It's Religion All the Way Down

It was a hot day and as Rebecca and I drove north on i70. I was looking straight ahead as I pulled up behind a beat up SUV with a bumper sticker extolling the virtues of Yoga. A hand poked out of the front window and flicked a cigarette.

In the passenger seat Rebecca craned to confirm the cigarette siting, "Is that person smoking?" she asked. "Doesn't seem to fit with their bumper sticker."

"True," I said as the light turned green and I accelerated to pass the SUV, my underpowered Suzuki Aereo struggling to hit 40mpg, "it's religion all the way down."

"What?" Rebecca asked.

"Oh, it's just something I've been thinking about recently. Everyone always says that there's no accounting for taste, but until recently it was hard for me to admit there were chain smoking Yoga enthusiasts. It just seemed too contradictory. The best model I've found for respecting those kinds of differences are to treat them like religions."

"Yoga isn't really a religion," Rebecca said. "And what's with all the way down? Is that a rip on the turtles?"

"Yeah, I just tack on the turtles bit because it's funny."

With the SUV slightly behind us, I quickly changed the topic, "Quick, can you see what the driver looked like? I'm curious what a chain smoking Yoga enthusiast looks like."

Rebecca swiveled around in her seat to look, "I can't see, there's too much reflection in the windshield."

"Darn, I guess it will have to remain a mystery."

Marketing is Just Sales Without Accountability

Over the course of the last 7 years, I've spent a considerable amount of time managing sales and marketing for my own 2 software companies as well as a web consultancy. I've attended conferences, made sales calls, printed flyers, and schmoozed at investor gatherings. And I've come to believe that marketing is just sales without accountability. Bear with me for a moment.

Marketing is Sales

Both marketing and sales are intended to produce the same end result: make more sales. Marketing is the top of the sales conversion funnel; the first interaction with a potential customer in which good salespeople provide value and make a lasting impression. But let's not loose sight of the end goal: that first contact with a potential customer is fundamentally about sales.

Companies don't sponsor trade conferences and staff booths because they want to spend tens of thousands of dollars on snazzy signs and airfare, they want to make more sales. Companies don't maintain blogs because they want employees to become excellent writers, they want to be noticed for their contributions to their niche and make more sales. Companies don't buy television, print, and radio advertisements because they want to support actors and graphic artists, they (you guessed it) want to make more sales.

Marketing Isn't Measured

The problem with marketing is that it's so far towards the top of the sales funnel that it's difficult to measure. Sales calls either result in a sale or not, so it's natural and easy to measure their effectiveness. It's a lot harder to measure the value of meeting someone new at a trade show. Will the person become a paying customer? Will they refer someone to your business? Will they write about what great work you are doing? Or will they trash you to a potential customer? It's just not something that's obvious or easy to measure.

As a result, businesses and business owners (myself included) simply throw up their hands at some point and start generalizing. "We have to get our name out there somehow, and people watch television, so why don't we try a TV ad?" Or perhaps "We know some of our customers attend the BigConf Trade Show, so let's exhibit this year."

The unspoken rule of marketing is that most of it is wasted effort, but for a myriad of reasons, it's too costly to figure out which part is wasteful. My suspicion is that for companies in established markets, what most folks would recognize as marketing probably is worth the time and energy. Most companies can't prove that, or course, but the fact that they are still around means that they are either spending little enough on marketing  that their failures don't sink the ship, or the marketing efforts are actually paying off.

Stop Marketing and Start Selling

If you work for a tiny company or are building a new product, you can't afford to waste resources on unprofitable marketing or sales. You have to start measuring marketing efforts in the same way that you might measure sales. At this point you might be saying "Yeah, well, that's all fine and dandy, but how the heck do you propose I do that?"

Simple: pick some metrics and start recording data.

Here are some examples of things that I've personally tried:

1. Enter people you meet at trade shows in a CRM and track sales as they come in later.
2. Attach a coupon code to print materials and track how many times the coupon is used.
3. Ask incoming sales leads how they heard about you, be bold and ask for a name if the person is willing to give you that information.

You might now be saying "but I'll never be able to measure activity X!" One striking thing I've noticed about sales is that salesmen and sales departments are loath to invest time or money in anything that can't be tracked. That's a good starting point for marketing too. That doesn't mean you can't do activity X. It just means you have to get creative about tracking it. In cases where you can't find any way conceivable way to track it, that's generally a good indicator that it's not a profitable marketing channel.

Conclusion

You might well choose the wrong metrics to measure the first time around. For instance, the first two years that I attended trade shows for my company, I measured booth traffic. I literally just wrote down how many people talked to me at the booth. It turns out that people visiting the booth didn't predict sales at all. But that data saved us more than more than $10,000 when it came time to exhibit in year three. We had hard numbers that suggested that either 1) it didn't make money for us to attend trade shows or 2) unqualified visits to our booth are not related to making sales to qualified leads. Dodged a bullet there!

Marketing doesn't have to be blind, and when it does, don't do it. Treat marketing like sales, measure everything you can, and in a year or two you should be able to make at least one decision that saves your company $10,000.

Sivers is Right: Customer Service is the New Marketing

This post was originally published on 8/21/2011.

Recently I read an old interview with Derek Sivers about his CDBaby customer support. I've always found his writing educational, but in this interview his closing sentence struck me as particularly true: "customer service is the new marketing." That got me thinking about my own experience with my company and how much we labored to do marketing "right."

We started Skritter on a shoestring to teach Chinese and Japanese. We decided that we would either make money and support ourselves, or stop working on the site. After the first year, the site was generating cash, but not very much. Nick, Scott and I decided we needed to hit marketing hard. We partnered with a great guy in China to help us reach more customers and we set about our marketing.

We began with Adwords. We attended several relevant trade shows. We tried promotions and giveaways on the site. We started Tweeting and developed a strategy for Facebook content. We had two marketing interns in China talking to learners on the streets of Shanghai. We were present at Chinese testing centers, we contacted schools to sell site licenses, we did webinars, distributed print ads, setup an affiliate program, spent serious time on SEO, created a content strategy to improve SERP results, collected quotes from marketing consultants and agencies, redesigned the appearance of the site, and conducted A/B tests.

Throughout all of this, we continued to grow, but our growth curve didn't look a hockey stick, it was the gradual upward movement of a well liked company.

We always made it a priority to bake in tracking to our marketing, and after a year with little to show for our effort, we decided to give it a temporary rest.

That week we watched in amazement as the site kept right on growing, despite negligible marketing outlays. We nervously twiddled our thumbs, waiting for the growth numbers to level out and go into free fall, but that never happened. Instead, Skritter continues growing to this day, and we have strong evidence to suggest that it's because of the way we treat customers.

First, we conducted a poll of our user base and found that more than 35% of our customers had personally been in contact with either Nick, Scott, or I. That meant 1 of every 3 paying customers had traded a personal email or met us face-to-face! Considering the size of the site at the time, that was an astonishing number of customer interactions.

Second, even after months of no marketing, growth continued unabated. Moreover, we continued seeing a trend we hadn't paid much attention to earlier: people arriving at the site and telling us (albeit not in an easily quantifiable way) that they were there because of referrals. This was corroborated by a forgotten poll we had done more than a year before which attributed most new inbound customers overwhelmingly to referrals.

So, while we were trying our best to find scalable marketing solutions, all those hours spent in front of keyboards being sociable and helpful to current customers drove site growth. Our customer service has been our marketing, which directly supports Derek Sivers' opinion.

Marketing doesn't always make sense for companies. Where marketing is concerned there are two schools. The first is the "build it and they will come" school of thought. The second is the "most companies needs marketing" school. We clearly followed the latter and although we are just one data point, when we started my second company,  we took Derek's advice to heart and disregarded marketing at least. When you've only got 5 customers, there's a real temptation to focus all your time and effort on finding more customers. But that's missing the forest for the trees.

In our experience, the key to a larger customer base is the person right in front of you, money in hand, waiting to see how you'll treat them.

Games: The Anti-MVP

This was originally posted on 6/6/2013.

My cofounders and I were sitting in the Mission on a sunny morning, eating crepes and talking with a survey and testing guru about testing business assumptions for our new startup.

"I would be hesitant to sink more than a week or two into the idea without data from customers."

My cofounders and I traded sideways glances. The truth was that we had been working on our new startup for 4 months already. We had been play testing our game for more than a month, but  it would be a while before it was ready for a beta launch .

Our meeting left us feeling uncertain, our previous startup had a minimum viable product ready to ship inside a month; why was this taking so long? It didn't hit us until later on when we were speaking with more experienced game devs.

"You have something ready to play in 4 months? That's great."

We had been operating on the assumption that like a B2B startup website, we could throw something down in a week or two, go to customers, test it, and begin iterating. But games defy MVP release schedules. Nowhere is this better illustrated than in 2D Boy's illuminating blog series about developing World of Goo. It's true that they churned out their first version in a week, but their first game tests (which occurred several months into development) were a mess. At CodeCombat we understand that all too well, unfortunately.

Games must be fun. The primary business assumption of any game is "users will like playing it." It's almost guaranteed that users won't be satisfied playing a two week old MVP. In fact, that's sort of the definition of a minimum viable product: it's rough, unpolished, and easy to change. If you're doing it right, the MVP is difficult to take seriously. By contrast, the quality standard for games has been raised so high that it's often not enough to slap something together to gauge user interest.

The takeaway we've learned is that startups with experience building websites need to step back and redefine their MVP expectations when building a game. It takes longer to make something fun rather than just functional, and unfortunately games are defined by the first term, not the second.

Your Child's Name Won't Limit Their Digital Future

What are you planning to name your children? If you answered with any common first and last name combination, your child may be at a digital disadvantage. They will be condemned to a life of appending numeric sequences to their user names, picking off-brand Twitter handles, and choosing unrelated domains for their websites. Their Facebook and LinkedIn profiles will be difficult for new acquaintances to locate, Google results will misrepresent them to future employers, and their children will have an even harder time of it.

At least, that's what I thought several weeks ago.

It's easy to imagine that any given name you choose for a child will be common enough that somebody out there has grabbed the domain, the Twitter handle, the Facebook username, the Gmail address, and countless other digital identifiers. In this paranoid world-view, your child is relegated to being a second class citizen of the net (or whatever it becomes) simply by dint of having a common first/surname combination.

But how large of a problem is it really? To answer the question, I did some research on full name variation and came away surprised.

It is trivial to find popular baby names for a given period of time, but finding the frequency of full names in the US is another matter. The CDC and census bureau both don't have full name information for confidentiality reasons, and the only place I was able to find a list of unique first and last names was from a company doing greyhat Facebook advertising and data mining [1]. Here is their list of the 100 most popular first and last names on Facebook in 2009 [2].

I then ran these names through a bulk domain availability search (where the name "John Smith" turns into "johnsmith.com"). Not a single .com domain is currently unregistered.

Of course, that test doesn't tell me much except that at the very edges of the full name distribution, most of the domains are taken. The bigger question is how long the tail is. If 98% of first name/last name combinations are not contained within the top 100, then your kid will be fine unless you do name them John Smith.

Since I wasn't able to get my hands on a dataset for full names, I decided to approximate the frequency of firstname + lastname duplicates based on individual surname duplicate frequency in the US. Because surnames have fewer letters than first names and surnames combined, there are fewer possible combinations. This means that there should be a lot more duplicate surnames than duplicate first name/surname combinations. So, if it turned out there was a lot of surname repetition, I wouldn't be able to conclude much about how common first name/surname combinations are, I would just know that they are less common than the data I can see. However, if duplicate surnames turn out to be relatively uncommon, I could conclude that first name/surname combinations are more uncommon still, and disprove my suspicion that the uniqueness of my child's name is important to their digital future.

Among surnames, the top 100 most popular constituted 16.4% of the US population, the top 1000 accounted for 38.9%. That means that the majority of surnames lie in the long tail rather than the head of the distribution. We can therefore assume that the distribution of full names is even more skewed towards the long tail since the domain space is substantially larger. This assertion is supported by the Facebook data above.

This Wikipedia graph suggests there were around 350,000,000 users on Facebook in 2009. The total number of names in the top 100 comes to around 592,000, which represents just .1% of users on the site.

The conclusion I have to draw is that unless you name your son John Smith and your daughter Sarah Smith, they will have perfectly viable digital identities available to them when they graduate from college and need to start looking presentable to the rest of the world. So rest easy; I know that I, for one, want to name my first child Ellis. According to White Pages Names, it seems it is suitably obscure, and hey, I notice the domain "ellissaines.com" isn't registered. I think I'll register that ... just in case.

Special thanks goes out to the staff of the Alden Reference Library at Ohio University for helping me obtain this link and other vital statistics I used in this blog post! Specifically Sherri Saines (@bibliosanity), Tim Smith, Kelly Broughton, and Cary Singer.

[1] The company was later banned from scraping Facebook, which is presumably why their data is so old.