May 28, 2009
Bid management –adjusting keyword-level bids in an effort to land in a certain region of the search results – is monotonous, inefficient, and (perhaps) better handled by machines. That said, it is also a crucial tactic that can be used to improve a mature campaign.
Though my colleague mentions it as a crucial weekly task (on Tuesdays), all too often, I see bid editing and bid maintenance as an ignored annoyance by less-devoted PPC practitioners. However, with proper execution and attention to detail, a test of alternative bidding strategies can bring forward great benefits and drive ROI to new heights.
Here are some of the ways to do it best:
Hold Everything Else Constant.
It should go without saying, but make sure that your test is with ad copy that you’re comfortable holding steady for a bit of time. You want to make sure that you only test the impact of alternative bids on results. For this reason, leave your headlines, ad text, display URLs, and landing pages constant.
Do It Over a Weekly Basis.
I’ve previously discussed how, depending on the product or service, different days of the week perform differently. In order to prevent overreaction to a new bidding strategy caused not by the bid, but by the day of the week, start your test on a given day of the week, and wait until the following week to check in on results.
Track Everything. Everything.
An inventory to consider – bid placed, price paid, ad position, impressions, clicks, click-through rate, conversion rate, total revenue, revenue per sale, cost-per-acquisition, ROI. I have seen an increase in bids show an increase in conversion rate, but with a corresponding drop in average sale that offset the increase in number of sales. (Follow that?) By tracking every metric, you will be sure to see what your ultimate results actually are.
Focus Only on High-traffic Keywords.
Depending on your site and your product line, there are probably just a handful of keywords that are your key drivers of traffic and conversions. These terms, and not the five word (including a verb) long-tail keyword of your dreams, are the ones to test. In order to acquire actionable results, you need actionable data, and this can only be achieved with lots of data. Consider working with your top trafficked term, and move from there.
Challenge Your Own Conventional Wisdom.
It is easy to become complacent with your thinking as a PPC manager, and to draw conclusions too soon. I am generally averse to bidding for top position, due to costs and a belief that the top spot brings in too many impulse clicks for my liking. However, a bid test on a key term with one of my clients showed the bid increase to bring in better traffic (that is, higher conversion rate, bigger-ticket sales) than I had been getting even out of the two or three spot. By the same token, I saw a client’s conversion rate and ROI skyrocket after a drop from mid-position to the bottom of the page. These things happen, and it is up to you as a campaign manager to make sure that you consider all possibilities.
Consider Initiating Tests at the Start of the Month.
For PPC vendors out there, this should be standard operating procedure. Often, clients are in an end-of-the-month review cycle. Initiating a test at the start of the month allows you the opportunity to fail. It is up to you to make up for loss in sales, leads, or registered users during those final three weeks (and, depending on the client relationship, to let the client know that you’re initiating a test in the first place).
Bid management is a tactic so simple, and yet so easy to ignore. I’d love to hear your thoughts on bid maintenance – and your bidding success stories – in the comments.