Probation

Last week I said that “two good days do not a recovery make.” Sure enough, the winning streak sputtered out after four days. That was just enough to cover July’s bills with a little bit left over. Business has turned bad again, but it’s no longer desperately so. I’m just glad Christmas is getting close.

Last month I paid Microsoft AdCenter (Bing and Yahoo) $115 to acquire just seven sales for over $16 apiece. I confess that I usually pay little attention to my Microsoft account because the outlay is comparatively small and the results are meager. But while I’m scraping bottom every dollar has to count. I deleted all keywords with a quality score below five and lowered my bids on the most expensive ones. That should shave $20 or so from my cost with minimal impact on sales. But I’m tempted to shut it down altogether and spend that money on Google Product Search instead. 

You’re on probation, Microsoft.

More Ups & Downs

Fixing the SSL certificate popup last week did not restore normality. I still limped along with one sale per day until Wednesday, when seven sales inexplicably broke the drought. Wednesday alone exceeded all of last week, and five more sales on Thursday reassured me that it wasn’t a fluke. I don’t know why I went 15 consecutive days with sales below $100 when a normal day should be close to $150.

I barely managed to cover July’s vacation-reduced operating expenses. July’s modest inventory purchases are still touch-and-go. It’ll all be OK if normal sales hold up. Two good days do not a recovery make.

***************

Another Sunshop user posted some code in the support forum that makes the realtime shipping modules return “commercial” (online) prices for Priority Mail rather than the full counter price that it has always grabbed before. Since this little price cut brings my Priority prices into line with my actual costs, I finally discontinued Parcel Post rates – something I’ve threatened to do several times over the years, since I never, ever ship via Parcel Post. I also cut my handling fee from 75 cents to 60 cents (padded mailing envelopes, at 55 cents apiece, are my most expensive packaging). I hope people will perceive it as a shipping rate cut, not a hike. But I also realize that virtually nobody will notice the difference.

My last experiment didn’t go so well, but I can easily resurrect Parcel Post if I notice resistance to Priority rates. Since Parcel Post doesn’t have an online rate schedule, Priority will actually be cheaper in many cases.

***************
 

Remember the battle of the payment processors that I told you about a couple of weeks ago? Well, no sooner did I win concessions from my current processor than they came out with “good news”: "…all CDG merchants will be receiving the new CDG360 bundle upgrade for payments, security, and compliance effective September 1, 2012. This powerful upgrade will include a $100,000 breach insurance policy, business-wide risk assessment, customized security alerts, vulnerability scanning and more for an additional cost of only $15/month.”

Yup, it’s a PCI compliance fee – the lack of which had been CDG’s main competitive strength. There go the savings that I extracted two weeks ago. Fortunately, there’s an opt-out procedure. They promised to eventually provide a “HackerSafe” type of site seal, but it’s not part of the initial roll-out. That seal would be worth $15. If the TSA has taught us anything, it’s that the perception of security is paramount.

Coming & Going

I suspected that something was wrong when Sunday became my fourth consecutive day with 0-1 sales. A test transaction uncovered a security warning popup at the checkout screen. The SSL certificate I had paid $50 to renew three weeks prior was never issued and my old one had expired. For three days I was paying to attract people and then scaring them away. It was the first time that MDD Hosting has dropped the ball, and they helped me solve the problem within an hour of discovering it, so I’m more angry with myself for letting it slide for three days than I am at them for causing it. Three days with no cash flowing is serious.

Business stayed in the crapper after security was restored. But at least now it’s normal crap, not extraordinary crap.

The more I watch my realtime Analytics, the more I notice very low-quality traffic. Visitors are coming in on all kinds of weird keywords. A webinar that I sat through on Wednesday gave me some ideas for coping with that. I can delete things that I don’t want to advertise from my Product Listing feed, for example. My first attempt at editing the .txt file in Excel cut the number of SKUs from 552 to 481 but rendered a lot of UPCs invalid; for example, “640346000120” somehow morphed into “6.40346E+11.” I figured that out and edited the damned file again, this time cutting the number of items to 458.
 
I’m probably overthinking this. On its best days, Google Product Search only delivers about 25 clicks. Of course, I could pay more to get more, and it might eventually be worthwhile if I bulldog my way through some of the more complicated optimizations, like segmenting my ad groups and building a good list of excluded keywords. But so far I’m only seeing increased advertising costs without increased business. I’m only persevering because Google promised a 10% rebate on my Product Search spend through 12/31.

Google Knows Best

Some sweat and cussing finally got Google’s Product Listing ads working after I grokked that I had to create an ad group without adding keywords or writing ads -- Google knows best, and I needn’t worry my homely little head over the details. I still don’t quite understand what triggers these ads, where they appear, or what they look like, but the campaign came roaring out of the chute with a click-through rate over 3% (six times the rate of my main AdWords campaign) on Monday. Twenty-three clicks only cost me $0.16 each and brought three sales for an astronomical 13% conversion rate (anything over 2% is pretty good). Sadly, it faded after that debut and I ended up with another lousy week. Traffic is running at 200 visitors a day, but nobody’s buying. I blame the competing demand of back-to-school shopping.

Because Product Listing ads are built from an exported file containing my whole inventory, I get (and pay for) hits from search phrases that I would never choose to buy, such as “beach towels under $5.” I had a few towels left over from my original product assortment, which I marked down to cost years ago. All of a sudden, they sold out last week. With no ad to turn off or keywords to deactivate, Product Listing continued to send me traffic for a nonexistent product. Whereas I only used to update my product feed once a month, I must now upload a new file almost daily as products sell out and reorders arrive.

I’ve been having fun (inasmuch as anything work-related is ever fun) watching the realtime reporting in Google Analytics – that’s how I figured out the “beach towels” thing. I feel like a voyeur. On average, only 2% of my visitors ever buy anything, but watching the other 98% come and go is fascinating. 

What’s weird is that virtually all of my sales come between midnight and 11 am. I have developed this superstition that if there are no sales waiting for me when I fire up my computer in the morning, I’ll go through the whole day with nothing. 

****************

The Squeaky Wheel Gets the Grease: A credit card processing company pitched me an offer good enough to tempt me away from CDG Commerce. This proposal would have cut $40 per month from an expense category that’s consistently over budget. 

The pitch looked like this:
 

(Existing deal from CDG)

2.35% qual (debit, and qualified)
4.01% Non qual (rewards and non qualified)
$0.10 transaction
$10 maintenance fee
$0.25 capture fee

 

(Proposal from competitor)

1.89% Debit
1.99% Qualified
2.47% Mid Qualified
3.47% Non Qualified
$0.10 per transaction
$0 statement or monthly fees
$0.25 capture fee

Superficially, that's a clear winner. But when I started probing I found a $10 monthly gateway fee (my current gateway is free) and a $96 annual PCI compliance fee (I currently don’t pay one). Moreover, they changed the original four-tier rate offer to a slightly weaker three-tier structure (1.99, 2.35, and 3.37%) in the contract.

Because I loathe perturbing a system that’s working and because CDG has been decent enough, I asked them to make a counteroffer. They finally came back with rate reductions to 2.05 qualified and 3.15 non-qualified…and they removed the $0.25 daily batch fee. That saves me a little money without the hassle of jumping ship and ought to bring my processing costs back into line with my budget. Assuming, that is, that I can stay afloat until Christmas season begins in another month.