The Gentle Breezes of Change

The excitement is building -- at least, I'm excited -- as Sunshop 4 is finally nearing release...slowly, ever so slowly. It’s been in Beta forever. Absorbing this major software upgrade and implementing its new features is my first major 2007 milestone. I had hoped it would be done a month ago, but of course I have no control over this. Sometimes the winds of change are a gentle tropical zephyr.

Once Sunshop’s digested, I can seek a new developer to implement a few of the features that I’ve had on the drawing board since Day One. I need to follow up on tentative expressions of interest from a couple of contractors last year, or seek another alternative.

Meeting a web developer might be the best thing that came out of last week’s Boston Gift Show (which was otherwise a showcase of banality, as usual). I don’t know if I can afford their services, or if they’re the right outfit to implement what I want. The woman with whom I spoke was strictly focused on HTML pages for search engine optimization. I prefer to stay within my PHP engine as much as possible. But my mind can be changed by the right professional advice.

This Sunday I will send out one last issue of my newsletter via Sunshop. Then I’m going to open a trial Constant Contact account. I’ll learn how to use it by creating some press releases for my small media contact list. Once I’ve got that figured out, I’ll upgrade the Curio City Chronicle to a professional-looking HTML newsletter. I have over 400 subscribers now – about twice as many as I’d have guessed. If upgrading the newsletter increases its results from 1-3 sales per issue to 3-6, it will more than justify the Constant Contact fees. Half a dozen sales out of a 400-person opt-in mailing seems realistic to me.

I’ve made very little progress toward my goals so far this year. March sales were disappointing, to put it mildly – and downright frightening at one point, as I racked up my longest sales drought ever. As of today, I just barely beat last March’s unimpressive performance by a few bucks. My goal of doubling 2006 sales just isn’t happening. So the prospect of finally making some tangible headway is heartening.

The Next Big Thing

One of Curio City’s main 2006 achievements was market testing different kinds of merchandise and establishing relationships with vendors. The experience I gained is key to realizing this year’s goal of earning a profit.

The oak DayClock was one of the first products I brought in, and it quickly became my #1 bestseller (subsequently nudged out by the phenomenal USB Fan). Its mahogany cousin is in 11^th place. Together, they delivered an astonishing 16% of gross sales in 2006. Their 50% discount with free shipping meets my markup requirement. I ran out of DayClocks briefly during the Christmas season, but restocked aggressively just as the season was ending. They’re such a sure bet that I felt safe investing my precious post-Christmas inventory cash there.

I’ve sold exactly three of them so far this year. I’m not sure why sales dropped so dramatically. The number of online retailers offering them has just about doubled since the last time I checked. Some of the newcomers are discounting. One, in fact, offers a “member’s price” that I can’t begin to touch. The American consumer is all about the lowest possible price, so whenever discounters move into a market, small independents die out. We can’t survive on those thin margins. I’m sure everyone’s familiar with the way Wal-mart has destroyed whole market sectors and business districts, or the way Home Depot has vanquished neighborhood hardware stores, or Borg & Noble's contribution to the death of independent bookselling.

To find out if pricing is indeed my problem, I knocked a couple of bucks off DayClocks for two weeks. I sold exactly one clock. After returning it to full price, I sold one more. This week, I discounted it again and sold one more. Transaction logs at Authorize.net revealed another customer trying unsuccessfully to buy one; their five rejections due to invalid addresss info will cost me $1.50. Altogether, this does not constitute clear evidence that the discounters stole my lunch. Maybe the product itself is dying. Maybe the marketplace is saturated. Maybe it's just seasonal doldrums. I just don’t know.

The Neverlate clock, my #6 bestseller, is another old standby that is now just standing by. It’s no DayClock, but it does bring in enough revenue to be noteworthy. I told you in “The Neverlate Dilemma” that this product’s markup is problematic for me. I was able to restock in January with a small one-time incentive discount, so I bought as aggressively as I could afford…and then didn’t sell a single one. One competitor has consistently undersold me all along, but I should still be moving one every now and then. I tried the DayClock discount experiment on the Neverlate, with similar results. I sold one clock during the two-week sale period, then subsequently sold one at the regular price. Again, the experiment told me nothing. Next week I’ll mark it down again and see what happens.

Thanks to dropshipping, typewriter key jewelry was a low-risk proposition with a decent markup. But the studio gradually put the screws on throughout last year. First the supply of the most popular letters became iffy, and then unavailable entirely. As the product consequently lost appeal, they raised their prices, discontinued some styles, and imposed a minimum order amount for customizations. I haven’t sold a single piece this year. I’m reluctantly looking for a replacement jewelry line now. It will be hard to find something that’s as clever and amenable to online sales – I really like the line. Jewelry is hard to commoditize. (Commodify? Is that a word?)

I don’t have any new product lines comparable to the heavy hitters that I mentioned above. Golf balls do alright, but aren’t spectacular. Ditto Switchables. USB gadgets are moribund at the moment, although the category retains potential. I need to either find the Next Big Thing, or create it with advertising or product placement. And I need to do that soon. March sales are going poorly. Unless I can turn this around pretty soon, I am going to be forced into Possible Future #4 (“A Curious Hobby”).

Well. The Boston Gift Show starts tomorrow. Maybe I’ll luck into the Next Big Thing at the show. I do have a few new products lined up, but with my inventory dollars all tied up in merchandise that mysteriously stopped moving, I can’t do much about them. None of them look like blockbusters anyway. But one never knows. I never would've figured that USB fan for a #1 bestseller. It's all about marketing, isn't it?

Future Subjects:

• Changes in the Wind
• What Sells, and Why

When things don't click

I’ve mentioned pay-per-click (PPC) advertising many times in passing. It’s my single biggest source of traffic and the only advertising I know how to do. I’m not sure that I’ve ever really explained it. Do you even know what I’m talking about? The rest of this post assumes that you understand the basics.

WTF is Google up to?

Around the end of February, at least a third of my long-established ad campaigns suddenly stopped showing. Keywords that I’ve been buying for a year are suddenly coming up “Inactive for Search.” Words that used to cost five or 10 cents per click suddenly have minimums of $1, $5, even $10. I spent two days rewriting ads, redirecting landing pages, working keywords into product descriptions, deleting the marginal ones...and even (grudgingly) raising my bids on a few of the better performers. It was a lot of work, but once my slimmed-down PPC campaign was back on track I noticed that my competitive position had improved on many keywords. Other users must have fallen to the same bloodbath. Maybe Google’s shakeup wasn’t entirely a bad thing. It prompted me to kill off many inefficient keywords and saved me a little money.

Last week I noticed another round of invalid keywords. Take, for example, Fluxx (and Eco-Fluxx). I’ve been selling these inexpensive products for nearly a year. Their ads cost me 5-15 cents per click. Suddenly, the word “fluxx” now wants a minimum 50-cent bid (for a $12 product? Nope.). “Ecofluxx” wants $1.00 per click, and “fast card game,” which once cost 6 cents, now wants $5.00 per click! The ads land directly on the product pages and mention the product in their text – how can I optimize that any further? I rewrote the product descriptions to include all the spelling variations that I’m bidding on (“eco fluxx”, “eco-fluxx”, “eco fluxx”) My ad and my product page both use the phrase “fast card game”. What the hell changed?

The same thing happened to Chrononauts, another card game by the same company. Six of my seven keywords are ineligible. They want 30-50 cents per click to reactivate them. Bullshit.

Another example: A Life with Wine software. Five keywords that were going for 5-12 cents suddenly want between 10 and 50 cents per click. It ain’t going to happen.

Maybe somebody has bid those keywords into the stratosphere. Sometimes Amazon or Ebay comes along and outbids everybody else, regardless of price, just to own the marketplace. But I don’t think that even the discount giants are stupid enough to bid $5 per click on a $12 product. It must be Google’s algorithm. So, rather than delete valid keywords, I’m going to just hang out and see if they come back on their own. Sometimes they do, just as mysteriously as they went invalid in the first place.

Meanwhile, over at Yahoo….

Around the beginning of the year, Yahoo Search Marketing completely overhauled their interface and their keyword ranking/pricing algorithms. I spent about a week wrestling my campaign into the new interface and rewriting ads – deleting at least a third of the words that I’d been buying. With their 10-cent minimum bid, Yahoo has always been more expensive than Google, so I only advertise my more expensive products there. At least the “upgrade” motivated me to streamline my campaign, just as the Google bloodbath did. It’ll save me a little money over the long haul.

Per dollar spent, Google generates more traffic, but Yahoo brings more conversions.

And now, welcome to MSN….

Last week I started a PPC campaign on MSN, who are offering $200 worth of free clicks. I spent numerous hours transplanting the more successful parts of my Google campaign. So far, there isn’t much traffic. Even though my MSN bids are much higher than the other two sites, I’ve only spent about $5 of my $200 credit.

There is never a shortage of things to hate about Microsoft.

Their conversion tracking is seriously munged. Their tracking code requires you to display a little text link tipping people off that they’re being tracked. Neither Google nor Yahoo do that, but OK, fine, whatever. It’s unobtrusive and I don’t care if my customers know that MSN is watching them. Right now MSN boasts 18 conversions from 9 clicks…and I haven’t had a single sale all week! Typically, less than 1 click in 100 converts to a sale, so two sales per click would be staggering. WTF do they consider a conversion? I can’t tell. Maybe it’s every time somebody puts an item in their shopping cart, regardless of whether or not they check out. Yesterday I turned conversion tracking off. Better to get no information than wrong information, and there appears to be no way to define a conversion.

Their pricing system is also opaque. The amount bid doesn’t match the amount charged. The cost per click is apparently based not just on the bid, but on ad quality and landing page relevance. Thanks to the $200 credit, I’m placing some crazy bids on keywords – 25, even 30 cents per click and beyond. I’d never bid that much if I were spending real money. It’s kind of fun to see what happens with these unsustainably high bids. I suspect that many of their users are doing the same thing, trying to use up their $200 in the allotted two months...and driving some serious bid inflation.

The biggest pisser: they keep rejecting keywords for vague reasons. For instance, the words “Switchable” and “Switchables” (yes, they treat plurals as separate keywords) were rejected for “Inappropriate website content.” Their guidelines say that you can’t promote alcohol, and one new Switchable style is a wine bottle. Did that trigger the rejection? Maybe, but my whole series of generic “wine accessory” keywords were approved. Maybe it’s because “Switchable” is a brand name, against which they do have some poorly defined prohibitions. How the hell can I advertise them without using their name? And if that’s the case, why didn’t they disapprove my Vinylux ads? They also have a rule against “adult content,” and they prohibit “illegal drugs.” Do they consider my smoking accessories to be adult content? Did they find my stash jars or Stoner Trivia game? Who the hell knows? “Night light”, “nightlight”, “nite light”, “nitelight”, “night lite”, “nightlite”, and their plural forms were all approved. But “nite lite” and “nitelite” were rejected for “insufficient content”. Figure that one out. Hardly a day goes by without another “keyword changes required” email from Microsoft.

Yes, there is an appeal process, and no, I have not used it. I shouldn’t have to. I don’t care what ridiculous reasons or poor algorithms they have for assessing keywords. This is a test, and they are failing. Unless something dramatic changes (like a rush of sales somehow being obviously attributable to MSN despite their unusable conversion code), I’m outta there after the $200 credit is used up. The whole campaign is just too high maintenance.

And the last straw: You have to navigate their site entirely by internal links. If you use your browser’s Back button, or hit the Backspace key, you go out to the login screen. Sheesh. The site loads too slowly. Keyword management is done from a different screen than keyword result reports. What a mess.

Future Subjects:

• The Next New Thing
• The Sunshop Upgrade

Adventures in Moving

Last week I told you about the Denial of Service attack that sidelined Curio City for a couple of days. I don’t think I’ve pissed anyone off enough to bring down their wrath upon my head. So either my server was a random target of opportunity, or the attacker was after some other site that shares my server. Either way, I quickly accepted MochaHost’s offer to move my site to a different server on Monday. I already planned to lose much of the day to doctor appointments and assorted personal chores and errands. It would be a good time to absorb a little more downtime.

I finally checked in around mid-afternoon. I promptly changed my name server addresses as instructed, and then checked my site. It seemed to work fine. There was even an email from Authorize.net (the payment gateway, see credit card processing) waiting for me. I like those. They mean somebody is giving me money in exchange for merchandise, which is, of course, the whole point of all this. Ordinarily these arrive simultaneously with a Sunshop email that details the order contents. Not this time. Well, it’s not the first time that Sunshop has inexplicably dropped the ball on notification. So I logged into admin to consult my transaction list.

Nothing. No pending transactions.

Hmm. Let’s look at the new customer’s account info.

Nothing. No customer account.

This is worrisome. I log in to my Mocha account and pop open my SQL database, the be-all and end-all of Curio City Online. Surely I can tease the order details right from the source.

Nothing. No transaction with that number, no customer with that name.

OK, now it’s officially panic time. Sunshop is connecting to the payment gateway normally, but not writing to the database! I email this tidbit to Mocha’s support and cc my own developer. Time passes, with no response. If I can’t resolve this before the charge batch settles at midnight, I’ll have to decide whether to void the charge (saving myself the processing costs, and potentially the additional costs of issuing a refund) or be caught holding somebody’s money for an order that I can’t even see. Reluctantly, I e-mail the customer (oh, no…an AOL address) requesting order details. No response. More panicky emails to my support people. Nothing.

Right about now I realize that I’ve only been receiving email sent to my “@krakenenterprises.com” addresses. The more numerous “@curiocityonline.com” boxes aren’t even drawing the usual spam. Sure enough, a test message goes into the ozone.

Another email comes from Authorize.net announcing another invisible sale. I think this is the first time I've ever cringed at a payment notification.

You can imagine the state that I’m in now. More panicky messages to the support guys, asking them to contact me @krakenenterprises. Now things resolve quickly, and the story turns boring. Mocha shows me how to flush my DNS cache and connect to the “new” database, where – sure enough – the orders are safely recorded. From here on in, it’s a matter of cleaning up the messes that I created in my panic. By Tuesday morning, everything was back to normal.

Almost. Yesterday I found another odd little problem relating to my new install. I’m not entirely confident that I’m completely out of the woods now. I suspect that all of this is a preview of what the Sunshop upgrade is going to be like.

Future Subjects:

• Customers Are Fickle
• PPC Advertising Evaluation
• The Sunshop Upgrade

Denial of Statistics

The last week of February (for accounting purposes, February ended on 2/24) almost became the first time that I failed to beat last year (LY). I was $55 short at 7 PM on the last Saturday night of the month, and Saturday is typically the worst day of the week. Then – miracle of miracles – I got three orders in the last few hours, and beat last year by $29!

I am running well ahead of LY. I’m even $569 over plan (which is double LY). That’s partly because I made $400 in January reselling my Google Analytics code (for which I still owe my developer), but mostly because I lucked into a very nice $900 sale to a high school in Tennessee. Without those hefty shots in the arm, I’d be just barely even with last year and about 50% below plan.

While I certainly can’t plan on extraordinary sales like those, I do need to find a way to encourage them. I’m always mindful of the benchmark that I’m setting for next year if $1300 worth of extracurricular revenue doesn’t materialize. Based purely on routine sales, February was a disaster – probably because I didn’t pander to Valentine’s Day.

As of the end of Feb, Quickbooks says I’m $255 in the red. Three major annual expenses – taxes, corporate registration, and tax preparation – are out of the way. My objective for most of the year is to minimize the loss that I bring into December – few retailers show a profit before Thanksgiving. My main challenge is to hold down costs without losing business or stagnating.

March is starting out dismally. Beating last year already looked iffy during this first week. Then, I lost most of Thursday to a denial of service attack. I turned off my PPC ad campaigns last night, which is tantamount to shutting down my site. Traffic immediately plummeted. Friday is usually my peak sales day; unfortunately, after being solved briefly this morning, the DoS attack has resumed. Curio City is still inaccessible and my ads are still suspended. If beating LY is nip-and-tuck this month, making plan is an outright fantasy.

Of course, today is only March 2. It’s a little early to write off the month. But my focus now is on what comes next. I’ve got two free print ads coming out in an upscale San Francisco magazine in April and May -- I don't expect much from them, but at least it's something. Fathers Day set the bar high for this June, so I need something to juice up that month. LY, Fathers Day was the last peak before Thanksgiving. I am going to test a PPC ad account at MSN, who are offering $200 worth of free clicks if I sign up by March 12. At my wife/marketing manager’s behest, I’m going to try another press release for spring/summer products. Finally, I’m going to learn how to use Constant Contact to upgrade my monthly newsletter. The service costs $15 per month. Typically, the Curio City Chronicle brings me one or two sales. If upgrading triples that to 3-6 sales, it will pay for itself.

This all assumes that Mochahost can defeat the DoS attack and get me back in business. I’m helpless to do anything about it myself. This is very frustrating. Ironically, when I recently researched a column on internet security threats, I concluded that DoS is a minor threat to small businesses and largely a thing of the past. Hah.

Future Subjects:

• Moving Market Targets
• PPC Advertising Evaluation
• Changes in the Wind