Home > Mobile > Follow-along Campaigns

---

From Level 3 to Level 4 Follow Along (+2 Tips to Efficient Internet Collaboration!) (14)

---

If you don't have conversions you can use micro-conversions (the smaller steps along the way), the data isn't as valuable, but it can do the trick to get you making decisions quicker and see if they lead to green.

Examples of micro-conversions are:
*Ad ctr
*LP ctr (to the offer)
*Lp engagement (the bot test of 300ms is a basic way to do it and see if they're humans coming from the placement, search for the post)

---

Come on dude. So many follow along campaigns on Decisive with no conversion. My advice : try another TS.

---

As mentioned briefly in the original post, we are experimenting with another traffic source (Adivity. In fact a whitelist campaign at Adivity was our most successful campaign to date. I am currently designing some banners for a new angle that we're planning to try to see if we can boost our ROI. However, Adivity has it's own fair share of downsides (like all traffic sources). Live stats take an hour to update, during which time placements can spend $10 in a few minutes, it's interface is slow and poorly designed, it sends you an email every time you do something like add, edit, delete, pause or resume a campaign (seriously my inbox is exploding).

We're wary of jumping from traffic source to traffic source, assuming that it will fix all of our problems. There are people on Decisive making money from app installs. We feel that at this point, the factors limiting us are probably our angles, copywriting, and general inexperience. Poor quality traffic isn't going to help matters, but if there are people out there making money on Decisive then we want to learn how to do that too.

---

If the traffic source doesn't work change it.

Take 5 of the best angles (which you think are the best) and set them up on 5 traffic sources
TIP: find trap sources with their own SDKs.

and run a $100 test on each

compare results.

---

Interesting tip, what's are 'trap sources'?

---

Ditto what Walt white said..

What are trap sources? And by SDK do you mean api?

---

Sympathy. When you're trying to get rolling and it's just not working, that's tough.

There are two other absolutely crucial variables that I didn't see mentioned in your initial writeup: offer and network.

At the end of the day, the offer is what really matters: usually I'll test a dozen or more offers for every one that I actually stick to. And whilst for very experienced marketers with cash to burn, just hitting an known-good offer until you get it to work is a valuable strategy, for newer affiliates I'd recommend being a bit less stubborn.

I'm definitely one of the "launch a ton of angles and one will work" guys, but with a major caveat - I'd recommend spending no more than two angles on a single offer that doesn't show promise right off the bat. After that, move on, move on fast. There are tons of offers out there, and you never know what will or won't work for you.

Likewise, you don't mention split-testing between networks. If you're not doing this, I really, really recommend you start. The difference in conversions between networks can be MASSIVE, even on an apparently identical offer.

(Traffic source is also a very big deal - definitely worth testing on other sources, as many people have mentioned. And have you done any bot tests?)

---

Many thanks for the advice. We had been split testing networks during previous campaigns, but for whatever reason we haven't been doing so with UC Browser... Perhaps that's one more thing to try before we move on. (EDIT: Turns out I was thinking of another offer and we were actually split testing UC across four different networks, none of which we could get to convert more than a couple of times).

We've debated back and forth about adopting what seems to us like a fairly ruthless strategy of cutting offers after two failed angles as described in the main course. We definitely appreciate that not all affiliates can make every offer work, even when it's proven, but after seeing others around our level succeed and investing hours in launching a campaign it's hard to trash it and move on. But, this could definitely be hurting our progress, and I'm sure when we hit our first profitable campaign we will quickly forget about all of the other offers that didn't work.

Do you define promising in the same way as in the main course? e.g., -50% ROI on initial testing? For us, that just seems unrealistically far away. But perhaps we need to put our faith in the law of averages and keep pushing out new offers until something sticks.

We followed your bot test write up for a campaign we were running in Russia on Decisive, which was extremely useful and illuminating. Do you recommend running a bot test on every new geo / traffic source? Or just when you feel there might be an issue?

---

Could you clarify a little in that please? Do you use the same banners for different offers and just change text content and logo for example?
The point is keeping the highest perfoming ones and adapt them to other offers?

---

Yes, as well as being able to launch camps faster and then optimise once something has potential.

---

Great work guys. Good to see a change of traffic source seems to have helped - I'm just going down that path myself now. Congratulations on finally breaking even and even turning a little profit. I know how it is to have campaigns tanking day in day out!

Look forward to future updates with your success!

---

Good perseverance Steve!

---

Great thread, I learn a lot here. Can anyone suggest me what is the most effective Traffic Source other than Decisive?

---

It’s been a while but we are still here!

Things are going well from a learning perspective but we aren’t millionaires yet. On average we are probably doing X/day. Since the last update we have been taking things slow and focusing on internalizing campaign optimization and finding the right approach since we spent the first three months doing everything but that.

We are also trying to systemize everything as a part of building the company, so it’s done from the start. It not easy but hey, nothing easy is worth doing anyway.

We have currently broken down campaign optimization into five steps which does not include split testing as we do this all the time.

1. Finding converting offers
2. Optimise them from -50% roi to +30%
3. Add volume within the same traffic source and geo
4. Re-optimise (assuming that adding volume will cut the margin)
5. Scaling to other traffic sources and/or geos (and further optimizing)

We are especially focusing on step two and three now and trying to learn as much as possible from various experiments. Our goal is to break down campaign optimizing into smaller areas and internalize them so when a campaign acts a certain way we will immediately have an idea of why and what to do about it. We hope this will become very useful later both for ourselves and when training employees.

There is some great material here on STM regarding campaign optimization but if anyone has any suggestions for articles or books anywhere else on the web, we are all ears!

A funny thing I have noticed, as it is with everything in life, when you do something for a while you start getting answers to old questions you had months ago from completely irrelevant situations. Love it. If people didn’t see affiliate marketing as a get rich quick scheme, I’m sure a lot more people would make good money doing it.. it’s all about expectations.

One major thing has changed since the last update.

We started working less and thinking more. By that I mean more time doing productive work and less time sitting in front of the computer checking emails, skyping, refreshing stats etc. This has turned out to be a huge game changer. Plus it gives much more time for reading.

As we both are interested in social psychology and the why’s behind everything, we realize that in many cases in affiliate marketing the why seems irrelevant or at least not worth the time researching it. We just need to look at the stats, realize that something isn’t working, cut and leave it at that.

Of course everyone has to figure out their own way of working but how much do you guys usually research when something isn’t working? I’m particularly thinking of external things such as how a traffic source interacts. To give an example, we were duplicating a campaign to significantly lower and higher bids ($3-4 CPM difference on each side of the original camp, as well as one a few cents lower to grab more traffic. What happened was that the camp with the highest bid converted (but isn’t profitable) and the rest of them stopped converting. Might just be one of those quirks with a traffic source that you get to know after sticking with it for a while???

---

Home > Mobile > Follow-along Campaigns