Microsoft Enters Witness Protection Program

Three weeks ago I put Microsoft adCenter on probation due to its high cost per conversion ($115 to acquire just seven sales for $16.46 apiece last month). The shame apparently drove Microsoft into the Witness Protection Program. This week they changed their name to Bing Ads, put a nice photo on their front page, and made it difficult to log in; the site still doesn’t work with Firefox, but I eventually got in with IE8 after numerous tries. What I saw was not pretty.

Here’s a snapshot of advertising results for the past 30 days, with a couple of caveats. First, not every click or conversion gets captured; users (like me) who don’t accept third-party tracking cookies are inscrutable to Analytics. Second, not all conversions are equal; some customers buy a single low-value item while others buy multiple pieces of higher price. Third, a click-through rate greater than 1% is satisfactory. A conversion rate of 2% is ideal.

Definitions:

·         Click-through rate = how many times somebody clicked for each time an ad was shown (impressions);

·         Average CPC = cost per click;

·         Conversion rate = how many of those clicks ended with a sale;

·         Cost per conversion = how much I spent to produce each sale.

From best to worst:

Product Listing ads (new)

·         Click-through rate: 3.25%

·         Average CPC: $0.16

·         Conversion rate: 2.99%

·         Cost per conversion: $5.24

Product category ads

·         Click-through rate: 0.94%

·         Average CPC: $0.18

·         Conversion rate: 1.47%

·         Cost per conversion: $12.14

Product-specific ads:

·         Click-through rate: 0.85%

·         Average CPC $0.17

·         Conversion rate: 0.68%

·         Cost per conversion: $25.54

Generic ads

·         Click-through rate: 0.61%

·         Average CPC: $0.17

·         Conversion rate: 0

·         Cost per conversion: n/a

Bing Ads (formerly Microsoft adCenter, formerly lots of other names):

·         Click-through rate: 0.45%

·         Average CPC $0.19

·         Conversion rate: 0.44%

·         Cost per conversion: $43.94

Ear buds were the biggest loser by far: $44.21 spent for zero sales of a $12 product. I have suspended that ad group. I might turn them back on when we’re closer to Christmas, since this is a popular small gift item.

Golf balls were a close second: $38.77 spent for no sales. August is ordinarily one of their strongest months.

Even though I’m only paying 10 cents a click to advertise keyboard stickers, it cost $21.63 to sell one $8.99 set. Bzzt! Suspended.

Lighted caps ran me $20.78 per conversion, but caps sell for $20 apiece and customers often buy more than one. The one keyword phrase that dragged the average way down has been paused. 

Last month’s changes brought the cost of Bing Ads down somewhat, but conversions fell from seven to two. I think Microsoft flunked its probation.

All of this pruning and tuning has an obvious downside: Business is already in the crapper, and shutting down ads risks flushing it away. I don’t know if that’s the way out of my cash flow death spiral or not, but I have to do something, and ad costs are one of the few major expenses that I can control.

To offset all of this scaling back, I increased my default Product Listing bid and I’m letting Google manage the Product-Specific bidding for a month. I figure that their algorithm can’t do much worse than I am doing, even though I know that its function is to maximize Google’s revenue. 

I’ll reevaluate all of these numbers in 30 days. If my Product-Specific results improve, I’ll automate my other two campaigns. If Bing Ads doesn’t improve, I’m going to kill it.