Welcome to Curious Business
Every Friday, I post a small insight into running Curio City and/or Blue Hills Editorial Services. My most recent posts are directly below. You can also start with the first post, or use the subject labels to the right to home in on particular topics. Feel free to comment on anything that interests you.
Friday, October 18, 2013
The Marketing Wizard Named Oz
This is going to be much longer than my usual post. Get yourself an adult beverage and put your feet up.
The guy who makes Corn-n-Tater bags (let’s call him “Steve”) is going to be on the Dr. Oz show on November 13. This TV doctor reportedly draws 168 million daytime viewers. I’m told that 50-60,000 people will try to order bags within four hours of the show’s airing. Maybe I’m off base, but I’m going to assume that most daytime TV viewers will order via telephone. But even if 20% of viewers prefer to buy online (purely a guess), Curio City could be slammed with 10-12,000 orders within those four hours. And demand will remain heavy for at least a few days, if not weeks, after the crush.
For perspective, remember that it took me eight years to reach 10,000 sales. We’re literally talking about doing 10 years’ worth of business in a day.
Order taking and order fulfillment are both epic challenges. I have about three weeks to brace for this and I’m already getting elbow-deep in Christmas. Bear with me as I work through this. Leave a comment if you see any logic errors or have any insights.
MDD Hosting has been rock-solid. I pay $9.78 per month for their Intermediate shared server plan. Their website says that 1 GB of bandwidth is equal to over 100,000 hits; cPanel says that I typically use about 6 out of 720 available GB per month. Raw bandwidth isn’t a problem. However, I’m limited to 25 concurrent MySQL connections. That means that only 25 people can navigate my site at any given moment – perfectly adequate under normal circumstances, but a serious bottleneck under the Oz scenario.
Google Analytics says that my average visitor spends 1:21 minutes looking at 2.6 pages, but only 2% of them actually buy anything. The Oz crowd is pre-sold so their conversion rate should approach 100%. If a real buyer takes 10 minutes (purely a guess) I could handle 25 simultaneous shoppers six times an hour, or 150 transactions per hour (not allowing for the normal Christmas shoppers who wander in). If 10,000 people are trying to place orders in four hours, that’s 2,500 per hour. The difference between 150 and 2,500 equals a lot of frustrated shoppers. Put another way, it would take my website 66 hours working at full capacity 24 hours a day to process 10,000 sales. And there’s an outside chance that I might get even more traffic than that.
Upgrading to semi-dedicated hosting doubles the connection limit to 50 at a cost of $50 per month. While doubling my capacity should cut my backlog problem in half, it doesn’t solve it and probably isn’t worth doing.
The next step up is a virtual private server. That runs a whopping $124.50 per month, but it gets rid of the connection limit and might solve the bottleneck. The Corn-n-Tater windfall would easily pay for a month or two of expensive hosting, and then I could revert back to my shared plan.
Other costs are involved besides the monthly rent. I’d need to hire my developer to install Sunshop on the new server. If he is even available on such short notice, that’s going to cost at least a couple hundred bucks. I’d have to either reissue or reinstall my SSL certificate, I’m not sure which. Google would punish me for having a new IP address (I’m currently paying extra for a dedicated IP; can they reassign it to a new machine? I don’t think so but I don’t understand these things). Then I’d have to pay my developer again to undo this and revert to cheap hosting; I obviously can’t do that in December, so I’d be stuck with the expensive hosting for at least two months. My best guess at the total cost to move to an upgraded server and back is about $750-900.
All of this assumes that Sunshop itself is up to that kind of load. I have no information on which to base or question that assumption. I might go through all the trouble and expense of moving only to find out that my little shopping cart still can’t handle the traffic.
SUMMARY: Taking orders is already possible, but at an unacceptably slow rate. Upgrading to overcome the bottleneck is expensive, but not prohibitive. Securing my developer’s support on such short notice might or might not be doable. I have no way of stress-testing the new installation before the big event, and I don’t have the expertise or time to troubleshoot if something goes wrong. Changing my IP address a month before Christmas would definitely pooch my normal holiday traffic; Google hates that. After all that risk and expense, it’s possible that Curio City won’t perform any better than the Obamacare exchanges did on their debut.
CONCLUSION: Risky and expensive, but not impossible.
Another option just occurred to me: Buy a separate URL and build a second store just for the corn bags! Running a second business would be a lot more expensive and trickier to set up (a new Sunshop license, new merchant services account, new bank account, new email addresses, new PayPal account, new SSL cert., etc.) but easier to run once it was implemented, and it would not put Curio City itself at risk. I could just shut it down when the rush is over. No time to dive into that idea today but I’ll explore it later if Steve is interested (see the Bottom Line below).
Suppose that I solve order-taking and find myself with 10-12,000 orders in a short amount of time. Yay! Now what?
One hundred corn bags fill an average-sized shipping carton. Ten thousand bags would be 100 boxes. The manager of the UPS Store that handles my receiving would freak right out, even if they were broken into four or five shipments. Plus I’d have to schlep them home in my Miata or Anne’s Fit, find someplace to store them…and then break them into 10,000 outgoing envelopes.
Have I mentioned that I’ll be in the thick of my Christmas business by 11/13? There is no way that is going to happen at home.
I would need a commercial space with a loading dock and a one- or two-month lease. There are plenty of empty storefronts nearby and our local real estate baron might be willing to deal. I’d need heat, electricity, and Internet access. Just for the sake of moving this to the next step, let’s make two huge assumptions and say that I could find such a space in the next three weeks, and that I could get it for $1,000 a month.
Have I mentioned that I’ll be in the thick of my Christmas business? I’ll be working flat-out by mid November just to keep up with normal business. I could spare a few hours to oversee a remote operation, assuming it’s not too remote, but there’s no way I can fill those orders myself. I would need to hire somebody. I don’t even know what all is involved in bringing a temporary employee on board; I’m sure there’s lots of paper to process. My wife believes she could find somebody who’d work for $10 an hour under the table; I’m deeply uncomfortable with doing something so illegal so openly. But to get to the next step, let’s make a couple of more big assumptions: That I can find somebody who’s trainable and willing to work for $10/hour in the next three weeks, and that this person (or combination of people) will put in 40 hours a week. That’s $400 per week in wages, plus approximately 12% in payroll taxes if I follow the law. And if my CPA is reading this: Relax, Ron, I'd keep it legal.
If s/he’s doing nothing else, this person can realistically print one shipping label and seal one package every two minutes. That’s 30 packages an hour or 240 per day or 1200 per five-day week. At that rate it would take 10 weeks to handle the expected volume. Steve plans to advertise 6-8 week fulfillment. I would need two minions for five weeks just to meet their guarantee. Labor is now up to $896 per week or $4,480 for five weeks. (Now I’m going to need a second month on that lease.)
These minions are going to need two computers, two Dymo label printers ($280 each), five or six 10-packs of labels at $326 each, and 10,000 padded mailing envelopes ($4,000). That’s about $6,500 worth of supplies plus whatever the computers cost.
So for a first pass figure $2,000 rent + $4,500 labor + $6,500 known supplies, plus the cost of two computers. I could buy myself the new laptop that was already planning to get in January anyway; figure $800 for that. My wife thinks she can scare up a crippled freebee for the second one. Total anticipated cost of fulfillment = $13,800 – round it up to $14,000.
Have I mentioned that I would have to make this all happen in three weeks while the Christmas season is gearing up?
SUMMARY: Fulfillment is complicated and expensive, time is short, and there are way too many large assumptions above. It will fail if anything goes wrong. There’s a reason Homecooked Shortcuts is outsourcing their fulfillment. I might be able to outsource fulfillment, too; there are lots of companies that do that. But I don’t have time to look into that today. In fact, the fulfillment industry is such a hive of scum and villainy that I wouldn’t have time to research and implement that even if I decided right now to go for it.
CONCLUSION: There is no freaking way I’m going to handle fulfillment. I don’t think I could set this up even if I could work on it fulltime for the next three weeks.
For perspective, Steve would give me his lowest price on these bags, so my markup would be a very healthy $9.50 per bag. Gross sales for 10,000 bags are $139,900 and I’d net $95,000. Payment processors take 3.25% off the top line, or $4,500. So $95,000 minus $14,000 fulfillment costs, $4,500 in bank fees, and maybe $1,500 in order processing costs still leaves an astonishing $75,000 profit. That’s lots of margin for error.
If I had twice as much lead time, and if it weren’t Christmas, I’d be all over that.
I have until Monday to come up with a proposal. Steve already plans to outsource telephone fulfillment and won’t hold it against me if I bow out of this gig entirely…which is the way I’m leaning. Even if I surrender my official online distributorship and sever my link to the Homecooked Shortcuts website, I’ll pick up some stragglers who see my ads when they google the product. That spillover alone could be hundreds of extra orders and might be enough to push this Christmas over the top. Steve says he’ll restore our relationship after business settles back down to normal. I’m not sure I believe that, but his product hasn’t been a huge success for me so far anyway.
BOTTOM LINE: When I call him on Monday, I’ll ask Steve if there’s any value in my taking online orders but not fulfilling them. I think I could swing that, but I’m not sure how much value to him or profit for me there’d be in doing it. This discussion is tabled until I find out.
Incidentally, Dr. Oz gives a vendor one free appearance to demonstrate the show’s immense selling power. Vendors can then make encore appearances by cutting Dr Oz in on the take. If Homecooked Shortcuts sells as many tater bags as they expect, they’ll probably repeat their performance. Maybe I could get in on an encore, with more lead time and no Christmas.
I’m already ordering untested new merchandise that I can’t pay for based on faith that Christmas bucks will materialize in time to cover late November’s bill. I’m not normally a faithful person, but this faith is based on logic and experience. I go through it every year…always with the same trepidation.
Last week was the busiest of the year. This week was merely average. Republican debt ceiling brinksmanship probably scared some customers away. It certainly frightened me into holding my Christmas orders until I was sure that the tea party extremists wouldn’t bring the economy to its knees. I’ll get my holiday stuff a little later than I’d like, and I’ll lose a few early sales. It could have been much worse.
I had expected to lose more ground in October, so even having a shot at matching LY is an achievement. On Monday I’ll get my first halfway decent paycheck since last spring.