Here’s What Actually Happens When You Launch A New Campaign

Simple_Maze

Lately I’ve been focusing a lot of time and attention on a new ad campaign.

(If you don’t know about me – I make money running ad campaigns on the internet. It’s one of the best things that ever happened to me.)

So I’m running the campaign… and after two weeks of focused effort – it finally had it’s first profitable day.

As I look back at how it happened – I laugh to myself. Because to tell you the truth… it was pretty wild, unorganized and hectic.

Let’s be honest.

You’re probably under the impression that every time a guy like me launches a new campaign, it works perfectly.

And while this does happen sometimes (or I wouldn’t be here), I’d be lying if I didn’t tell you that there are plenty of launches that are total chaos and die in a red-hot fire of total failure and lost cash.

So to keep it real – I’m going to give you a sneak peak behind the scenes of one of these chaotic launches to let you see what the ugly process really looks like.

Warning: It aint pretty.

The Beginning

Let me start by saying that I do actually have a pretty solid launch system that I run every campaign through. Most of the time I run it through this process and within a week I’m profitable and can just work on optimizing it to make more money.

So when I launched this campaign I put it through the steps and… two weeks later it was still losing miserably.

Here are some of the things that I was dealing with.

1) The traffic source was sending me 1000s of extra clicks of bot traffic. If you’re not into online advertising this won’t make sense to you – but ‘bot traffic’ basically means fake traffic (robots) that hammers your website but don’t click links or buy anything. This means there is a shady website owner out there that is sending fake traffic and getting paid for it.

When you start a new campaign on a traffic source you might see a lot of this. Some people pause the campaign and run for their lives, I just let it run a few days. Then I start to ‘weed out’ the bad traffic.

I was getting 1000s and 1000s of these bad clicks. I talked to the ad network and they agreed not to charge me for them (thank God) but the clicks were still showing up in my stats. So when I looked at the numbers in Voluum, it was a nightmare. I had to wade through the mess to see what actually going on.

2) I had several offer options and none of them would get on track. I like having multiple versions of an offer to run from multiple networks. It’s always nice to have some backups in case things go nuts. But for this campaign I had 6 different offers rotating and every day – a different offer would do best.

My job is to look for patterns and exploit them. And when a different offer does best every day – there is no pattern… unless you call losing money every day a pattern.

As the first week drug on, and no winner emerged – I got more and more frustrated.

3) My daily test budget was low. When I start a new campaign I keep my test budget low. Most of the time I’m losing 50–100% of the money I spend, so I keep it under control as a form of damage control.

But the safety I get by limiting the budget means I’ll get data slower. Which can be frustrating.

By week 2 – things were still pretty rough but I was close to breaking even. I tried doubling and tripling my daily budget a few times – but every time I did things went wrong. The new wave of clicks would come at an almost total loss.

But I kept going.

4) I tried a completely different angle. I was using a set of ads and a landing page that had worked in the past. It was doing ok but wasn’t working as easily as I’d hoped. So I decided to come at the campaign from a totally different angle. That means new ads, new copy and a new landing page.

5) The new angle gave me confidence to spend more. Now that I had the new angle in there it was easier to spend more per day. I was really interested to see what would happen. Now that I was spending more – the data was coming much faster and I found a few winning ads that started doing really well.

6) The new angle died completely. My excitement for the new angle that came on day one was quickly squashed on day three. The new ads and landers that I was using crashed so badly that I had to pause the whole campaign.

7) The failure gave me the confidence to retry my first angle with a much higher budget. At this point I was used the higher daily spend and I decided to just stick with it. So the next day – instead of spending a couple hundred dollars on the test – I let it rip.

Now, with the new, more aggressive spend and all the new data that came with – I saw some really good things happening.

8) Patterns showed up. Remember, my job is basically looking for profitable patterns inside a massive amount of data and attacking them. And now that I had a big pile of data to dig through, I was able to see a few small patterns of profitability that could make the campaign work.

9) I paused everything that wasn’t’ working. One of the best tips I ever got was from a huge online media buyer who said “when something works – pause everything else.” I paused all the ads that weren’t working and funneled all the traffic through the profitable pattern (ad>lp>offer) that I’d found.

10) I made money. After more than 2 weeks of being all over the place, the campaign finally made money. I had narrowed the campaign down to a working set of ads with the best landing page and the best 2 offers.

And just like that – it was hitting a nice, steady ROI.

< end of story />

What’s The Lesson Here?

If there is a lesson here, it is this: Expect Chaos.

Things don’t always work out as planned. I don’t care what you’re doing, if you expect things to be easy – and then they aren’t… you’re going to be mad as hell and will probably quit. Why? Because you’re not expecting the craziness that will come.

Want to lose 30 pounds? You’ll probably lose a few, then gain a few back, then mess everything up, then binge eat and gain 10 lbs, then start over, then quit, then get focused and then you lose the weight.

The trick is to expect the craziness before you start – so you’ll be more mentally prepared for it when it happens.

My process is something like this:

  1. Attempt to follow a proven system.
  2. See what happens.
  3. If everything falls apart, ride it like a rodeo bull.
  4. If it succeeds, document what happened and try to repeat it when needed.
  5. If it fails, add it to the pile of other failures and move on.

Then, every once in a while…

Write a blog post about it.

  • ✠BOLO

    Damn good read. How did you detect the bot traffic , High ctr and xtra low cr ? Also was the traffic source known for Bot traffic like some or was this the first experience with it. SuitF

    • http://malandarras.com/ Malan Darras

      Thanks BOLO – bot traffic normally = high ad CTR and low (or zero) LP CTR/CVR. Traffic source is known to have bot traffic so I wasn’t surprised. Honestly most traffic sources send bot traffic – the decent ones let you block it.

      • ✠BOLO

        Lots of discipline on your part to give it two weeks also. I’m way 2 A.D.D. I constantly adjust bids and usually scrap it after 3 or 4 days at most. I really need to slow down .My slow loading LP,s are priority at the moment.

    • http://www.benokumura.com Ben O

      one way to detect bot traffic is to setup an alternative placement and track it through another campaign with a 1×1 button or link and see if the bots click through it. You can then look through the stats and see which site/placement id’s to block. [edit from above]

  • Derrick

    get started, keep going and never give up…… Thank you once again Malan

  • Tim L.

    Thanks for sharing the inside workings of one of your campaigns. I’ve been away from AM for a while to build my budget up again and this was just what I needed to light some fire under my ass and give ‘er!

  • Alexis Avellan

    Great post Malan.

    As far as bot traffic is concerned; do you see it come mostly from a particular type of traffic source or can it be any traffic source? I can’t imagine it coming from Facebook.

    • http://malandarras.com/ Malan Darras

      any source, even FB.

  • Drako Blackwell

    sex, drugs, rock and roll!

    • http://malandarras.com/ Malan Darras

      okay then

  • http://www.benokumura.com Ben O

    [replied below]

  • Mdm.C

    Thanks – this is more helpful then reading the morning news! I am going to finish my thesis now <3

    • http://malandarras.com/ Malan Darras

      the morning news is BS, don’t read it.

  • Prakash

    Great Post Malan.

  • CreationNation

    Awesome. Was this an offer your AM recommended that was working for others, or were you just “going for it”?

    • http://malandarras.com/ Malan Darras

      i always work on recommendations. never ‘going for it’

  • Chris

    Thanks Malan, very helpful post!!

  • Camden Forrey

    Hey Malan,

    Loved the article, had a quick question if you don’t mind:

    1) How should I prioritize which parts of a funnel to split test first? My thought would be to start at the front of the funnel to optimize CTRs, then optimize headlines/main images, and finally optimize offers. Any other insight or advice about which order to tackle these different leverage points?

    • http://malandarras.com/ Malan Darras

      offers first. then about the same.

  • http://fitaffiliate.com/ Andrew D

    The beauty of being persistent .. great post once again sir

    • http://malandarras.com/ Malan Darras

      thanks Andrew

  • majdi105

    What you use for traffic ?

    • http://malandarras.com/ Malan Darras

      the internet

      • Bret Royster

        haha

      • Jon G

        Boom!

  • Yassine

    Great Post Malan
    how do you know it’s a bot traffic ??
    is there any specific tip to know about it??

    • http://malandarras.com/ Malan Darras

      already answered. look for BOLO’s comment below

  • Carlos

    This sounds like something I’m going through right now. Thanks for the post, it was def helpful.

  • Francisco Bernabe

    I got from your posts and vids that in order to get a representative data set a budget of around $100/day is the ideal. However, I am a little confused about the daily 5X or 10X testing budget rule.
    Can you please clarify this?
    Thank in advance! 🙂

    • Astra Sleepz

      The 5X or 10X rule is very vague. You can’t judge a winning campaign based on this rule unless you are trying to test a fresh new offers and want to see whether it has potential to perform well or not with very low cost testing budget. If you know that an offer is hot and you still see 0 conversion after 5X-10X spending, it can mean that your ads suck or something else is missing in the equation. I try not to worry about this much, just run it for 24 hours or even 24 hours for a week and see the pattern and decide from there.