Applying PPC Methods to SEO

Jun. 16, 2010 | by Rob Green

Recently, I have been working in both the PPC and SEO teams at iCrossing. Batting for both sides in this way has highlighted the traditionally very different ways in which these disciplines are generally analysed and reported on.

It feels like we are getting closer and closer to the point where we will be able to use the same metrics across search as a whole, and the points here are where I think that Paid Search methods can now be applied to Natural Search campaigns.

…..and I want to go beyond simply reporting the ROI or CPA of Natural Search campaign because, although it generally makes you look very good compared to Paid Search, I don’t need to explain how to do that (hopefully).

What you should already be doing

Focusing on Revenue

Any PPC campaign is always going to boil down to how much it is costing to drive a sale, or how much money is being made for every pound that is being spent. Taking this mindset into how you target terms (for instance) in your SEO work is going to drive you the right kind of results.

Splitting Brand vs Non-brand

Brand traffic is arguably navigational in the most part, and non-brand traffic is arguably the valuable new customers that you are trying to acquire. This applies to all of search.

Just like you would in PPC, you should be looking at these SEO visitors very differently. Jumps in brand traffic could be an indicator of offline marketing successes, whereas jumps in non-brand traffic are going to be all about your performance for your core terms.

It’s tempting to say “you can’t really take any credit for brand traffic in SEO, because you will always be #1”, but that isn’t always true. If you are a new company or brand, brand visibility and traffic might be valuable KPIs when you begin SEO activity. In addition to this, if your brand has a generic name like “eggs.co.uk”, you need to be addressing the benefits or disadvantages this may be bringing you.

CTR Testing

The Google Webmaster Tools now gives you CTR data. There is no excuse for not testing the titles and meta descriptions for your key landing pages now. It’s not as quick or easy as it is in PPC, but you can get a good idea of the sort of things you should be including. If you have invested time and money into achieving a good ranking for a competitive term, it seems logical to optimise the result you are showing – a no-brainer in my opinion.

Forecasting

Not the easiest of tasks for a Natural Search campaign, but if you are hoping for significant investment from any part of any business, the chances are that they are going to request some sort of forecast. Paid Search teams have been providing forecasts split by product areas, including seasonality, for a long time, and there is enough data at your fingertips to be able to do this for an SEO campaign.

What you might need to start doing soon

It is not uncommon for metrics such as “contributing clicks” to be measured in Paid Search campaigns, where non-brand clicks earlier in the conversion path that end on a branded click are given some value. Nor is it uncommon to measure the uplift in branded searches during a period of good visibility for high search volume non-branded terms. These are the kind of metrics SEO people should prepare to have to report on.

More and more money is going into search marketing, and there is more and more scrutiny over how that money is being used, as the people who control those budgets become more and more savvy. The people running SEO campaigns now need to provide better forecasts, thoroughly explain their activities, and show the true value of the work that is being carried out. There are a number of lessons we, as SEOs, can learn from our friends in PPC.

Be Sociable, Share!

    Comments (4)

    • Phill Ohren

      Awesome tips. Forecasting is something I have been doing for a while now, It really helps me plan ahead and structure a campaign right from the start.

      The other point I liked is CTR testing, However the thing thats frustrating me at the moment though is that you can't export the data from GWMT easily.... its a pain!!!Jul 29, 2010 09:17 am

    • Jordan McClements

      I agree that justifying spend on SEO is harder to do than it is with PPC, that's why I stick to PPC! :-)Jun 17, 2010 01:24 pm

    • David

      @Rob thanks for the nudge to use GWT for CTR testing of terms ranking high but maybe not getting as many clicks as possible.

      The problem with forecasting is that you don't know what Google is going to change and it can make a huge difference if you have multiple elements such as Google News, Youtube clips and similar widgets also showing in your results.  It is very hard to foresee the massive impact these can have on your CTRs
      Jun 16, 2010 06:37 pm

    • Dan

      Great post,  I also think Multivariate and A/B testing is often overlooked as a great way to improve conversions from natural search, once visitors are on site.

      With Google Website Optimiser being free and very easy to set up there's no excuse for not doing it really.Jun 16, 2010 06:30 pm

     
    Please note: the opinions expressed in this post represent the views of the individual, not necessarily those of iCrossing.

    Post a comment

    SUBSCRIBE