers built into the offers.

A tip on running direct-linked offers:

If you're running carrier-billing offers, don't spend too much budget on testing each offer. Carrier traffic doesn't have bot traffic like wifi does, so there isn't much you can cut to improve ROI. And because you're not using landers, you don't have much room to optimize the camp. If you're already bidding average (i.e. traffic quality is decent), and if you don't get a conversion after 5-10x payout in spend, I'd say ditch it.

Instead of spending lots of budget on testing each offer, spend a little bit on testing each offer but test a TON of offers.


Looking forward to your next update!



Amy

---

I'm ripping + cleaning + spinning a couple variations with my own angle/copy/ideas

And Wow! @Amy, I knew by reading the other follow alongs, that input like this would be so helpful, but it's really different when you apply yourself, and you're the one getting the specific feedback and help. So thanks so much for replying!

The offer + geo + lander combo I found to convert is actually in the UK, not India. And considering that the UK is a big geo too, I'm assuming it will take a while for the offer to become saturated in there as well too, right?

I'm running all connection types on RON with the UK campaign. With this, I've also ran a preliminary bot test and blacklisted the bad placements before getting the conversions.

Anyways, I've decided to continue running this campaign, also because I've seen the daily cap on this offer/geo across the network it's in is only a quarter of the way there and it's only been half the day. The average bid on the geo in the traffic source I was running was about the same as the other tier 2/3 countries, so another reason why I've decided to continue running it. Although the avg bid is slowly increasing the past couple days/hours, which just goes to confirm how competitive and volatile this niche is.

So with the converting UK campaign, I've blacklisted the bad placements (spent x1 of offer payout with no conversions), and added 4 more landers to the mix for a total of 8 landers being tested.

Also, I'm heeding your advice and testing a ton of offers with a small budget for each: Just launched a campaign in a geo w/ 5 offers and a second campaign in a separate geo w/ 4 offers, similar to the way Mr Payne explains his "Point and Shoot Method". Both campaigns with $10 budget each and direct linking just to see if an offer will convert. Offers all came from recommendations from an affiliate manager or from a top offer list in their specific geo's. All are also either 1- or 2-click flows, so hopefully will show some conversions with DL.

Things I've noticed:
- Speed is essential here. Being able to find an offer, create and launch landers, setup campaigns, etc. has to be quick in order to keep up with the competition
- There are so many things that can take your focus away, (for example, an affiliate manager recommended me to switch verticals and gave me a list of top antivirus web offers doing well.) I saw that it is absolutely possible to make those work, but I wanted to keep my focus on mobile and master it before moving on to something else, since I thought if people can make mobile work, then there's no reason I shouldn't too Another thought on that is: if I can make it in a vertical as competitive as this, then making it in a vertical less competitive should be easier!

Update on these campaigns coming soon...

Grace

---

Update:

For the UK campaign, here are the stats for landers:



What I've done:
- cut the 2 worst performing landers according to stat sig calc
- adding more budget to keep collecting data (currently spending around $4 per hour at current bid, which is slightly below avg bid currently. When launched the bid was slightly above avg)

Next steps:- Create/prepare more landers to test for this campaign, hopefully to find something that will make it break even (currently the lander + offer combo is at -50% ROI) Any suggestions to get this campaign to breakeven faster?
- Continue cutting any placements spending x1 of offer payout
- Learn how to do bidding tests and try one here (is there a specific time in which I should start testing bids? or is now a good time?)
- I just checked the device analysis on this, and noticed interestingly that tablets have a much better ROI and volume than mobile phones (see here) Should I cut mobile phones, and just rely on tablets to bring the campaign to breakeven/profitability? I know doing so will greatly reduce volume and opportunity to scale on mobile phones, but maybe another strategy can be to work on the funnel first with tablets and then open up volume to mobile phones again?

Question:
- Am I going on the right track with this? The campaign has an overall ROI of -75%, and the best lander has an ROI of -50%. What are the top actions I can do to further move the needle on this, getting closer to breakeven?



For the smaller geo campaign testing 5 offers: no conversions yet, traffic is slow

For the other geo testing 4 offers: got several conversions, stats below:



What I've done:
- cut the worst performing offer, and continue running traffic to find the best offer to go with (or should I pick the top offer already and focus on testing landers with that one?)



Thanks,
Grace

---

Update:

Finally got green!

Here are the stats from the UK campaign yesterday from the time I made changes to the time it got cut off (due to AM pausing my link from tracking issues, but hopefully that's being fixed today).



I managed to breakeven with just a tiny bit of profit, after cutting mobile phones, mobile carriers, and chrome browser, since those were the ones converting poorly.

I know cutting mobile phone devices will reduce volume greatly, but my strategy was to try and get tablets profitable and then scale to mobile phone devices when ready.

I'm also competing with other advertisers for the daily cap on this on the network, so it's quite a race.

My thoughts:
- If I can just make the funnel better and ROI higher, I can have a much better and easier time scaling this (any tips on the process for doing this? I was testing 8+ landers in the past 2 days, but it didn't go very well, because I was adding/pausing landers every hour or so due to emotions, and making a mess out of analyzing it, so truth be told, lander testing didn't go very smoothly.) However, when I was looking at lander stats, it looked like even after 10k-20k visits on each lander, a winner was still not coming out confidently, so I would have to just keep running them until a winner came out, then just pause the losers. Is this the right way to approach it? Or is there a max spend or max number of visits to feed a lander before deciding on one winner?

Update on geo 1 testing 5 offers: may have to rerun again since a ton of wifi traffic was leaking through

Update on geo 2 testing 4 offers: offer that got most revenue and conversions is a 1-click flow gaming offer (see stats below) any lander tips on getting this to convert better?



Next steps:
- Continue looking for the same offer in other networks to test (Just found one network, applied to it, and just waiting for approval now. Also asking my other affiliate managers if they might be able to get it)
- Do a lander immersion and look for/brainstorm more angle ideas for UK campaign with hopes to better the funnel enough to increase conversion rate and ROI significantly. Also keep a lander testing process down intact so I don't keep making changes in the middle of testing
- Test bidding on UK campaign by cloning campaign and using separate bids for each to see which one converts better (although, wouldn't it be "self-competing" against oneself if someone runs multiple campaigns at the traffic source targeting the same settings at the same time?)
- Brainstorm/research some popup lander angles for converting gaming offer (although, I have an idea that display ads might be better catered for this offer?)


Thanks,
Grace

---

Much progress has been made!

Nice!


Yes! It's not even as much keeping up with competition, as being able to keep finding profitable camps to replace dying ones.

Pop camps don't tend to remain profitable for long. If you aren't constantly testing and scaling, you won't be able to stay at the same revenue levels, let alone grow your revenues.


Good thinking. Plus antivirus offers can be difficult to run if you're not cloaking. Aggressive marketing (e.g. "Your machine has been infected with xxx viruses and will explode unless you install this software!") is normally required to make profits, and yet most traffic sources don't allow this sort of landers.


Good.


-Are you already down to your best offer+lander? -50% ROI may be a bit far from breakeven to rely on cutting placements to get to 30%+ ROI. It would be better to test more offers instead - that's usually the quickest way to increase ROI.

-Drill down to OSs, browsers, devices....etc. and cut those that are doing the worst ROI.

-It would be better to finish your lander+offer split-testing first, and then test bids. That way you won't have so many variables you're testing at the same time (you'll get "noise" and less accurate results from so many variables). But if you really want to experiment with bids now, don't let me stop you. If you want to test bids now, take what seems to be your best lander+offer, set up a camp, clone it a few times, and assign different bids to each. Run them for a while and observe placements stats mainly - see how many profitable placements are triggered by each bid. Keep the camps that have enough profitable placements and optimize further, and pause the camps that aren't able to make enough placements profitable.

-Yes you can absolutely break tablets off into its own camp. See how much traffic you can get - if there's not much then I wouldn't bother.


If traffic is slow and you're already bidding above-average, then you should ask yourself whether it would be worth your time to even test in this geo and on this traffic source in the first place.

If there's not enough traffic to make beyond a few bucks a day - I wouldn't bother.


In the beginning (i.e. your current stage) the main focus should be on cutting down to a winning lander. In order to reach this goal you'll only need an offer that converts semi-well - basically well enough such that the testing won't cost you an arm/leg.

So yes, you can take any offer that you know converts and just use that. Sometimes I would throw a bunch of offers and landers together, wait until an offer reaches 2 conversions, and pause all other offers and just use that one offer to cut landers down to a winner. But then I would retest all the offers again - because I had not waited for stat sig the first time around.


WOOHOO!! Green is always good!


If the traffic volume is simply not there, I would suggest to test more offers to make MORE of the traffic profitable.


A resounding YES on focusing on making your funnel better!

If 2 landers keep "going head to head" by performing the same - just cut one of them and leave the other, then add more landers to start another round of split-tests. For details please see this post by caurmen:

https://stmforum.com/forum/showthrea...ugh-Clicks-Yet

Paying particular attention to point 3: "3. If you've had a few hundred clicks worth of data for each landing page, all of them have probabilities of being best between 25% and 75%, and you're seeing high, overlapping graphs like the ones below, the remaining landers all perform about the same. Pick one to continue running, pause the others, and start a new split-test (see "What To Test Next"). "


Update on geo 1 testing 5 offers: may have to rerun again since a ton of wifi traffic was leaking through


I would say ditch the offers. When you're direct-linking to carrier-billing offers on pop, there just isn't much room for optimization. The 2 major tasks would be to cut placements and test bids. But the CR are not looking promising enough to justify cutting placements...


-Testing more offers will often get you better results then testing more landers (assuming you've tested enough landers so you're running with one that converts decent).

-For carrier-billing gaming offers, it can be difficult to make landers that will do more good than harm. Landers are a double-edged sword - basically it will need to pre-sell so well that the increase in CR will outweigh the decrease in the amount of visitors that will actually click through to land on the offer page. And for something like broad-appeal mobile gaming offers, pre-sell is usually not a necessity - it's just not a place for most landers to shine. Don't let me discourage you from testing! Just because I didn't succeed at it doesn't mean you won't! But if you do, try to keep it very simple - an extremely attractive graphic with a short headline and a CTA may be worth a shot. I've seen similar landers used for video subscription offers with success. May just work for gaming offers.

-If you're wanting to try gaming offers on mobile display, here's a rough outline you can follow:

https://stmforum.com/forum/showthrea...l=1#post299115


Will be waiting for your next installment!



Amy

---

Hi Amy,

Thanks so much for all your advice!

It's been a couple days, so here's an update on what's going on...

First, it's been an emotional roller coaster ride, and I haven't even gotten to the point of scaling a profitable campaign yet. The UK campaign hasn't been consistent at breakeven/positive ROI enough for me to scale, even after cutting mobile phones, placements, and other segments. The daily cap for this offer at the network it's on has been reaching 100% for many days, leaving me with such little room to test and collect more data. However, I've just got accepted into a network with the same offer and going to test as soon as I get approved.

I also reached an emotional down, when I didn't feel like doing any work for days...Did some meditation and self evaluation and figured it was maybe due to overwhelm. I would be launching many campaigns with multiple offers in multiple geos, and the process of having to check numbers with every campaign, ensure not one placement is eating up all budget, analyzing offers/landers, adjusting, etc. It got to the point where my brain couldn't handle it, and it made me realize a system is very important in managing multiple campaigns. And not just a simple generalized system (where 1. research offers 2. create landers 3. launch 4. analyze numbers 5. adjust), but a system that is extremely specific down to the point where you don't really have to think about anything; you just do, and take action, leaving emotions out of everything.

Well, I've developed something and wrote it down on paper out of everything I've learned so far about testing and cutting offers and landers, that I think I've got it down somewhat. Now I'm at the point of launching 2-3 tests per day, managing about 5-10 campaigns at a time, all without going crazy. (I've picked up Amy's advice somewhere about widely mass testing offers and picking the best of the best ones to optimize rather than testing just a few offers and picking a mediocre one to optimize).

I've also read somewhere that some of the most successful people on the planet end up having to work years before their hard work is paid off, which eases me more seeing as how I had the expectation that I needed to master this in 1-3 months.

Anyways, good news, is that one of my campaigns unearthed a statistically significant best offer, and best part is that it's in the green!



Bad news(?) is that it spent only $0.71 total on that one offer, which means consistency is very unsure of. So only running more traffic to this one offer will let us see.

Questions:

- Sometimes, I launch the same campaign (same geo + offer) on multiple traffic sources, and these sources often times get different results. When I was trying to cut offers at the time, I had one TS get one conversion on an offer and the other TS get another conversion on the same offer. When analyzing and cutting these, should you analyze the numbers on a per traffic source basis and not combine the two traffic source data together? Or is it okay to use both traffic source data when identifying offers to cut? (For example, in this scenario, if I used both TS data, I will reach a stat sig best offer, however if I analyze the traffic source data separately, I will still need to keep running traffic to both campaigns before reaching stat sig. Which one would be ideal?)

- I've launched ~4-5 campaigns and ~10 offers before finding this one converting offer (vertical is mobile content). And after browsing many affiliate networks, I noticed that a large majority of offers are mobile content. Is this true? or has it just been on the aff networks I've been on? And also, if the large majority of mobile cpa offers out there are mobile content, rather than sweeps/anti-virus/etc, does that mean it takes more offer testing before finding a working mobile content offer, rather than a sweep/antivirus/utility app offer? Or is this question just missing the point overall?

Next Steps:

- Continue managing campaigns and testing to find converting offers, while also working on tweaking processes and systems
- Continue driving traffic and optimizing converting offers

---

Suggestion: Before doing any heavy cutting, try to decide whether there are enough profitable traffic segments (a group of high-volume placements, a major OS, a mobile carrier with good volume...) as early on as possible.

A typical newbie trap: I've seen many new people try to cut their way into green, when there aren't even enough profitable traffic segments in the first place - meaning the cutting efforts would be futile - i.e. no amount of cutting would lead to significant amounts of green.

Basically you'd want to run the camp until you can decide either:

1)There ARE enough profitable segments to justify cutting the unprofitable segments, or

2)There are NOT enough profitable segments, in which case you'd want to either ditch the camp, or improve the funnel, or test other bids.

I'm not sure if this was the case in your particular case, but either way, trying to make this decision as early as possible can result in cost-savings.


Good - very very good!

Having a system that you can refine over time is definitely one of the best things to put in place for your business.

Well done!


I can understand that sentiment for sure!

I know how a lot of people are goal-oriented and how all the motivational speakers in the business field are touting the importance of settings goals and deadlines and tracking results...but I really don't believe that approach to be the best when it comes to getting started in running paid traffic campaigns (or getting started in ANYTHING that has a steep learning curve, for that matter).

You almost can't give yourself deadlines or even plan profits projections, at least not before you've built up the self-confidence and experience required to be making profits on a consistent basis. As you've mentioned, the emotional roller-coaster can result in not wanting to take any action at all. After all, a camp that was never set up, cannot fail.

A much-better goal to set, would be a weekly/monthly budget that you can aim to exhaust by the end of every week/month. If you're able to set aside some disposable income that you can afford to lose every week/month, you can aim to spend the entire weekly/monthly budget, every week/month. This way, you're not overly focused on profits - your focus would be on finding things to test (offers, landers, traffic sources, different traffic source settings...) instead. You'll still spend your money as efficiently as possible (after all, it's still your money!) But now you can overcome the inaction issue because the pressure of making the campaign profitable is OFF, and all the experience you accumulate will eventually and naturally start reflecting on your campaigns. I've made the same suggestion to another member a couple days ago in this post:

https://stmforum.com/forum/showthrea...l=1#post320724


That's tremendous!! Good job! I know it doesn't look like much, but green is green - that color sure is easy on the eyes!


I get this question all the time. When you change the conditions, the test results will vary more or less. Even when you're running on the same traffic source using the same targeting, if you change the bid you can get different offers/landers that win out. Heck, even if you were bidding the same, running during different times of day or days of week may give you different winners.

I'm going to have to bring up the same concept I've talked about time and again when it comes to campaign optimization, and that is the concept of accuracy vs. efficiency. These two have a yin-yang relationship, where when you have more of one, you'll have less of the other.

There's no doubt in my mind that you'd get more accurate results by cutting offers separately at every traffic source. But the question is, is the extra accuracy worth sacrificing efficiency for? (Efficiency = time and money.)

That's a personal question that only you can answer. Test different approaches and see which one works better for you. To me personally, I strive for efficiency because I don't have much time to run camps - I would only go for accuracy where it matters most (e.g. checking for statistical significance when cutting landers and offers).

For your specific example, I would just test offers and landers on a single traffic source, find the winners, and scale to other traffic sources - without testing them at every source.

Is this the most accurate way? Maybe not. But this way I save time and money that I could use to test new offers - which would arguably have better profits potential than any extra ROI I could get by striving for that extra accuracy (by testing on each traffic source). I don't have concrete stats to back up this claim though - so please feel free to do your own testing to arrive at your own conclusions.


I can't say I've had to test more mobile content offers to find one that works, compared to sweeps/antivirus etc.

However, if you're referring to carrier-billing mobile content offers of the 1/2-click variety, then yes, you tend to need to test a lot more of these to find one that's profitable (unless you employ some blackhat tactic). This is because these offers are mostly promoted via direct-linking, so there's really not much optimization one can do, and therefore competition is fierce. But then, it takes a lot less budget to test this type of offers as well - because there isn't a lot of room for optimization, if the offer doesn't convert after 5-10x payout, then you probably won't be able to make it profitable.


Sounds good! Keep on keeping on! You're doing splendidly!





Amy

---

Hi everyone,

Apologies for a delayed post.

But here's what I've been working on...

Last 7 days of current campaign:
http://imgur.com/a/axHDe

With this campaign, I've tested:
- 4 offers, and found 1 converting (screenshot of all offers tested)
http://imgur.com/a/O5xhN

- 16 landers, paused 8 worst, still running 8 landers to find a winner (screenshot of top 8 landers)
http://imgur.com/a/92I90

- Tested bids (both smart and legacy on Popads), found a slightly higher than average bid on legacy to produce the highest ROI, however traffic speed is very very slow, and I want to get my ROI a little higher before going back to smart bidding with increased traffic speed (or should i take a different approach?)

- Checked placements for anything over 2x payout spend and no conversions, and haven't found anything; I'm assuming using the legacy bidding feature on Popads is a small culprit in this

Questions:

- As you can see in the first screenshot, I've managed to get green for two days, which is the weekend so I'm guessing weekend works best for this offer, but today it's red, and overall, the campaign is just barely breaking even. How can I move the needle on this campaign, and get it to "work better"? In other words, get higher ROI, and keep it consistent enough for me to scale it? Is this campaign good enough for me to keep running budget on it, even though it's just on breakeven? I know they say to only focus your resources on the higher potential campaigns instead of the lower potential campaigns... so is this what you would call a "low potential" campaign enough for me to ditch it? Or could I still work on it to make it better?

Next Steps:

- Besides this campaign, I already have a new campaign set for approval (1 geo + 2 mobile carriers + 10-20 offers/carrier + 3 top landers/carrier) and 4 more geo's worth of offers to setup.

---

Home > Mobile > Follow-along Campaigns