Many Body Theory

Coming back from SMX London – I decided to write this article on the plane about optimizing keywords versus optimizing the funnel (landing pages and ad creatives). Too many presentations occurred this past week where advertisers and agencies were told to look at their keywords when optimizing their accounts – that keywords are the first step.

I disagree with the philosophy of “keywords keywords keywords”. If my product/service never changes, there are only so many different combinations of keywords that I can upload to my account. Eventually, I’ve exhausted every combination. Sure, over time there might be new competitors that I might want to bid on their trademark terms, or new slangs/variations of my keywords may become popular (from dating to “social dating”) – but for the most part that is just spending an afternoon every few weeks making sure everything is in check.

Eventually, I can only update my bids so many times. You get to the point really fast if you are in a competitive space such as online dating (over 60 advertisers in the US alone on Google) where you can no longer make any dramatic changes to your keyword bids (sure, you might be able to shave off a few pennies over the course of several weeks/months of getting sufficient data – but nothing that a day of work once in awhile can’t handle) – where do you go from here?

I believe first thing first is to optimize either your landing page and/or backend. If all you care about is having a lead filled out on the landing page – then you are killing two birds with one stone – if you are subscription-based (Match, eHarmony, Chemistry) then updating the funnel may or may not be in your control and area of responsibilities (if you are Match with 400 employees – probably not). For us, let us assume that this is some other guy’s problem – and that even though you two should be working side-by-side for whatever reason you aren’t (example: you’re a search marketing agency and your client does not want you poking around with their back-end).

You are left with your landing pages and the creatives that drive users to it. For a website such as Zoosk, we do a lot with the content network (it is no surprise when you hardly see us show up in Google Search). If you work on your creatives, you might be able to boost your click-through-rates from 0.1% to 0.12% which can be great for additionally traffic – but you aren’t freeing up anything to increase any of your bids (which is key if you want to dominate in Google Search). So for us we need to rely heavily on our landing pages (by the way, we must be doing a great job if we spoke at Search Marketing Expo London 2009 )which requires hours upon hours of brainstorming, designing, developing, testing, and measuring.

At Zoosk we have tried multiple methods of converting our users. With our great presence on Facebook (over 8 million monthly actives if I’m not mistaken) we applied what worked on Faceboook to the traditional online dating world. Our process was not effective for users in the dot com space (can’t discuss the numbers too heavily but if you are/were in the online dating space and saw what we were doing – you can tell it needed some optimization).

Needless to say, we are currently using the traditional online dating model for landing pages of getting as much user information as fast as we can and then hope that if the user gets trapped into the funnel they will eventually convert because we believe that our product is unique with interesting features that many other dating sites do not offer their users.

I would like to think that there is a better way to convert my users and that a landing page can help speed up the process of getting users to subscribe and pay for the service. Even if I decrease my conversion rate by 10% but increase the rate of my user acquisition to wanting to subscribe by 15% — I now have a higher cost-per-acquisition for my “free users” because they are more likely to subscribe than before. This higher cost-per-acquisition can mean the difference from us bidding on position 7 in Google Search to maybe 5 or 6 without changing my keywords or ads – changing this much in my position is a lot more than shaving a few pennies off of my keywords during a multiple-month process (which may not even be effective because by then I can enter into varying shifts in seasonality/trends).

Yes, over the next couple of weeks/months instead of changing my keyword bids from $1.20/click to $1.17/click, I will be making dramatic changes to Zoosk’s funnel and allow us to go from $1.20/click to $1.80/click or even higher by changing the cost per acquisition for new users.

As I mentioned before – by doing this I might drop off on how many users actually complete the new user registration form, so once I open up a new target cost-per-acquisition and meet it, I can then optimize my ads and creatives to increase my click-through-rates and compensate for the missing users.

Keywords… I’ll get to those later. If users start to use terms other than “singles” and “dating” – please let me know – but I’m betting that after 10 years of online dating that I can afford taking a 10 week break from keyword research/expansion to focus on something else.

Note: the cost per clicks used in the post above is just arbitrary numbers… Zoosk would kill me if I disclosed our CPCs

This entry was posted by Nicholas Abramovic on Thursday, May 21st, 2009 at 5:37 am and is filed under Search Marketing. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a thought provoking comment, or trackback from your own site.