Shop Codes PPC experiment statistics revealed

When I cashed a cheque for 17,100 that Google sent me last week, I knew how I was going to spend that extra 100 on the first rainy day. So on Monday, I dusted off the cobwebs of my Google Adwords account and set about creating my first PPC campaign since 12th April 2007.

I’d said to a few of my fellow code peers at a4uexpo and online that all my traffic is via SEO and I no longer take part in PPC as I haven’t seen any evidence to suggest it is beneficial. So I told them that I’d pick a random day and join in and then take an indepth look at the stats.

  1. I was to measure how many visitors I can buy for 100.
  2. How much time 100 will buy me.
  3. How much revenue that 100 investment would make me.

However, I also had to measure the benefit. And so I’ve collated traffic and commission figures for the day previous (Sunday 18th November) and one week back (Monday 12th November).

So here’s the backdrop. Sunday 18th November 2007 was Shop Codes highest traffic day ever, since launch. All this traffic was from SEO with Google contributing to over 1,000 of those visitors.

So to experiment on the following day was not probably the wisest choice as traffic may be on the increase and I may find the best is yet to come. And so it was. Yesterday Google (organic) delivered an extra 168 unique visitors, to make Monday 19th my highest ever.

I created one new campaign in Adwords that was to simply highlight that Shop Codes has 140 exclusive codes and I targeted keywords such as “voucher codes”, “discount codes”, “promotional codes” and “coupon codes”.

The minimum spend I was allowed to pay was 0.25. I set the daily budget to 300 and the maximum cost at 0.30.

As soon as the ad went live, it appeared in the top slot in the main part of the Google page. So within minutes I had maximum exposure.

Then soon after I could see that an affiliate network had started clicking on my advert. During my campaign an IP address of Affiliate Window appeared in my analytics.

The stats say that the first visit was at 15:10:53 for the search term “dixons discount codes”. The green url in the image is the url clicked on. The second url is the landing url. If the url is appended with “?clid” then that shows it is a paid-for link on Google.

The statistics summarise that this visitor stayed on my site for 2 hours 18 minutes and 28 seconds.

It shows they clicked on those adverts that were costing me 0.25 a click each, 26 times in those 138 minutes.

It shows that searches included “dixons discount codes”, “dixsons discount codes”, “shop codes” and “voucher codes”.

Some clicks were within seconds of each other, which suggests the visitor was using their browser back button and re-clicking.

After only a few hours, my campaign was at 95.91 and before closing it down completely I reduced the cost to 0.10 to see if the ad would show at all. It didn’t and so at 18:14 the last paid-for link was clicked and normal SEO traffic resumed.

It was not until this morning that I could assess the success or failure of paying for an extra 380 visitors. Was it a success?

Well, yes and no.

On Monday 19th November 2007 I received my highest days level of commission received this year.

Not surprising when I’ve joust bought 380 visitors but remember my organic traffic is on the up with Google organic sending an extra 250 visitors a day compared to just seven days ago.

OK, if I made the most money ever this year, why wasn’t it a success?

Well, the total commission was only 8 more than what I made on 12th November. That day there was no PPC.

I also made one fewer sale on Monday 19th than I did on Sunday 18th. That day there was no PPC.

So, 95 bought me only 8 worth of extra commission.

What else can I measure? Well, lets take a look at my RSS and newsletter subscriber figures. Did those 380 people contribute to an uplift in sign-ups? No, in the last two days I’ve had 5 new subscribers on both days.

So, I’ve summarised that I don’t receive any material gain from PPC. Yes, I’ll make more money (just) but I’d rather carry on making money without spending 95 a day. Or the 89 it would have been if a network wasn’t so keen on my site.

Blog Widget by LinkWithin

9 Responses to “Shop Codes PPC experiment statistics revealed”

  1. An interesting little experiment!

    I personally haven’t ever done any PPC which has always lead me to wonder if I should be. The fact is though that things are going well as they are so I think I’ll be sticking to my non-PPC policy.

  2. Ray

    After our discussion yesterday, I went and checked on my stats following your alarming screen print showing AWin happy clicking affiliates PPC Ads.

    Sadly I have found the same thing happened to me on a number of occassions also. Personally I think this is wrong on AWins behalf. Adverts have a show URL and therefore, the network should be typing in URL rather than costing us money and screwing around with out stats.

    I understand they may need to click once or twice but my stats mirror yours and personally I find this extremely excessive. Have you mentioned it to AWin yet? I haven’t but will be when I get a minute to gather my evidence.

  3. Hi Ray

    I’m afraid that this was all a case of bad timing. Following activity over the weekend, we had a conversation with Dixons about affiliates’ ads appearing on voucher code terms relating to their brand. This includes things like [dixons discount codes], [dixsons voucher codes], and so on. Over the weekend a few affiliates had begun to appear on these terms, which is a contravention of the Dixons programme terms and conditions. In pretty much every case I’m sure this is down to broadmatching and over-enthusiastic search engines, possibly combined with these affiliates increasing their paid search spend for the big Christmas rush.

    We gave a list of the key terms to a new member of our team to investigate. For each keyword, we asked them to provide the affiliate ID so we could get in touch with these affiliates and work with them on avoiding these terms.

    I asked one of our techies to have a look through our firewall log to find out where this activity took place, and it was all tracked back to this member of staff’s computer. I’m happy to send you over a screenshot of this so you can take a look if you would like. Our member of staff was working through the terms and checking for affiliate IDs.

    One thing I wanted to reassure everyone about was that these clicks were not caused by our brand monitoring tool, Snoopy. We have an arrangement with Google meaning that all activity by Snoopy is completely discounted from billing and account details.

    It was really unfortunate that your first day of paid search coincided with us checking through these terms on Dixons.

    We advise that all new campaigns avoid these key terms for paid search, along the lines of [dixons voucher codes].

    I know Shopcodes does really well in natural listings for this kind of term anyway, and luckily there’s are far fewer broadmatching obstacles in SEO!

    Thanks

    Julia

  4. Interesting post, I use PPC for quite a large proportion of my traffic and do see it turning a profit. However my latest blog post shows Im going into offline marketing as a method so will see how that works. Also need to work on SEO which you seem to be successful at. So keep good work up 🙂

  5. Shocking to see AW activity costing you that much money! I thought multiple clicks from same user/IP were discounted and only billed once by Google, obviously not.
    There’s no reason to click on ads, you can visit the link without actually clicking the ad, it’s not rocket science.

  6. Hi Joe, of the 380 clicks 9 of those were recorded as invalid by Google.

  7. Wow, that’s a shocking amount of activity. I’ve mentioned the adwords preview tool on my blog before. So how’s about all networks start using this?

    https://adwords.google.com/select/AdTargetingPreviewTool

    Or how about this?

    http://www.google.co.uk/search?q=+&adtest=on

    Or how about this?

    The firefox plugin which Gav created.

    http://mycroft.mozdev.org/download.html?name=adwords+preview&category=all&country=all&language=all&submitform=Search&sherlock=yes&opensearch=yes

    Paul

  8. hi ray.

    Sorry about the PPC activity; but i just cannot believe AWin felt the need to click on your ads.

    A tech (in fact my nan) would know that you only have to copy the link location via your right hand clicking the mouse to find out the tracking links. Obviously some affiliates hide/mask the URL, but most don’t and I’m pretty positive your advert would have said ShopCodes.co.uk – so there was now need to click.

    Plus surely the tech can spot what keyword you are bidding on, because they are in bold. if Dixons was not in bold print on google results, then you did not use the keyword dixons. If it was, then rightly you should get a warning to stop this activity.

    I’d ask for your money back. PPC can be very effective if you use the negative matching and Geo etc.; so i’d try again but may be negative match Awin merchants so you dont get the same thing happening.

  9. […] In my last post (PPC experiment statistics revealed) I reported that I didn’t see much benefit in promoting my site via Pay-Per-Click as I wasn’t seeing a clear return on my investment. With PPC spend at 25p per click, it is now clear that it’s not worthwhile, when I make 26p for each visitor. […]