Are your PPC campaigns eating your organic traffic? Google love for you to bid on your brand name in AdWords, but how to you know if you are wasting money on traffic you could have got for free? Of course, to find this out you need to run some sort of test.
I don’t know how you could run a good A/B split test for this because I don’t know how to randomly segment people into the “see’s PPC ad” and “does not see PPC ad” buckets. Without random segmentation our test will not be fair but hopefully we can get a test that is “fair enough”.
- Use AdWords ad scheduling to set time periods where your brand adverts run and do not run
Be aware of that you can’t split a day into more than 6 segments. You should also think about if your traffic changes much depending on the day of the week. - Wait. Be the surfer. Get ready to ride the gnarly wave of data.
- Use Google Analytics to get the data. I’ve done a post in the past about how to segment by time .
- Draw some graphs (call them “data visualisatons” if you want to impress) and calculate your averages.
Bar charts will give you the info you need; I’ve used stacked bar charts to show a bit more detail about how PPC has performed.
We are interested in the incremental value of adding PPC which is the difference in value between the “No PPC” and “With PPC” averages. This is the figure which you need to compare with the cost of your PPC to decided if it is worth running or not.
The idea of incremental value is important here. Just looking at the value of the PPC traffic will almost certainly give you a positive answer because brand clicks are so cheap.
Ha! That is a truly disturbing photo Richard!
Useful trick. I’ve always run campaigns on our brand, now I can see if it’s worth the price :)
Man I don’t know what is best. the pic you chose or the post..lol NICE
It was actually the photo that kept me reading. Good post.
I’m not approving the photo but the concept is very true, in campaigns ive worked on in the past it has be sometimes around 20-30% of my brand traffic, and usually conversion based terms that have been taking by the cannibalized by the PPC agency.
The problem is that the client doesn’t understand and if the PPC is converting slightly higher and the average CPC is not too high it can be accepted. The problem is where the PPC rate paid for brand traffic is not sustainable and can often be the desperate actions of a PPC agency struggling to continuous provide campaign improvements as they promised to the client in their initial pitch.
As for giving a boost, i have certainly seen this but without the direct correlation of data as Google offers some of their bigger advertisers you maybe wasting time arguing with a client. If you decrease or increase PPC it seems to benefit all site traffic.
PPC agencies can benefit from the knowledge gained by the web analyst and it shouldnt be a race to the bottom…
Gruesome picture! But interesting read.
If Pay per click Advertising is to be a cash generator, then one must decide where it makes sense to stop. Maybe one can try experimenting with different efficiency targets to evaluate the ROI for each increment.
Although I could do a time segments I wouldn´t know what data to believe since my traffic varies too much from day to day and from hour to hour. On week monday might be busy and next week it´s the most quiet day.
Hi!
That’s a very good idea to test the “Branding Effect” in PPC; My experiments have shown that there ARE User only clicking on the brand text ads and are not clicking on the organic results…
@ulf yes i’ve seen that on a number of campaigns, but examine the conversion rates you will likely find that with the correct message its higher. Also look at bidding for lower positions…
The reason you usually get users only clicking on your brand ads is that your quality score is so much higher than the competitors you are shown at the top just below the search box making it look like the first result, so play with preferred placements :)
@Tommi You make a very good point – my methodology relies on being able to segment traffic into roughly ‘equal’ buckets. Without the ability to do this I’m not sure what you should do.
@Ulf The issue I am trying to address here is whether or not users would click the organic results if the ads weren’t there. That is what you need to know to decide if PPC is delivering useful *incremental* value. As @David says quality score is very high for brand ads to only looking at the PPC data will show a successful campaign even if it is not delivering incremental value
@David I have never heard of anyone using position preference bidding with brand terms. Generally I am against position preference bidding. Is this something you have done yourself or just an idea you’ve had recently?
@Richard yes ive had to for one client, it can depend on the strength of your brand and how many potential competitors Google decides to show. The higher the traffic the more choices in placements, as the clients brand grew it it proved easier but initially it just didn’t work.
It was much of a choice, it usually was between 1.5 and 2.5 but it kept the client happy and also shaved some % of the average cpc for the brand terms which gives you an even better ROI.
The caveat is that if your competitor is wanting to be in that position and your brand terms leads to sales/leads its sometimes best to be #1 in paid and organic and accept the costs.
@David So why did you use position preference rather than just reducing the CPC?
@richard the main incremental value i had experience with, were the adwords sitelinks: there it’s up to me to decide what to offer the user, not up to google..
advertising special promotions there raises CTRs and CVRs!
@Richard have used a combination depending if client is running a manual campaign or using a bid management platform, as sometimes according to the reports run later Google ignores both…
As you know reducing bid rate can do a similar job, but currently i have one test campaign that Google seems to be going out of its way to break this technique, bid too low, not shown on first page increase bids get top positions.
@David Cheers. Sounds like an interesting tactic. Thanks for the insight.