Here at Ionic we do things a bit differently from most agencies: hallway debates over Google Analytics tagging methods, discussions on iframe implementations, and the occasional ruby versus python rant. As an example, for the past several years we’ve been using Marin Software’s excellent paid search management software. It was robust, provided exceptional slicing & dicing capabilities, and performed great on larger accounts.
Problem was Marin wasn’t the best solution for *all* accounts – just certain ones. Never satisfied with partial solutions, we decided to create our own! Meet MARS.
For the past four months our hearty dev team Art, Tim & Raffy have been cranking python code to create what we believe is a better solution: MARS. MARS provides our clients with unique capabilities:
· The same level of slicing, dicing and control as Marin,
· With a bidding algorithm that we have complete transparency and control over,
· And the ability to factor and model historical performance trends
I’ll write more about some of the features we’ve built into MARS, but I wanted to highlight one of the most unique patent-pending approaches to bidding we’ve encountered: time windows.
The idea behind time windows is that performance over 75 clicks may be a good indicator of performance, or it may not (depending on the number of clicks per day). So we wanted to not just look at performance over n clicks, or y days…we wanted to be able to do any timeframe, and compare that performance with another timeframe. So for any account, our paid search team can set multiple windows to view and measure performance:
· a 3 day window that catches current trends and can rapidly respond to current market conditions
· a 14 day window that measures a wider segment of performance and allows more data to be factored into trends
· and a 28 day window that gives us a wide view of how a keyword or campaign has trended over an entire month
Each of the windows can be completely customized (highly volatile accounts can have shorter windows, while more slow-building accounts can take a wider view). And each window can be weighted in the final recommendation, so we can weight short-term performance higher for one client, and weight longer-term performance for another. For even more control, we can have multiple rules working within the same account, so high volume core brand keywords can get rapid response, while long-tail slow building keywords can take a very long view to performance.
This gives us an ability to create a bidding algorithm that custom fits every account situation, from a low spending long-decision-cycle client like a boating company, to a high spend quick-fire account in precious metals. Here at Ionic we never believe in one-size-fits-all approaches, and by creating MARS we now have a robust scalable solution that fits every client’s needs, regardless of spend levels, phase of growth, or volatility.
Now we are all Martians here at Ionic!