Trotline Marketing: An Interview with Rob Snell
Spinning Off Multiple Micro-Focus Stores

I started selling old comics in a traditional storefront, Gun Dog Comics, when I was a college student -- first old comics, then new ones, then games, each time trying to grow the business. Finally, when I had saturated the market and run out of space in the store, I figured I needed to clone it. Store #2 was a lot of work and a steep learning curve. But after that, stores 3, 4 and 5 were pretty easy. I ended up with my comic book stores in five real world locations -- Tupelo, Hattiesburg, Jackson, Columbus, and Starkville. I would drive 800 miles a week visiting each store. But then the comics industry started folding up. The Kinko's copy shop in Starkville closed with only 19 days notice. I decided it was time to trade up. So with $500 in my pocket I signed $1/4 million dollars worth of Xerox leases and opened the Copy Cow, the name in keeping with the agricultural university in town.

Back in March 1997, there was a PetSmart opening across the street from my parents' dog supply store and my mom was afraid it would cut into her business. She decided she needed to go on the Internet to diversify her sources of income. My parents had a mail-order catalog company and she was looking to get more catalog requests.

"Nobody is making any money on the Internet," I told her. But thank God for my mom -- she insisted. She made me do it. I did a lot of homework, and after reading an article on how Yahoo! works, I wrote my first Yahoo Directory listing. A few weeks later were getting 100+ catalog requests a day.

But we couldn't afford to mail out all of these catalogs not knowing if they were qualified prospects, so my dad asked if I could build an online catalog. After two attempts of doing some cgi-bin shopping cart programming, my brother found Viaweb.com (now Yahoo! Store). I built a store and it took off like a rocket. The Internet sales at GunDogSupply.com completely changed our family's business dynamic, and I was hooked on the Internet's potential. With the Internet, now I was no longer limited by geography and population, the two largest limiting factors with my retail store.

That fall I developed several online stores including a comic store, a games store, an art supply store, and a collectible card game store, but nothing succeeded like my parents' store did. The next summer, the Pokemon craze hit and people couldn't get Pokemon trading cards fast enough. Based on the five retail stores I owned, I got a huge allotment of cards -- $12,000 worth on the first shipment. But I knew I couldn't sell them in Mississippi stores, since Pokemon hadn't exploded here yet.

That night I built a Yahoo! Store using elements from the manufacturer's website and was listed in Yahoo! Shopping. Then I signed up for a paid Business Express Yahoo! Directory listing which got me into the directory quickly. In June, 1998 we were the only online store with the keyword "Pokemon Cards" in a Yahoo! listing. After we had sold our first allotment in a single week, we had to shut it off until another pallet arrived.

I decided to sell Pokemon cards at normal retail prices, rather than at the crazy prices they were going for on eBay. I worked to build a relationship with my customers, and limited the sale to one box per family or address to increase the number of customers. We also added around 100 other Pokemon-related products to sell while waiting for the next pallet of cards to arrive.

To build my in-house e-mail list, I would also give away one box of cards each month to a new subscriber. I found when I'd do an e-mailing to that list I would experience a $10,000 bump in sales. It was my introduction to the power of e-mail marketing. We sold a quarter of a million dollars of Pokemon cards and other related merchandise those first 18 months.

An online store is NOT a retail store, but "store" is the best metaphor we've come up with. Let me explain.

My retail comic shops had to be general -- everything to everyone -- since each one served a different rural Mississippi college town community and had to cater to a variety of needs. On the Internet it's exactly the opposite. The online stores that seem to do better are super focused on a narrow niche of a business. The Pokemon store had a micro-niche focus. I treat my online stores more like direct marketing ads in magazines than I do as physical retail stores.

I found as I began to build Yahoo! Stores for clients, that a lot of them didn't realize what they're going to be selling a year later. They'd start with pet supplies, but end up selling alligator food -- a very niche market -- and make more money selling to a particular kind of customer. One client began selling shoes, but found a niche market selling extremely large sizes of women's shoes, presumably to male transvestites.

It was from this realization of the power of micro-niche stores that I came up with my concept of trotline marketing.

When I was seven or eight, my dad and I would go fishing for catfish with cane poles and come back with five or six fish. But on our way home we'd go by an old-timer who had caught 20 to 30 mudcats. He had two secrets: first, he knew where the fish were, and second, he had more than one hook in the water at the same time.

The old man had a trotline attached to a tree on the bank, a series of floating milk jugs connected together, each jug weighted with a little bit of sand, spaced five or six feet apart -- each jug with a line and hook attached to it. The old man would bait each of the hooks and then come back the next day to find a mess of fish.

Trotline marketing means spinning off half a dozen "baby" stores from your original online store, each with a particular narrow focus with a segmented market. This isn't just mirroring the same site six or seven times over. Each of these baby stores has a separate focus, like a separate hook in the water -- and some of these hooks will catch fish faster than others. You don't really know until you try. Usually one out of every three baby stores I build for clients hits a home run. Pokemon was this sort of breakout hit, but there are so many things still left to try. This fall, a month before Halloween, I set up a Halloween costume site and was able to generate 90,000 visitors that grossed $32,000 in costume sales from a single, highly focused Yahoo directory listing. I didn't have time to get anything else up and running…

Once you have a store in operation, you find out what is selling and then you a better idea of how to focus. One of the nice features of Yahoo! Stores is that you can study the search terms people used to find the site, and which of these search terms resulted in the most sales. When I go through a year's worth of raw data I may find that nearly all the sales were made to buyers in one to three micro-niches.

By the time you figure that out, you realize that the name of your store is horrible for each of these micro-niches. Another name would be much more effective. So we end up spinning off baby stores. And lots of times they end up being more profitable than the original store that spawned the baby store.

I look at websites to see how this business can help customers to get what they want more effectively. I see so many websites with impediments to buying -- no 800 number, they only take Visa and not other credit cards, etc. I look to see if the store going after the right fish? Are they building a relationship for repeat sales of other products or are they relying on one-hit sales?

I try to help them find micro-niches that aren't too crowded so their store can show up on the first page of the directory. I look to see how many competitors are NOT capitalizing on a particular niche? And I try to see if they are selling products with the right price point?

This September I developed a website for my sister-in-law to sell GPS (Global Positioning System) units. She had a LOT of traffic, but sold only one item in the first six weeks. The price point was too high, though it is the only item that comes up in its category on Yahoo! The hook is too big. The fish can't swallow the $900 hook. It needs a $100 or $300 hook, but the margins on those are so low, we're probably going to punt and move on to something else.

I get my clients to reverse engineer what their costs are on a sale. You need to net at least $20 when packing a box or you shouldn't pack it. I encourage them to get rid of $1 and $2 products or any unprofitable sales for them. Even the very first transaction has to be profitable.

I charge my clients development costs of about $1,500 to $3,500 per store. I encourage them to aim towards $1,500 per store and then beef up those that do well. The more hooks you have in the water, the greater your chances of catching fish.

But if you've got too many jugs on your trotline, you can't keep your bait fresh. You get to the point of diminishing returns and can't provide good customer service. Perhaps my clients will start with five micro-stores to see which hooks the fish are biting on. Then if one goes well, they put all their energy into that one -- and convert the low and non-performers into new experiments.

Yahoo! Stores have great statistical analysis tools. I'll examine the logfiles for the previous year to determine which keywords are driving sales in their store. I also try to determine which keywords that should be working aren't working and look for the differences.

I'll look for 10 top keywords and keyword phrases in terms of traffic and run them through WordTracker (www.wordtracker.com) to get a relative idea of how often a word or subphrase is searched for. Then I'll examine the top 10 keywords in terms of dollar sales in the store and compare that to the first list. If a keyword comes up #1 in traffic, but only #9 in sales, then I try to figure out the discrepancy.

WordTracker is also tied into search suggestion tools and provides a competition search, a ranking based on what folks are searching for, and shows you where the holes are. It's awesome.

Next I go to Overture.com (formerly GoTo.com) to see what competitors are paying for each of these top keywords or keyword phrases so I can determine what is the relative retail value of a keyword phrase. This process helps me find the right bait for a possible micro-store.

"Right bait" means using the keyword phrases that folks who are looking to buy actually use when they are searching. For example, instead of a store that sells all kinds of T-shirts, focus on "Britney Spears T-shirts" or other popular-culture T-shirts. (Be careful of using trademarks in your company names or domain names. Though when you are selling their licensed merchandise rather than trying to exploit a trademark, you are on safer ground.)

I use WordTracker as my base method. WordTracker samples meta crawler search engine data and then extrapolates these figures to approximate what the total Internet searches would be. WordTracker is more accurate than the Overture.com keyword feature, since on Overture.com, some competitors check their competitors' bid prices every 15 minutes, which skews the results.

Google (https://adwords.google.com/AdWords/Welcome.html) helps you test how your ad will do. When you buy advertising from Google, they tell how many searches they had for a specific phrase over a 30-day period.

A search in the Yahoo! Directory will suggest 1 to 9 related phases. For example, if you search for GPS, it'll kick out several categories containing GPS in them.

Inside a Yahoo! Store, you can view "Top 100 searches in Yahoo! Shopping" (under My Store Manager | Statistics | Shopping Searches) to determine what people are searching for who are looking to spend money. (This week "Harry Potter" was #1, followed by "xbox," "dvd," and "turkey thanksgiving cookbook.")

It is best to include important keywords in your company name, domain name, and Yahoo! Store account name. Though your store URL can be your own unique domain name, an alternate URL to get to your Yahoo! store will be http://store.yahoo.com/accountname/pagename.html Having a keyword in the URL path helps in search engine positioning. You have 30 characters there to name your account and I load that up with keywords.

The Yahoo! Directory brings you a ton of traffic and it starts within 7 business days after you get accepted. It used to take forever to get a listing in the Yahoo! Directory. But now you can immediately cash in on a new business idea using a Business Express Listing ($299). You still must play by their rules, and they aren't obligated to accept your proposed listing -- only look at it. You can also have multiple listings in their directories, but no mirror sites where the pages are the same as another listing. Your new site must add value to the directory by having unique content. I always encourage my clients to write articles such as product reviews, or create buyer's guides and comparison charts in order to add value. 

I don't like it as much as I used to -- about half as much -- but the addition of Yahoo! as a PPC partner has really piqued my interest. But Overture makes it very easy way to get a lot of traffic quickly, especially since they partnered with AOL and other major search engines. You have to be careful of keywords that are too generic, since they often don't convert into sales for you.

But Overture does work. I had a client for whom I set up 2,500 or so Overture listings. He inadvertently let his Overture account balance drop to zero, so for three weeks he wasn't getting any Overture traffic. He found that sales dropped precipitously when Overture shut off. Now he realizes the value of this traffic.

I go into the statistics section of a Yahoo! Store, looking for the top 50 referral sites. Then I look at the raw data to determine which search words from those sites that are used to find the store. I download these files, merge them into a huge comma-separated value (CSV) file, and parse the data into a database. This can generate a keyword report of 25,000 keywords people have used.

I take the top 20% of the words that generate 80% of the traffic, usually up to 5,000 or 6,000 words, and bulk-submit them to Overture.com, bidding only a nickel for each word, no matter what #1 bid for the keyword phrase actually is. We might miss out on the high dollar keywords, but get thousands of keywords that others miss. Of the number bulk-submitted, I end up being accepted for about half of the keyword phrases. If you only get a single click through per phrase, per day, think of the hundreds of extra folks you can get tripping across your web site for a nickel a person!

The URL I send Overture visitors to is actually the Yahoo! Store search page that would be generated if someone searched the store for those keywords. Those URLs are predictable. For example, if the store URL is http://store.yahoo.com/ACCOUNTNAME/ then the search results page for the search words "GPS Systems" will be http://search.store.yahoo.com/cgi-bin/nsearch?catalog=ACCOUNTNAME&query=GPS+SYSTEMS. When the bulk submissions are checked for relevance by Overture's staff, they'll see a URL that contains the products mentioned in the keywords. The page itself will also show the 10 most relevant pages in your site for that phrase. By accident we found that this makes all of the Overture.com traffic trackable for sales AND traffic through one reference in your Yahoo! Store stats.

For the Overture description in these bulk descriptions I use something like: "Looking for a keyword phrase? Click here." My description is meta description, but I try show the surfer a benefit of clicking through to our site. It usually takes about 1-1/2 weeks for words submitted through bulk submission to be approved and show up in a search. In addition, we've found that sales increase substantially when visitors are directed in this way to pages where the products are.

The beauty of this approach is that the store very quickly begins to receive a lot of relevant traffic for only 5 cents per click-through. Since many of these keywords show up in the top three listings with Overture, they also show up at the top of Overture's network of partner sites that get a great deal of traffic -- AOL, AltaVista, and now Yahoo! It's easy to see on the traffic logs which buyers came through the bulk Overture buy, since the URL of the landing page is unmistakable.

GoodKeywords.com Pro version (http://softnik.com/products/gkw/gkwpro US $59.95) helps you monitor the position of your keywords. But my favorite is WordTracker. I just learned that Did-It.com has a program called PPCMax (http://did-it.com/services/max.shtml) that looks really cool, too. It has some tracking features so you can automate your bidding and track your conversion rate from your Overture traffic.

FindWhat.com (http://www.findwhat.com) is pretty good, I hear, but I just use Overture. My approach is quick and dirty. You just ram your pickup truck into the trunk of the tree, and the low-hanging fruit falls into the bed of the truck. Then you move on to the next tree.

(Grin.) A lot of people are banging their head against the wall, determined that, "I'm going to make this work no matter what!" My approach is to have twenty different micro-store ideas out there. One will surely pay off big. Throw anything you can against the wall and see what sticks. The more things you try, the more likely you are to find that next Pokemon.

One thing I do is monitor what's hot on the Internet. I search through the top 1,000 searches on WordTracker, the Yahoo! Shopping Weekly Top 100, and things like the Yahoo! Buzz Index (http://buzz.yahoo.com) to see what folks are looking for and then I figure out how can I sell something that would appeal to those folks.

I read a book recently that made the point, "Don't sell what you make, make what sells." The key to good sales is to study what people are looking for, what keywords they are searching on. You ask: "What are people looking to buy?" and then build a micro-store to meet that specific desire.