So what happens when you discover that a feature isn’t working properly in one obscure browser?  If you’re like me, you let it go. 

There are quite a few obscure browsers out there, and my obscure meaning that the percent of the overall users using them is very little: Opera, Konqueror, Seamonkey, SRWare Iron, K-Meleon, Lunascape, FlashPeak SlimBrowser, Songbird, Beonex, Camino, Galeon, Epiphany, Kazehakase, Avant Browser, Maxthon, and I’m sure the list continues.  Even well known browsers (though older versions) fall into this category: Firefox 2, IE 5.5, etc.

I pretty much stop at 3%. If your browser’s share of use is less than 3%, I’m not fixing the issue.  Even if your given browser is 100% standards compliant and scores 100 on the Acid3 test, you’re out of luck with me.

I also stop at 3% on not only browsers, but also operating systems. Still hanging onto Windows ME? Not going to test or fix my apps on that.

When it comes to mobile, you’ll have to find the percentage you’re comfortable with. There are just so many handsets and form factors out there.

Finding out the Cut-off

Where do you find out what to cut out?  The main thing I use is Google Analytics stats for zKorean. Since it is a high-traffic site with a good spread of users across many ages and demographics, I’m pretty comfortable with it.  I also tend to use W3Schools stats (yes I know there are W3Schools haters, don’t give in to hate).  But be warned that sites like this have very tech-heavy visitors like programmers and designers which can skew the OS and browser numbers towards Linux, Mac, Firefox, Safari, and Chrome, so take their stats but keep it in mind.

See how Opera is below the 3% cut-off? Sorry, Opera users.  Then I have to dig in to each browser and see where its cutoff is for its per-version market share.  Your favorite spreadsheet program comes in handy here.   W3Schools breaks it out for you already, so you can put away your calculator.  But if you’re doing it on your own, and browser X has a 55% share, but it’s newest version only has 12% of that, then the newest version is at 6.6% of overall share (0.55 x 0.12 = 0.066, then multiply by 100).

Love Thy Customer

But when responding to a customer about it, just be apologetic with a little friendly prodding:

“I’m sorry that you’ve had an issue with using the sorting feature on x browser. I would like to help, but our company does not have the resources to test across a large number of browsers and platforms.  The best we can do is make sure that it runs on the most common configurations. In the future, hopefully we’ll have the ability to run our tests across even the rarest of browser/OS combinations.”

Some of you may say, “well x browser uses the Gecko engine, just like Firefox, are you saying you don’t support it?”. I support the browser, not the engine. Since browser makers are always looking to add some feature to make their browser special, all bets are off if they break the rendering or goof up behaviors.

Don’t be That Kind of Designer

Don’t take the other extreme, either. I’ve seen many a condescending message from a “better than you” designer on a website.  I’m talking about the kind of web designer that thumbs his nose at users running IE, or visitors that are using browsers that aren’t standards compliant (and I could rant on how browsers that tout standards-compliance always add in special sauce to push their own standards, but I won’t). If a browser is below 3%, like IE6, you can show a nice message to your visitor, or let your pages degrade somewhat gracefully.  But if I happen to be on IE8 and checking out your site, I don’t expect “Your browser is weaksauce and doesn’t support web standards.  Go away and come back on a real browser.”  Almost as bad are sites that just fall apart on IE (even IE9) because the designer didn’t think it worth their time to even check.  Yes, the box model is different.  I get it.  But regular Joe web user is not going to blame their browser for how bad your site looks. They are just going to think you suck as a designer.

Going Large

If you are Google or Yahoo, you have the resources (people power) and extensive automation/testing systems to test across a larger range of possible platform/browser combinations. So Yahoo and Google don’t cut off at 3%.  They go below 1%.  The reason is they can afford to not alienate users on obscure configurations.  And each 1% represents millions of people.  It’s a noble goal to be inclusive as possible, but within the limits of what you can handle.

What Do You Think?

Comment below on where you think the cutoff should be.  How far do you go to ensure the experience is pleasant to a variety of users?

Next post: I’ve Joined Earndit as an Advisor

Why dumb? Because once you’ve had people contributing to your site with their hard work and diligence, you decide to sell or rent all the work they did.  Even though you may continue to let those who built it use it for free, it’s still a jerk thing to do.

I was going to write an analogy about a farm and community garden, but, like most analogies, turned out to not mimic the scenario.

Here’s the bottom line. Don’t be a content punk. Don’t try to make money off the hard work of your visitors. You wouldn’t expect Wikipedia to all of a sudden put up a paywall?

What about sharing in the profits?

Now that’s a business model! This has been done before.  There are a number of topic sites where people write on different topics and get paid for traffic that their pages generate. It depends on the site and the philosophy of the site. If users know they are writing content to get paid, it’s quite different than writing content to help others.  The reason? It crosses the line from a social transaction into a monetary transaction.

Next post: Solo Focus: DuckDuckGo

Let’s say you’re hosting web sites for customers.  You can’t just turn it off.  Your customers may have their business running on their site.  This situation is one of the hardest.  You have to work with the customer and help them migrate.  If you just shut down their site with little warning you’ll be getting calls from lawyers.

You may be wondering, if a site actually has customers, why would you shut it down?  Good question.  Sometimes it’s just not financially feasible to have a handful of customers.  You’ll have to do the math for your situation.

The site shutdown scenario doesn’t affect just web hosts.  What about a Flickr-style site?  Are you going to do one last backup then turn it off?  If your site has the only copies of the photos from some kid’s third birthday party you’re going to have some anger heaped upon you should you take it away.  In this case, where you’re storing items for little or no cost, and you can’t make enough to afford the scaling issues, you’ve got to give plenty of warning to customers and make a way to get them their files.

Any startup that holds information for a customer, whether it be images, a social graph, transactions, even links (like tinyurl or bit.ly), can create legal troubles if taken away.

The safe thing to do is to never start this way.  Make sure your business model can support itself.

For small self-startups, freemium can wipe you out.  You may overestimate the ratio of free vs. paid customers you’re going to get.  Then you’re stuck with a crippling amount of data to manage/store/serve and the costs involved.

With the right business model where revenue from paying customers dwarf your operating costs, you’re in the clear.

Next post: Professional Design at a Bargain

I use three methods:

Pay by the Month

This is an example of a SaaS payment model. Customers may use your product daily, but they pay on a monthly basis. Customers may get unlimited use of your service, or it could be metered in some way (5 faxes per day, for instance).

This model is lucrative when businesses are your customers, since they tend to review products they pay monthly subscriptions for on a quarterly or yearly basis. Many times your business customers won’t be using the product at all, but will still be paying about $15 to $100 a month (or more). Conversely, individuals are more likely to use what they are paying for more often and you’ll get a lot more turnover, especially when your subscription fee is less than $10. Individual customers tend to be fickle.

Pay per Use

Your service may not warrant a pay-by-the-month charge. Perhaps you offer a one-time on-demand service or something that will be used irregularly. Charging a fee is a good way to go. In many cases you don’t even need to have the user create an account.

It may be the case that your service is process intensive or uses an API that you have limited use of. Offering this as an unlimited service on a monthly payment basis could put you out of business.

Hybrid

Some customers may not need your service every day even every month, and just need to use your product briefly. You may choose to offer both methods, offering a limited use for a slightly premium price.

Conclusion

However you decide to go, I recommend running some numbers beforehand before committing, because the switching cost could be painful.

Next post: from BlackBerry to Android

He who can destroy a thing, controls a thing.

  – Paul Atreides

And if that thing is valuable, you become the monopoly.

Providing a reference tool is a good business idea.  I’ve been in organizations where having a large-size data repository and searching capabilities keep people coming back.  This is a solid method for building a long-term SaaS (software as a service) where you can charge a monthly subscription for its use.

My Experience

For zKorean, I spent years building an online dictionary.  Over time I added more entries and added features that made searching it easier, too.  Now the dictionary has tens of thousands of entries (just recently added 400 and have immediate plans for more).  People can use it for daily use on a limited basis (25 searches per day) and get unlimited use for $4.95/month.  There are also audio pronunciations for all the Korean words (voiced by my lovely wife Suzy).

Am I bragging or trying to make a pitch?
No.  The point I’m trying to make is that it takes a lot of work and you have to keep it fresh to keep competitors at bay.

Once you have a resource like this under your control, it gives you an advantage that your software doesn’t need tweaking or maintenance very much.  A few features now and then where you see the need, but mainly your data is where you’ll make improvements.  And adding data means not having to touch your code.

Data is Expensive

When I worked at a company that offered an online medical coding reference, we licensed data from other companies.  Our technology was based on bringing multiple data sources together to give the best information display to the people that needed it.  And the service was expensive, partially because of the cost of licensing the data in the first place.  I won’t give numbers but once someone comes along and wants to license a data store that you have, it becomes even more profitable.  They pay you for usage of the data, and they make the sales and handle their service.  Your responsibility then is to deliver quality data in a timely fashion.

Is It Possible for You?

Take a little time and consider a hobby or area of expertise that you have.  Is there a large data store that you need or that people like you need?  Think about it a second.  There are databases of Scrabble words (by word length), World of WarCraft game items, medical terminology, and automobile parts.  With some effort and time, what interest do you have that could turn your hobby into a business?

Next post: one browser is not enough

If you go to Rand-McNally nowadays, it looks like MapQuest did back in 2005.

My question: Why wasn’t Rand-McNally the online leader the whole time?  They had the data (the hard part) and just needed a way to get it online, then have a very small group of experienced web developers and software engineers make improvements over the years.

Google’s draggable maps were the rage back in 2005, using AJAX technology that was not widely supported at the time.  There’s no reason why Rand-McNally couldn’t have owned this space, and run the MapQuest style maps for browsers that didn’t support AJAX at the time, and offering draggable maps where supported.

Now in automobiles you have TomTom and other GPS systems for navigating the road without a paper map.  Rand-McNally has its own, but still nowhere near Garmin or Magellan.

This is not meant as a rant against Rand-McNally.  It’s a rant against companies who have the resources but shallow vision.  Where is the long-term vision?  Or is the vision there but you’re too large to move quickly?

If you work in an industry that could do so much with what it has but crawls in terms of progress, perhaps it’s time to go somewhere else, or take advantage of the situation and create a competitor that blows them off the map. 

In these days a small player can get word-of-mouth faster and more widespread than ever.  MapQuest (an unknown) beat Rand-McNally.  They blew their brand off the internet.  Did you even know Rand-McNally had a map site? I didn’t.

So don’t be afraid if you’re going against a giant company.  If you can deliver a better product, the brand name is inconsequential.

Next post: a business lesson from Dune

When you’re running your own startup and trying to make your product better, it’s tempting to take everything your users say and build it into your product.  After all the users are the ones closest to it, right?  A user makes a suggestion and you’re more likely than not to add it in.

Doing things this way is part of calming our ego.  When a complete stranger who is kind enough to use your site gives you a helping hand, it’s hard to say no, right?  They want to know that you care.  Well caring is not the same as always saying yes.  Sure, your kids may want candy for breakfast or want to stay up until midnight, but you don’t let them.   It’s not part of the vision you have for their health and well-being.

Let me make a distinction.  A user is a prospect, a site visitor, a person who uses your site (perhaps as a freemium structure) but is not a paying customer.

Customers help drive company direction, whereas users provide valuable input.

If you always listen to users and follow their advice:

• everything will be free and therefore make no money
• your products will be built by committee (of everyone)
• your vision will not matter

So should you do everything customer suggest?

Nope.  It’s your business.

Quote from Henry Ford:

If I’d asked my customers what they wanted, they’d have said a faster horse.

So even when your customers suggest or even demand something, you need to think about how it fits into the vision you have for your startup. If it doesn’t fit, you must reject it.  Now there may be an astounding insight that comes up that shakes up your vision and changes it.  That’s a good thing.  But don’t go backwards, only forwards.

Won’t a user become a customer?

Yes, but getting your customers satisfied means that once the user does convert to a customer, their experience will be so much better than when they were a user.  But don’t make the user’s version suck.  Just make it not as great as the customer’s version.  If your users do not convert to customers, then the product wasn’t for them in the fist place.

Do you have evidence to back this up?

Luckily. Here’s a good, short post on BlueMango’s experience with the launch of their ScreenSteps Live product.

But my business model only has users.  It doesn’t have customers in it.

That is why you fail.  Listen to this podcast and make your hard efforts worthwhile, and not just a hobby:

37 Signals podcast: “Making people pay and targeting nonconsumption”

Next post: ad solutions for publishers/content sites

Today’s angels and VCs are holding on to their money.  It’s not good enough that your startup gets tons of traffic.  Traffic does not equal money.  If you don’t have a product or service that people are willing to pay for (either sponsors or users), then you don’t have a service worth building.  It might be a service that makes you happy to offer, but you won’t get funding.

So why am I talking about creating a story to get funding?  It’s a good test to discover that what you’re doing is worth your time.  If you can’t honestly come up with one solid idea that will create long-term revenue, then why are you working so hard?  You need to sit down in a quiet place and figure out that one sure-fire revenue method.  And a revenue estimate that should make it worth your time.  Otherwise, why are you taking time away from your spouse, children, and other family members?

You need to start making money on day one.  If you start making money within the first day or week that you launch, isn’t that better than working for months and then making nothing?

Forget ads.  Ads will not make you rich.  Ads are not worth the time.  Some sites like TechCrunch make a lot of money from ads, but they also have many servers, office space, IT staff, article contributors, and other people they have to pay.  If your revenue model is “get traffic, then run ads”, you need to come up with something better.

Be the VC

Take off the rose-colored glasses (honestly does anyone in the 21st century wear these?) and pitch the business idea to yourself.  Or even better, your spouse.  If they aren’t excited about it, then they don’t buy it.  And neither should you.  So figure out a better way to make money with your startup, or dump it.  Life is too short.

Outliers

You may be thinking, “what about Twitter?” Sure they just rolled out a business model and have $150 million in funding so far.  But the investors are expecting a return on that investment.  Probably a 4-6x return.  And no one can cash out shares until that $150 million is paid back to the investors.  Twitter, Facebook and many other “successful” startups like this are statistical outliers. 

Playing the lottery is not a business model.

Next post: serving your business email, part 1

Here’s a perfect example from This Week in Startups episode 7 (Don Dodge of Microsoft was the guest).  A caller calls in with the question.  Click play to jump to it (lasts about 12 minutes).

So how did this situation get resolved?  Click play below to find out (from episode 11 – segment lasts 5 minutes).

Call for comment: have you had this problem?  How did you handle it?

Next post: our first “ask SND” question

Once you click it, you can never go back.

As a newsletter subscriber or “insider”, you’ll get startup opinion and information that won’t be published anytime soon, or maybe ever.  You’ll also get beta invitations to upcoming services that will be available on Startup Next Door in the near future.

What are you waiting for?  Embrace your destiny and click the button.