Category: Case Studies

PPC Case Study #3 – Same Campaign, More Data!

I’m back with another case study. As you may remember from my previous ppc case study, I was working with just Yahoo for that campaign. With this new ppc campaign I instead went with Bing, Yahoo and Bidvertiser because I had a bit more money to drop on this test and wanted more data across the networks.

Note that these numbers are from November 2009 through Jan 2010, and I’m still working on this campaign so I still won’t be telling y’all what I’m actually running with. Look for a new case study coming up in the next few weeks or months. Probably months knowing my slackass style.

For this campaign I ran things a little differently. If you remember from my last post, what I did was have about 600 keywords spread out over a few ads in one ad group. This time though I did it differently and set things with a new method. I went with;

  • 20 ad groups
  • each ad group having 1 keyword
  • 3 ads based around those keywords.

I ran with this formula on Bing and Yahoo, but with bidvertiser I just ran with the highest performing ads from the two networks combined with some simple math to calculate which was the overall best ad.

So, on to the numbers shall we?

BIDVERTISER

With Bidvertiser, as with all my previous attempts with them I had horrible click through’s with relatively high impressions. I know some people love bidvertiser, but I just haven’t figured it out yet. I probably won’t try much more with them till I’m baller like some of you guys and can afford to throw away money 🙂

Bidvertiser Spend: 12.54
Bidvertiser CPC: $0.96
Bidvertiser Impressions: 75,291
Bidvertiser Clicks: 13
Bidvertiser Sales: 0

Bidvertiser Statistics

YAHOO

As with the last campaign, I started out with Yahoo, and then once I started getting traffic and clicks I paused the underperforming groups and migrated the performing ones to Bing and Bidvertiser. I did see one sale, so I made some of my investment back on this one, but not nearly as much as I had hoped, especially with my highly targeted ad groups, and ads per ad group. I am still confident in yahoo though because I had relatively high impressions over this amount of time, however my CPC was way high, I wished it would get down to what BING was doing, but apparently not.

YAHOO Spend: $156.92
YAHOO CPC: $0.71
YAHOO Impressions: 37,193
YAHOO Clicks: 220
YAHOO Sales: 1

Yahoo Statistics

BING

This was hands down my best experience ever with PPC. I had 3 sales getting me around $144 before removing my initial investment of $2.12. This is simply awesome. I don’t know why it dried up but I will be going back and revamping the campaign to hit BING with again because I haven’t seen sales like this ever before, especially with limited impressions and very limited clicks. Fucking awesome.

BING Spend: $2.12
BING CPC: $0.30
BING Impressions: 198
BING Clicks: 7
BING Sales: 3

Bing Statistics

SUMMARY

So let’s total it all up.

Total Spend across all networks: $171.58
Total Impressions across all networks: 112,682
Total Clicks across all networks: 240
Total Sales across all networks: 4
Total Commission amount: $192
Total Profit: $20.42

As you can see, I was profitable. Not very much, but I was profitable. So, this campaign was definitely a win for a PPC newbie like me. If you remember last time, I was still 80$ shy of actually turning a profit, so even $1 would be a win for me at this point.

What have I learned off this one? Bing apparently fucking rocks, or I just hit the right market at the right time – I guess we’ll see on the next test. I do think I’ll drop bidvertiser indefinitely and probably yahoo too since my spend was so high compared to Bing. Yahoo, nothing much, same old shit – I just need to refine my process with them, and probably develop more ads along the way – I just ran out of money to keep testing.

Now on to more testing.

Campaign 2 Case Study Details and Results

As mentioned before I launched and ran my second campaign a little while ago. I’m just now getting around to posting an update about how it went. Which by all standards was much more successful than the last campaign but still not profitable. I won’t be disclosing which niche I was trying because I plan to give it a shot again with what I’ve learned from this attempt.

Let’s get down to it.

I started out with around a $225 budget to run with, and again I was only using YSM for this campaign, I did not branch out with BING or Google or Adbrite this time, but next time I think I will try expanding out to BING to see what kind of results I see from there. YSM is kind of weird, and so far I’m not a very big fan of their interface, so using the ad Center desktop app will probably be a refreshing change.

I ran this campaign back in September for about 11 days (there were two days where my CC wasn’t charged so I wasn’t running the full 9 days though).

As you can see compared to my previous campaign test, which was in a different niche, this one worked much better. However, the CTR was much less – I think this was due to my bid price, and the placement of the ads, which we’ll get to later.

Overall1

I spent $202.88 over the 11 day period, and had nearly 72,000 impressions. Not bad considering I wasn’t using the content network at all and only search. Speaking of this, this is something else I might try next time is using banners and the content network instead of text links and search. We’ll have to see I guess.

This time around, I was using some demographic targeting along with my regular bids.

demographics

So, if I understand this correctly I was bidding a max of $1.00 for any keyword, but if it was female 21-29 I was paying up to $1.24 for each click. I don’t know if this really helps much or not as I’m still learning how the bids work and how to turn everything to my favor, but as you’ll see later I think this was beneficial to this campaign.

I didn’t bother to set any geographic options because the offers I was running were compatible with both the US and Canada.  I was however only marketing to these two countries, but that doesn’t cause an increase in the bid price, only if I wanted to bid higher for the states or Canada.

geotarget

As you can see from the following graph I had about 297 clicks over the campaign period, with an average CPC of $0.68, not bad but I should have been around the $0.28-0.32 to bring the campaign really profitable. I just didn’t have the funds to extend the campaign this long.

Overall2

Now to the good part.

Even though my CTR was much lower with this one and my impressions much higher – I made three sales this time. This means I made $120. I was only $80 (2 sales) away from bringing this campaign profitable. Not bad for my second attempt.

What I did for this campaign was a little different than the last campaign I tested out. This time I created many more ads, and keywords – I also used Market Samurai to get keyword suggestions instead of doing it by hand – saved a lot of time. This time I also used keyword targeting within the ads, which I didn’t do last time and I also had the content network turned off from the start of the campaign.

Overall Campaign Stats.

  • Spend: $202.88
  • Sales: 3
  • Sales Amount: $120.00
  • Clicks: 297
  • Impressions: 71,551
  • Keywords: 588
  • Ads: 12
  • Ads Using Keyword Targeting:  7

Now, this is not entirely accurate because I was watching this campaign like a hawk. Every time Yahoo would update their stats I would login to pause keywords or change the ads around. I ended up with the following changes.

  • 588 Keywords Initially, Paused 210 of them throughout campaign
  • Started with 12 ads, ended with only 1 un-paused.

Why did I only run one ad in the end? Because it was performing really well and it seemed to be getting the most clicks, as well as the cheapest clicks. I probably should have created a few more ads here and ran tests on them, but money ran out so I couldn’t do that with this test. Next time I will.

This time I also ran with 2 landing pages to begin with, then after 100 clicks to each one, I switched around the strongest performing page, and created three separate versions with different offers on each page to gauge offer success. Because I ran out of money sooner rather than later, I was not able to fully test the 3 identical landing pages as much as I wanted, oh well.

What did I learn?

Overall – much more than last time.

This time,  I already had the knowledge of my first campaign so I knew to have the content network off, I also was aware of keyword targeting in the ads this time, which I didn’t know about before and really helped with the clicks and impressions. I also knew to insert a bunch of keywords into the list, instead of just 20 or so – this let the ads show up with things I wouldn’t have thought of. In fact, all three of the sales came in off keywords I never would have expected, or thought of so it just goes to show when testing, insert a shitload of keywords to test with.

As a side effect of having a bunch of keywords, when testing a campaign like this you also are able to take those keywords and use them in article marketing, or blogs and even running ad sense you know you will be getting high click value off those keywords if the ads are targeted appropriately.

Another important thing I learned during this campaign was that Yahoo will go in and charge your credit card every day even if you’re not running ads. They base this on your average spends.  I had to call them up and get them to switch my account to manually insert funds; apparently there is no option to switch this in “my version of the interface” whatever that means.

Overall this has inspired me much more, even though I didn’t obtain a profit, I was much closer and have increased my knowledge two-fold, next I’ll be testing all this out on YSM content network and BING to see what kind of differences I see.

I hope this has inspired you to just to get shit done and try out a campaign.

Second Campaign Launched. Case study to follow.

Just so you, my avid readers aren’t left in the dark I thought I would do a little post outlining what I’ve been up to lately.

So, here goes.

I’ve been setting up a new campaign to test out in a different niche then last time. [ADHD] Speaking of niche, we were always told its knee-sh, not nitch… which is it? [/ADHD]. This time I’m running it a little differently than last time, and arguably with less work. So, let’s dissect what this one is all about aight?

Only trying one traffic source this time instead of two.

I’m not running banner ads and text ads this time. Just text ads.

I also ensured the content network was turned off from the get go on this one, instead of needlessly getting clicks from here when I wasn’t intending to test it. This is a big thing I learned from the last test.

I am running two landing pages. One with only one offer and another with two offers running side by side. The landing pages used this time are drastically different both in tone, intended reader profile and products. Keep in mind this is in the same niche, I’m just marketing differently between both niches.

Instead of writing my own copy this time, I opted to invest $40 USD in order to get a professional copywriter to re-write the copy that I had already written. He did a great job and went over all of it with a fine tooth comb while I spent my time designing. While I do think my copy is pretty rad, sometimes it’s worth it to get a professional to look yours over and provide editing services. This is what I did, and not too much was changed. However, some pretty important things were pointed out to me and modified by the writer I found on the gay webmaster forums. Even with the expense of the copy, if it leads to one sale it’s paid for itself so in my opinion this was well worth it.

This time, I also loaded in 690 keywords, instead of my measly 25-50 that I used last time. I figure this will dramatically increase data that I receive on long tails and other weird things people are searching for.

Another thing I did differently this time around was instead of only 5 ads, mainly directed at people in the “buying” frame of mind I created 12 ads of varying consumer mindsets. I feel this will give me a broader test to go with compared to last time and will give me more opportunity to lessen the CPC on this one and find converting long tail keywords.

I’m also limiting my budget significantly (50%) compared to last time. I lost too much money on my test last time with absolutely no converts so I want to spend less, and aim for more data retrieval and testing on this campaign.

So, that’s pretty much what’s been going on and of course, if I don’t end up making a googol with this campaign I’ll be sharing my results with you in a later campaign follow-up.