Good Debt, Bad Debt

I mentioned in Kraken Enterprises Begins that I’m trying to become profitable without taking on debt.

So far, it’s been easy enough to avoid borrowing. My startup cash is dribbling away on schedule. Now, though, I’ve hit an unexpected sales slump just as I need cash to buy Christmas merchandise. Revenue was supposed to increase steadily throughout September. Instead, it's coming in as my worst month yet. My open-to-buy is stubbornly stuck in negative numbers again.

No merchandise, no sales.

Most established stores buy their merchandise on “net” terms. That means that they place an order, and then have a defined amount of time from the ship date (usually 30 days) to pay the bill. Net terms let you sell some of the merchandise before the bill comes due. It’s a nice cash-flow trick for high-volume stores.

Curio City Online has net billing terms with a few vendors, but, like most small accounts, I usually prepay orders via credit card. Disregarding the float between the time the order is placed and the time payment is due (determined by the statement period, rather than the order’s ship date), I effectively have to pay for my stock before I can sell it.

Right now, I’d like to bring in some Halloween and autumn-themed things, but I have no open-to-buy. I do have a credit card with a high limit and no balance. It’s dangerous thinking: I can’t remember the last time I failed to pay off a monthly charge bill. If I start using deficit financing now, my open-to-buy might never return to black numbers again, interest charges will erode my markup, and I’ll be in a death spiral.

When is debt a good thing?

Debt makes sense as an investment in something that will pay off in the long run – a mortgage, education, home improvements, or business expansion. Debt is wrong when it’s used for consumption, or to subsidize routine expenses.

Going into debt for merchandise straddles that line. If I’d borrowed ahead in August in anticipation of a September sales recovery, I’d be deeper in the hole and racking up interest charges right now.

I’m fairly sure that I can service any debt incurred in October with increased November and December sales. If those don’t materialize, I’m sunk anyway. Next week I will repurpose some of my remaining Phase 2 startup cash to buy new inventory, and run up some credit card debt to replenish some depleted items (such as my magnetic globes and winestoppers). You have to spend money to make money, right?

Long-term, I will take out a bank loan to either (a) expand and revamp Curio City Online after Phase 3; or (b) open a physical Curio City store – depending on which way I ultimately go. I need to show a profit before I’m comfortable with doing that, though. I am not going to borrow money for a losing business.

Other Forthcoming Topics:

• The Royal We
• Where Money Comes From
• Long-term Prospects
• Planned features
• O, Canada

(SEO) The 900-Pound Marketing Gorilla

First, please welcome a new Friend of Curio City. FluidPC uses the same Sunshop software as I do, and its owner has encountered many of the same problems that I’m facing. We’ve been talking about the subject of this week’s post: Search Engine Optimization. Check out their high-quality, reasonably priced, customizable gaming computers.

As I mentioned in If You Build It, Will They Come?, I have known from Day One that search engine results have to drive most of my traffic. PPC advertising is too expensive by itself. Making one’s site appear prominently is a dark art that has spawned a whole industry of SEO companies.

Very few of my product pages ever appear the top 1,000 results for keyword searches, if at all.

I know that the problem is Sunshop. I don’t understand the nitty-gritty well enough to do anything about it.

Consider my bestselling DayClocks. If you search the literal string “DayClock Classic” on GoogleRankings.com, my product page ranks #1 on Google and Yahoo (but is invisible to MSN; go figure). Trying just “DayClock” ranks me at #24 on Google, #18 on MSN, and a respectable #4 on Yahoo. The bots are clearly indexing my product title. However – and this is a huge however – searching the keywords embedded in my META tags (“dayclock”, “day clock”, “week clock”, “day of the week clock” and others) returns no results at all. If you look at page source, you’ll see that the keywords appear properly in the header. Why, then, are the bots ignoring them?

I tried to help the bots by creating an XML site map (as well as an HTML version for customers). The free online sitemap generators, and the one commercial program that I tried, all choke on Sunshop’s sloppy URL generation. I screened out some of the garbage results with filters, but even my best effort still produces lots of duplication and missing category information. I’m unsure whether a decent site map will really help my page results or not, but it's all I've got to go on. So I am going to take another shot at hand-editing an XML sitemap.

A post in the Turnkey Support forum indicated that adding a couple of lines to my header file might make the bots see my product META tags. I just edited that file a few days ago, so it’s too early to say whether that made any difference. I hope it is mere coincidence that a bad sales slump began at about the same time.

Sunshop 4 reportedly contains major SEO improvements. I am hoping that the upgrade will alleviate my problem. Unfortunately, it keeps being delayed. Even the Beta won’t be ready for another month. The release version will almost surely come after Christmas.

If the upgrade doesn’t help automagically, I could change the way Sunshop builds my site. Instead of letting the PHP engine build pages on-the-fly, I can make it generate standard HTML pages, which should be bot-friendly. The drawback is that I’d need to regenerate my whole site every time I make any changes. It is probably a huge amount of extra work to maintain the site if I go that route. Another alternative, which some Sunshop users are trying, is to make a main page outside of Sunshop. This kludge would enable me to directly edit the header information that goes to all of the assembled pages. Unfortunately, it wouldn’t allow me to have custom keywords for every product page.

I suppose I could try to work variations on product titles into my product descriptions, and see if the bots index those. I could rewrite my DayClock descriptions to refer to “day clocks” and “dayclocks”. It would look like a style error to readers, but I might have to try it. I don’t know if the bots are reading product descriptions or not.

The last resort is hiring a professional SEO company. The one’s I’ve talked to are all ruinously expensive – $6,000 and up, which is considerably more than half of my entire Phase 3 budget. SEO is also a shady industry that I would rather not engage unless absolutely necessary. A mistake here would probably be fatal.

Solving this problem is absolutely critical. Sales will improve dramatically when my pages start coming up properly in keyword searches. They will never amount to anything if my pages remain invisible. All the listing and submitting and linking that I’ve been doing is ineffective.

I need help. The Sunshop upgrade is the first step. A decent site map is the second. Professional SEO is my last resort. If you have experience with SEO, please leave a comment or shoot me an email.

Other Forthcoming Topics:

The Royal We

Where the Money Came From

Good Debt, Bad Debt

Long-term Prospects

Planned features

O, Canada

The Neverlate Dilemma

Consider the Neverlate clock.

It’s a great item during back-to-school season. But the economics are difficult. Each clock costs $18 plus about $2 freight, if I can afford $960 to reach their 12-case discount (48 clocks). Because my open-to-buy is chronically struggling back up to zero, I have to go for their 3-case minimum instead, and pay $22 apiece.

The Manufacturer’s Suggested Retail Price (MSRP) is $34.99. To me, “.99” price points mean sale items, so I use $34.95. That four-cent difference happens to make me the cheapest vendor on the internet, and lately I have sold a goodly number. Each sale brings in $12.95 ($34.95 retail minus $22 cost). Profit! Right?

I wish it were that simple.

My cost structure requires a 50% average markup. 50 cents of every dollar goes toward replacing merchandise; the other 50 cents covers all of my other costs (or will, when I'm profitable). When I sell a Neverlate, only $17.48 goes back into my OTB. Because the clocks cost me $22, my OTB actually declines by $4.52 (the $22 cost minus the $17.48 infusion) every time I sell one. That’s another way of saying that I take a $4.52 loss.

When sales are broad-based, some of my merchandise that costs less than 50% of retail makes up the difference. But when an item like the Neverlate surges, as it did last month, I have a real problem.

This is a pickle, George. This is a pickle.

I know of five other web retailers selling this thing. Their price points are $34.98, $34.99, $39.95, $39.99, and $44.99. It’s tempting to simply raise my retail. But with two other online stores underselling me, I’d never sell another one. Only bricks-and-mortar stores can get away with price gouging (because in-store consumers can’t easily comparison shop). You can bet that the $44.99 retailer is not selling any clocks online.

When I went to reorder recently, I learned that the supply is limited; soon everyone will run out. I immediately raised my price to $39.95. And as expected, sales immediately stopped. Hooray! I'm no longer losing money on Neverlates.

That's obviously self-defeating.

What to do?

First, I’m going to have to cough up the $960 to get my cost down to $20 apiece. That cuts the loss from $4.52 to $2.52. I’ll have to risk getting stuck with 48 of these things if sales drop off when the back-to-school season ends. Second, I can inflate the weight by a few ounces to nick people a little bit extra on shipping. It goes against my own policy, and I hate doing it; but I’m desperate here. Let’s say that I can recoup another $0.52 that way and reduce the loss to $2 apiece.

By squeezing my budget mercilessly, I increased my OTB percentage from 50 to 50.75%. Now each sale replenishes $17.74 instead of $17.48. The gap is down to $1.74. I can’t figure out how to make that last $1.74 go away. I’ll have to absorb it, and make it up on other products.

After considerable deliberation, I decided that building volume is more important to my business than maximizing revenue per sale. So I dropped my price to $34.99. Now I'm tied for second-lowest price, with the difference being only a penny. The manufacturer sold me two more cases (eight clocks) from his limited supply. That gives me a total of 12 more in stock between today and the time the supply returns to normal in November. If I can sell all 12 between now and then, I’ll assume that the demand is sufficient to justify placing the big 48-piece reorder next time. If the item dies on me despite my loss-leader pricing, then I will instead try to trickle them out at the 39.95 retail.

Footnote: More than half of my recent Neverlate customers have paid an extra $2 for giftwrapping. If only I could get everyone to buy giftwrapping, I’d be all set.

(Update, August 2007: I've restored the original version of this post. It replaces a bowdlerized version that removed all price specifics at the vendor’s request. Since I decided to sever communication with that vendor after he removed my page link -- and especially because the post didn’t make sense when purged of numbers -- I have replaced the censored version.

(Ultimately, I ended up marking my stock down a couple of bucks to clear it out. At this writing, I’m down to four left. Thanks to its already inadequate markup, I just about broke even on this product. That’s too bad; it was a nice clock. I actually bought one of the last copies for myself.)