• http://www.buzzmaven.com/ Scott Clark

    Timely article! Are there any specific thoughts on methods of deploying RLSAs across multiple domains (same client-owner)…. We’re looking at the subject-area-blog approach, with audience definitions spanning domains, not just site sections.

  • Sam Owen

    Hi Scott,

    You can do some pretty cool stuff with that if you have a tailored list of remarketing tags you put across domains. For example if you have 3 domains that are PPC Blogs, and 2 that are SEO blogs, you could create master SEO blog readers/PPC blog readers lists in your AdWords account and then bid up your PPC Conference search ads just to the PPC Blog readers etc.

    Thanks for reading!

  • neotrope

    Great information, thanks. And, basically: “wow” :-)

  • Ronnie’s Mustache

    Very nice!

    Setting up lists based on the visitor on-site activity has worked for all of our clients.

  • http://pennystockfrontier.blogspot.com/ Dave

    Hi Sam,

    Great article – just a quick question for you: you mentioned that by adding a remarketing list to an existing campaign and having it set to “bid only” (without a higher bid adjuster) will allow you to pull in stats without inferring anything – where would I be able to see the difference in these stats on AdWords? Thanks so much!

  • d_a_t

    In the data on your best results case study, it looks like average position was much better for RLSA but with a significantly lower CPC than the ‘search total’ data. Is this because the ‘search total’ contained data for keywords not launched on RLSA?

  • Sam Owen

    Hi Dave,

    You should be able to see those stats in the Audiences tab in AdWords once you add remarketing lists. You won’t be able to directly compare keywords on RLSA alongside not on RLSA, but you can get a good feel for ad groups/campaign stats with or without it.


  • Sam Owen


  • Sam Owen

    Thanks Ronnie’s Mustache!

    We try and break it down as far as we can before the data becomes too sparse. What kind of on-site activity do you use to base your lists from?


  • Chris Zaharias

    Great article, Sam. EBay
    Labs published an SEM study (Consumer Heterogeneity and Paid Search
    Effectiveness: A Large Scale Field Experiment), in which they proved
    quite convincingly, IMHO, that neither brand nor non-brand paid search
    bring any incremental revenue to eBay at present. Two findings from
    that study have direct bearing how advertisers might think to explore
    RLSA, namely that non-brand keywords (NBKWs) generate no incremental
    sales from eBay customers who either buy frequently (>3X/year) or
    have bought recently (past 90 days). Granted, eBay has become a
    near-perfectly well-known brand – the authors themselves say these
    findings should not be generalized – but since many large advertisers do
    have relatively strong brands, using RLSA to bid *down*
    NBKW’s for segments corresponding to recent and/or frequent customers might possibly result in limited/no loss in sales (the hypothesis being that, as eBay saw, those people get to the advertiser’s site regardless),
    while AdWords spend would go down moderately. This, in turn, would
    allow advertisers to spend on marketing channels whose
    conversions are likely to be more truly incremental.

    Do you have any thoughts or anecdotes on bidding RLSA segments down?

  • Sam Owen

    Hi Chris! Just saw this comment but it’s an interesting one. I hadn’t read that paper before so thanks for pointing it out.

    For bidding things down I tend to focus on websites that have login areas. I’ll create a remarketing list for anyone who gets to /login, /myaccount etc. and exclude that remarketing list from all my Search Campaigns (so a negative RLSA list). You can get clever with this on monthly subscription service sites by excluding users who have logged in less than 30 days ago, but bidding up those who haven’t in an attempt to get them to come back.

    As for generally bidding down repeat website visitors, that isn’t something I do as a rule but it might be effective for some of my clients. Typically I’m not working with huge brands but mid-size enterprises who aren’t necessarily household names and so the findings of the eBay report don’t apply so much.

    Great comment though, really appreciate it!