Home > Mobile > Follow-along Campaigns

---

Noob Affiliate Marketer, Experienced Copywriter (This Should Be Interesting) (30)

---

Hey,

So I figured the best way to learn is to start, so here I am. I'm going to write this like a journal because I have soo many questions; this follow-along is basically going to be a place for me to brain dump my thoughts as I run my first campaign. This should be interesting.

I'm going to with mobile for my first campaign because I think it's something that'll be around for a while.

Anyway, here's where I'm at right now.

I've completed as many steps as I could in "The Appetiser" guide. I'm still waiting for approval from Decisive; I've sent them an email and haven't gotten a reply. I did apply on Friday night, however, so I'm guessing they won't get back to me until Monday. I'm going to give them a call tomorrow afternoon and see if I can get through to someone.

I'm confused about how networks and offers work, so I'm going to start reading more guides about that after I make this post. From what I understand, you sign up on networks as an advertiser and try to get in touch with an affiliate manager; from there, it seems to me that you request a list of offers that are currently converting at a high amount of volume.

For the graphics, I'm going to use Canva. I'm surprised no one on these forums has mentioned it; it's a great tool. Easy to use, and it's free.

So that's where I'm at right now.

Moving forward, here's what I have to do:

- Figure out how offers & networks work (F5 was no longer avail, btw)
- Get confirmed on Decisive
- Figure out what's next

One thing I'm considering is building a lander and collecting emails. I've been writing direct response copy for two years now, so I'm comfortable with converting cold leads into buyers. One of my mentors right now has 10 years of email marketing experience and, at one point, helped a biz handle 280,000 email leads per week for several months. The experience he gained from that was, as I'm sure you can imagine, very valuable. Anyway, he's teaching me how to write to a mass market at that level. Everything I've been taught about copywriting has gone at the window. And I like it.

In sum, I know a lot about copywriting, conversions, and how to use emotions to get people to do stuff - but I know absolutely nothing about how to run an affiliate marketing business. I mostly write the copy, not handle the traffic and campaigns.

So that's where I'm at! Updates coming soon.

---

Update.

I found myself an offer, but it's not an offer from a network. The offer is a free + shipping deal - the lead gets something for free in exchange for the shipping cost ($4.95). There are two upsells after they enter their credit card info - one for $9.99 and one for $47. I earn 75% on all commissions - including the shipping cost. It's a US only offer.

I'm going to use Airpush to send this offer as a push notification to someone's Android phone.

Update: I've funded my Airpush account with $100. I'm running a 728x90 banner ad. Waiting for approval.

Note to self: look into Airpush's "Abstract Banner" options moving forward + reach out to the AM of my offer and get a HD video.

---

what the payout from the offer? if it's over 20usd, which it should be for a cc submit, then your testing costs will be very high. like many thousands of dollars.

---

Hey Raymond. Sounds like you've got a great skillset to leverage in this industry. I'm trying my hand at email again - it's pretty intimidating. Couple things standout in your post:

What do you mean no longer available? Too late at night?

Is the offer you've selected already a proven offer, with other affiliates pushing decent volume to it? That's the really important thing.

The second thing is I'm not sure why you're starting out on Airpush. Have you spied others running this offer on that source? If not, I wouldn't start your mobile journey there. It's not very newbie friendly and quite likely to blow through your budget and overspend your daily cap, leaving you no results or meaningful conversion data to show for it, especially for a CC submit offer. However in fairness to them I haven't used them since 2013 so things may have changed substantially.

It's really important starting out that you get conversions of any kind so you can learn what works on a given traffic source. I think you'd learn much more on a DSP like Decisive or Go2mobi with a low payout offer ($3 or under, even as low as $0.25).

---

as folks already said - there is too many unknown variables in your case. you don't know the offer, you don't know the source network, if all went wrong - you don't know who blame to.

for start - an offer must be proven (would be in a top of network or has personal recommendation from trustworthy person) and payout for a newbie must be in $0.27 - $1 range. lesser payout - too risky to go deep negative, higher - to stay without any statistically significant data (to burn money for nothing).

---

You mean from the affiliate? I believe it is $100.

---

F5 is no longer accepting affiliates.

The offer I'm sending is proven. I have data to back this up.

I choose Airpush at random; not the best decision for starting out, I know.

I'd have liked to use Decisive but I'm still waiting for approval. I didn't know about Go2mobi - I just signed up for it.

---

You're probably right - but I'm willing to take the risk. I have a cap of $15 spend a day, and the cpc buyout is really low.

The offer I have is proven; I know for sure 150k people opted in for it in 2014. It is one of many offers in a hyper-responsive niche.

The payout violates the what the guides say, buuut it's something I'm willing to test - I'm too curious to not, not do it (could be the stupidest thing I've done all year, who knows...)

My ad's been approved, now I'm just waiting for payment approval from Airpush.

---

"$15 spend a day," that is your testing budget? with a 100 payout that's gonna take months for you to get meaningful data. u gotta do something like 1.5x offer payout per placement. and you also have to test all the creatives.


that will be like 5k or so minimum just on airpush to get reliable conversion data.

do you know the offer converts on mobile push traffic? generally cc submits perform better on desktop. but there are exceptions.

---

if ever

---

I have no idea how attractive this offer is, but I would suggest that you model out what you would have to believe in order for this to be profitable for you.

What does your CPC/CPM have to be, what ad to landing page CVR need to be, what does the initial offer CVR need to be, what does the upsell(s) CVR need to be, how low do the chargeback rates need to be, etc?

This should at least give you a clear idea of what you are up against. Right now, I honestly get the sense that you may be flying a little blind.

Knowing this for ALL of your campaigns, i.e. exactly what each of these metrics need to be and whether they are realistic or not is a great discipline to have. Getting someone to take out their credit card and put in their details (ESPECIALLY on a mobile device) is not easy, so if your model shows that you need to have traffic costs of CVRs well beyond the realm of industry benchmarks, then this may lead you to rethink your approach.

None of this is meant to discourage you. Quite to the contrary, I want you to have the best first campaign set up before you spend money so that you maximise your chances of success.

---

I don't understand the connection between testing budget and payout, can you explain? Shouldn't $15 of test give me insight on conversion rates? Perhaps I'm misunderstanding what payout is; isn't payout the threshold set by the entity I'm getting the offer from? Or is it something else?

I don't know how the offer converts on mobile, but I could reach out to the AM and ask. I'm going to test an opt-in as well as the direct sale.

---

Test budget and payout are not necessarily related, but most affiliates use it as a rule of thumb (in my view actually incorrectly).

However, in your case, $15 of test most likely will not give you any meaningful insight on offer conversion rates for this particular offer.

---

Ah, okay... makes sense.

Update on my first campaigns: I got accepted to Decisive, so now I'm going to test that out while I wait for Airpush payment approval.

---

Update: Airpush ad is live. So excited.

---

Good luck!

---

Paused the campaign.

Spend: $7.05
CTR: 11.01%
Clicks: 47
Earnings: 0

I paused it because out of those 47 clicks, not one went to the "enter shipping details page". This means that no one seems to be interested in the offer, which leads me to believe I'm off with the targeting. I'm going to review the data and try again.

---

Hey cmdeal would you mind explaining a bit more why you feel the relationship between offer payout and testing budget is often misunderstood?

---

If you have a stats or econometics or mathematics background, you will immediately see that when you are calculating chi squared or p values for campaign conversion rate tests, there is no formula input in either the numerator or the denominator for a payout amount. Statistical significance for conversion rates is driven by sample size (n) and the magnitude of the conversion rate differential from an expected value: (expected conversion rate - observed conversion rate)^2)/expected conversion rate.

The amount of payout does not play any role whatsoever in this equation.

In other words, the amount that you should spend on testing is whatever amount will get you a large enough sample size to get you to statistical significance on your conversion rate. The size of sample size you need, in turn is related to determined by the size of the differential in conversion rates you get from each test.

Two simple examples will show you why.



The statistical significance of campaign's conversion rate does not care about payout.

These maxims like "spend 1.5x payout on each campaign," or "test campaigns with low payouts on cheap traffic source because you can get to statistical significance faster" are heuristic rules of thumb. Nothing more, nothing less.

Heuristics are designed to give good enough and directionally approximate answers to questions that come up especially in situations where sound decision processes making may be to costly or time consuming.

In that sense, they are like any other rule of thumb that people use, whether it is "Attend any party or event to which you are invited," "If you are not sure whether to include some text in your writing, cut it out," "When in doubt, do what Jesus would do," "Neither a borrower nor a lender be," "Spend 3 months' gross salary on an engagement ring."

They are not meant to be followed blindly like sheep.

---

Thank you very much cmdeal for you reply.

I understand that payout value is not directly related to a test of statistical significance. However, as you mentioned the conversion rate is. And in general, conversion rates are much lower for higher payout offers. This isn't because the payout is high of course, but must likely reflects increased necessary action by the user, like submitting their credit card details.

So I guess it is not high payout offers per say that require a large testing budget, it is low conversion rate offers.

When you are testing placements/creatives, do you use a conversion rate heuristic rather than a payout one?

With low conversion rate offers, is it possible to test them on small budget? Or do statistically significant tests require large spends to get the necessary conversion data.

I imagine, one way is to fix as many variables as possible, either through spying or past experience. So that you can focus your spend on 1 or 2 key variables at a time.


Though, I suppose with cheap enough traffic anything is possible...

---

Not exactly.

This is assuming an equilibrium condition where the payout from all advertisers is set at just above the marginal cost of generating the next incremental conversion across the entire market of traffic sources and campaigns. In such a situation, the most that anyone would be able to earn from affiliate marketing is a return not much greater than the market cost of capital (which these days is just a couple of basis points). As we know from those who are doing decent numbers on their campaigns, the market is VERY far from an equilibrium condition.

You need to be careful that you are not confusing the direction of causation and the extant of the correlation across the various metrics you track. The payout amount by an advertiser is driven by a number of reasons, many of which can have absolutely nothing to do with the marginal cost of generating the next incremental conversion across the entire market of traffic sources and campaigns. The payout amount can be because the advertiser needs to meet a certain quarterly goal for the number of installs, it can be because they want to pursue a burst campaign to expand discoverability of their app on the various app store leaderboards, it can be because their internal head of acquisition got poached by Pinterest and now they have no skills to run any activity in house, or it may be because they really do not understand the Indonesian market and they need someone else to figure it out for them.

Indeed, the very fact that the market is not in constant equilibrium is what allows you and many others on this forum to generate excess profits well above the market cost of capital, and hence we should all be grateful. This is why competitive advantage is even possible. Otherwise we would all be better off in a 9-5 job.

---

Therein lies the source of our friendly "rule of thumb".

It is not the payout that is that important, it's that the payout is generally commensurate with the amount of user intent required or the difficulty in moving users to the end of a funnel (it really depends on the offers) - i.e. higher payouts often result in lower conversion rates, ceteris paribus. This isn't necessarily true, just often observed. A trend I suppose.

This is a good point to be mindful of given the above ^^
To get to a practically useful situation where you actually have conversions and thus a numerator in the fractional metrics you are often looking at, you generally need to spend more if the conversion rate is lower. That much is common(ish) sense.

My bullet point thoughts are:

- When initially getting data and having few conversions, I consider if adding or subtracting a single conversion from any test would drastically change the significance of the results or your decision. If it would, you don't have enough data (or statistical power). Note that simple calculators won't necessarily tell you this by design and they again don't use payout, so cannot compute the change in EPC you might get from an extra $50 conversion.

- When getting data and having no conversions yet, this is a situation where you can use something like binomial confidence interval calculators to get a predicted conversion rate range. The lower limit clearly will be zero, but the upper limit will give you a much more statistically driven cut-off point for killing things off (assuming of course you are taking into account confounding variables, gotta love mobile traffic!).

- If you have multiple tests running and the data shows that one of them is performing much worse, for whatever reason, it is illogical to keep that one running. If the predicted conversion rate based on the data tells you that there is virtually no chance that it will outperform the top variation, then spending some arbitrary amount more on it is again illogical (keep in mind my one-conversion-effect rule of thumb).

- The statistical methods and approaches are quite clear but not necessarily the most economically practical, so it's up to you as the intelligent marketer to factor in things like anticipated conversion rates, payouts, data costs, confounding variables and so on rather than blindly following either rules of thumb or pure statistical analysis.

---

haha, true.

I guess i need to find some high pay out, high conversion offers!

---

Update: making minor progress; feeling like I'm not getting anywhere, yet learning stuff at the same time.

Nothing major to update everyone on - just going back and forth with different aff and traffic networks, trying to understand this side of the aff marketing biz.

The more I get into it, the more complex it seems; which is annoying because I like simple. What's confusing to me is: how all the networks seem to connect to each other in some way or another. For example, I signed up for a traffic network but had to use an aff network to get my offer, but while signing up for that aff network, it was recommended I go with another traffic network. Raymond Duke does not like feeling like a ball in a pinball machine.

I've actually funded $100 in two different networks, so I'd like to stick to those.

Anyway, latest update: still looking for an offer.

---

Another update: good news (but no results... yet).

I'm resorting back to my strengths: understanding a market, and giving it what it wants.

I'm going to use Decisive as a traffic source, and an infoproduct as the offer. Now, the key to making this work is to create my banner as a "native advertisement" on the website I'm targeting. Meaning: the banner I'm creating will look like the same type of content as the other articles on the website - same font, picture size, and whatnot.

A tip for doing this right: load the URL into Buzzsumo, and see what's popular on the website - then, create your ad accordingly.

For example, on the site I'm targeting there is an article that's relevant to my offer that has 98k shares! This tells me: people like this type of content on the site, which means they'll relate to the banner/offer I'm placing on it.

I could go the extra mile and make a native article of my own - that looks like an article on the site... and maybe I will, but for now, I'm just going to test:

- a direct link to the offer.
- a link to a lander and capture the email (why a lander? because the offer has a VSL (video sales letter - and I'm thinking, people on mobile will not want to sit through a 30 minute video)

As for the images: I'm using Dollarphotoclub to download images for $1 each.

More news on this coming soon - I'm going to run the offer by my AM and get feedback about it before I begin.

---

Another update on this.

Still talking to AM and figuring out this side of the biz (it's so funny because I've been writing copy for the longest time, yet I don't know how to set up campaigns).

I just went live with a US offer (Lyft). I know starting with US isn't the smartest thing to do, but I'm doing it anyway. I'm stubborn... and I'm confident I can find a creative that works based on what I know about copy.

Yet, I also realize the market is what decides what works - not me or my copy.

Anyway, I'm still moving forward and learning along the way.

Oh, one more thing: I wrote a post today about the worst thing you can do when it comes to making decisions - deciding not to decide. If you feel stuck, check it out.

---

Another update.

Just not making the progress I want with this.

And that's okay.

My expertise is in writing copy - not running campaigns.

I may return to this at a later date.

For now I'm going to focus on what I do best, rather than learn something new.

The biggest problem I had with this mobile camp?

All the tech stuff.

I'm not a tech guy, I work mostly in creative.

Numbers and code and tech stuff gives me the hibby jibbies.

But I shall return ...one day.

---

Hey man, I just wanted to drop into your thread, since you also came to mine, and tell you not to give up! There's a LOT of money to be made in this industry.

And if it's just the tech stuff that scares you away - that's the easiest part! Especially with all these new trackers like Voluum, it's a breeze! You just need to get used to this whole new thing.

I come from a creative background too, and although I've had success in the past in this industry, right now I'm only losing money, I haven't seen even a single lead so far from mobile, but this doesn't stop me at all.

You should see me try to code an LP from scratch for example - you'll start pulling your hairs when you see how slow I work, or how crappy the code I write is. I've had coders curse at me when they see the type of stuff I code, literally, and it has happened more times than I can remember. But this doesn't stop me at all, if it's good enough, it's good enough!

Sometimes I still get frustrated when I have to setup a more complex campaign's tracking - lots of variables, links, things that can go wrong, but once you've done this a lot of times, it becomes second nature and won't bother you at all.

All I'm saying is, you shouldn't let some minor things like coding and tracking get the best of you. And numbers aren't all that scary either, you just have to look at a few columns of data, and most importantly the kind of money the campaign makes or loses, no need to over analyze and over complicate things. Or ultimately, if you like keeping things as simple as me, it's all about seeing money in the bank, or ideally in your hands! CTRs, CRs, EPCs and all that crap means nothing at the end of the day if you don't have cash in the bank, keep it simple and don't let it frustrate you. Besides people here would be happy to help if you're stuck with the techie stuff.

You've got the right attitude, and probably one of the most important skills - copywriting. Try and talk with some more AMs, and just jump into the offers that are actually working today. Ask them "What are your hottest, newest, most promising offers?", "Can you recommend some traffic sources?" etc. Screw Decisive, everybody starts there, try something less known, or go to pop traffic. Do some epic spying to see what ads and LPs are being used - with your skills you can surely improve their copy.

Or why not go to where your strong points are - build some mailing lists with paid traffic! I've done it in the past and they made me sales and profits, and I'm a below average copywriter at best. It took me forever to write some half-decent follow-ups.

Don't give up!

---

thank you so much for the positive words.

I'm back in the saddle, just had to rethink my next steps.

Your post here definitely helped.

---

Totally agreed with @shishev. I would add that you take some professional online course, it will save you from those early-stage annoyance of paradigm shift.

---

Home > Mobile > Follow-along Campaigns