Monday, November 1, 2010

Automatic continuously updated web design - the site is alive...

So I have an idea...don't we all?

Ever since adwords was able to automatically optimize the display of ads based on click-through rates I've been curious about automatic content and web design.

Shouldn't web widgets, such as those at jquery automatically rotate through colors, size and perhaps even functionality from the moment they are launched?

This way they would be regularly updated and improved.  Some of the website control would be turned over to an optimization program that would cycle through a predefined range of options (colors, size, location, functionality) which would then feedback and choose the best combination set.


I can see one bit of the admin console that looks something like this:

Each in this way, multi-variate testing becomes even more scalable so that widget option rotations could be linked. The purple button rotates into orange when the content switches to internal vs. external.

In its simplest form a button is optimized with no input from the site editor. In it's most complex form, the entire size can morph and change based on site visitor behavior....


The site responds to traffic and builds itself in a way that optimizes whatever goal is set...

Saturday, October 30, 2010

The math behind bit.ly and other url-shortening services and other numbers on the web

My son, who is 12 years old, was recently assigned homework dealing with outcomes.  These are the calculations that feed into probability calculations. We use outcomes all the time to calculate our chances of an event occurring given a predetermined number of possibilities surrounding prior events.

If Janie has a red blouse, a blue blouse and a green blouse and she also has blue jeans, khaki shorts and black capri pants, how many outfits can she create?
Aside from the  possible Glamour donts that Janie could create, we can calculate Janie's number of outfits by multiplying the number of blouses (3) times the number of leg-wear (3) and get 9 total outfits.  This can be generalized by saying that the number of outcomes is the product of the total possibilities of each subsequent event.  This can be represented by a tree diagram as well.


This math is actually the basis of our numbering system. Every number we use every day from our age to our shoe size can be represented as the outcome of a tree diagram.  Each digit in any number is represented by one of 10 possible characters 0, 1, 2, 3, 4, 5, 6, 7, 8 or 9. (These are called Arabic numerals.)  The second digit of a two-digit number can also have 1 of these 10 different characters.  Each digit in a number has 10 possibilities. If we draw a tree diagram for a 2 digit number, it has 100 possible branches.  Below is a tree diagram for a 2 digit number (I left out 2,3,4,5,6,7,8 for space reasons).

Each additional digit increases the number of possible outcomes by a factor of 10.  So a 3 digit number has 10^3 possibilities or 1,000.

You may see numbers mixed with letters on the, such as FF6.  This is a number, but instead of using 10 Arabic numerals for each digit, the character set has been expanded by 6 to include A, B, C, D, E, and F.  This yields a total of 16 characters for each digit.  This is called "base 16" or "hexadecimal". Our out numbering system is "base 10" or "decimal". Each additional digit in a hexadecimal number increases the number of outcomes by a factor of 16.

Anyone who has some knowledge of Cascading Style Sheets or HTML has come across hexadecimal in color codes.  All colors on the web can be represented by numbers and those numbers are often represented in hexadecimal instead of decimal. You can see breakdowns of the colors in both numbering systems on many websites. (for example, http://html-color-codes.com/ and http://www.december.com/html/spec/colorhex.html)

The reason behind this has to do with how computers think and how early computers stored numbers in memory.  Computers at their simplest operate in base 2 meaning they have two options, off and on. This is typically represented by the numbers 0 and 1.   Each on/off digit was called a "bit" and in fact on/off fields database tables are still called bit field.  Early computers, notably the 6502 microprocessor (which ran the Apple ][+), used 8 bits together to process calculations.  Half a byte was 4 bits. And with a 4 digit number in base 2, you have a total of 2x2x2x2 or 2^4 combinations or 16 total.  And instead of writing out 4 digit numbers, computer memory was often represented in base 16. (The name of this blog, FF69g, is the computer instruction that the original Apple ][+ used to exit from machine language mode and it refers to memory location FF69, which is a hexadecimal number)

Let's take a moment and talk about what Bit.ly and other url shortening services do.  They are very simple services in concept.  With an url shortening service you can take a long web address and shrink it down, like putting a wool sweater in a dryer, only when you put the sweater back on it magically returns to its correct size. Say you have a very long web address, for example if you search for "what is the capital city with the longest name?" on Google the actual results are in this web address:

http://www.google.com/search?q=what+is+the+capital+city+with+the+longest+name%3F&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a

But wit an url shortening service, that search becomes

http://bit.ly/db3e3Q

The longer url can be hidden on a web page using the tag.  That search is linked here, for example. But with non-HTML services, such as text emails and text messages and Twitter, the longer url is simply inconvenient or impossible.  So url-shortening services evolved and now there are hundreds of them. They generally have short domain names such ".ly" or ".co".  (See recent controversy surrounding ".ly")

So bit.ly and other search services save the long url in a database table and assign a code to it. And when your browser requests a bit.ly url, the code that looks random (db3e3Q in this case) is used to look up the correct url in the table and redirect your browser to the destination, in this case a Google search.

Each url that bit.ly stores get it's own unique lookup code.  And sinceI assume bit.ly stores billions of urls, the lookup codes could get very long (and I'm assuming there are billions since I don't really know).

So a 10-digit lookup code using Arabic numerals would have about 10 billion possible outcomes  from 0000000000000 to 9999999999999.  In the computer world and the search world, that's not that much.  That means if everyone in China searched 10 times the service would run out of lookup codes if they wanted to use only 10 digits.

In order to increase the number of possible outcomes without increasing the number of digits, bit.ly added upper and lower case letters to each digit.  So instead of 10 possible characters each digit has 62, because there are 10 Arabic numerals, plus 26 lower case letters and 26 upper case letters. The number of possible outcomes for a 1 digit code with 62 characters is 62.  The number possible outcomes for a 10 digit code with 62 possible characters in each digit is 62 to the 10th power or 839,299,365,868,340,000 or, in words,  eight hundred thirty nine quadrillion two hundred ninety nine trillion three hundred sixty five billion eight hundred sixty eight million three hundred and forty thousand.




Now, math purists (and bit.ly mathematicians) may say I've actually undercounted the possibilities because not all 10 digit numbers are used, so, in fact, the remaining digits in any combination could be 63.  For example, 0000000000 and 0 could be considered different lookup codes by the bit.ly servers (and they probably are).

This means that the actual number of outcomes is the sum of the various-length codes from 1 digit (62 codes) to 10 digits (the aforementioned long number).

I've never seen a bit.ly url that has only one digit in the lookup code, so perhaps they've limited the number of digits in their lookup codes. And in fact I have no idea if they even use 10 digits.

In fact, I don't know anything about how bit.ly works at all. I'm just assuming this since I had to create a similar system for my QR code tracking service that I built for fun (http://www.bloty.com) which might one day become a way for people to tell you that your lights are on in your car or that you are illegally parked while sitting at a restaurant (http://www.notesonawindshield.com, but it's in beta, so apologies for errors).

My code to generate the look-up using 62 possible digits is written in PHP and I've cut and pasted it below. No apologies if it doesn't work for you, I only recently learned PHP and it may or may not work exactly as planned. "UUID" stands for universal unique identifier.

  function uuid()
  {
//62 ITEMS IN THE ARRAY
  $arr = array(0=>0,1,2,3,4,5,6,7,8,9,0,a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z,A,B,C,D,E,F,G,H,I,J,K,L,M,N,O,P,Q,R,S,T,U,V,W,X,Y,Z);
  $uuidtemp='';
 
  $i=1;
    while($i<=12)
          {
          $uuidtemp = $uuidtemp . $arr[mt_rand(0,61)];
          $i++;
          }
  return $uuidtemp;
  }


Wednesday, October 27, 2010

The Best Source of Market Analytics in the World: Customers (And how to use them)

It's funny, when I started LandAndFarm.com I never expected to be talking with customers about their kids or the weather in Tulsa, but when you do customer service and you do it right, you talk.  And if you can steer the conversation the right way, you can gain a fount of information from customers.

The customer service team is the best source of market analytics available to any company.  No complex funnel analytics report, no consultant, no website or blog will give you more valuable information than talking directly to customers. (OK, maybe exaggerating a bit, but seriously...do not ignore the customers...)

Whether the customer has called to complain or congratulate, they have equally good information, and a case can be made that complainers actually are more valuable than satisfied customers.

Let's talk about what kind of information customers have;

1.  Customers have information about the product.

Both satisfied customers and complainers have information about your product. They bought it. That means they've done more QA on your site than you have because they put real cash money through your site and actually bought something.

If it all worked out, great. So ask them what else they want. If you sell cars, ask them why they bought it (color, make, model, price, etc.) This will give you an indication of what your competitive advantage is.

And if they are calling to complain, you've hit an information gusher.  DO NOT LET A COMPLAINER GET OFF THE PHONE WITHOUT A CLEAR UNDERSTANDING OF WHAT WENT WRONG. 
 
2.  Customers have information about the market.

When I talked with brokers selling land (my customers) I always asked them how the market is. And brokers love talking about the market. Now, you have to know a little about brokers to know that they will rarely say, "The market sucks." But they might say, "Things are getting better" (translation: the market sucks).  Get trend information from them.  And if you hear one thing from Frank in Tulsa, you might tell Bobby in Austin that "Someone in Tulsa just told me..." And then Bobby might give you some more market insight.

Get it?

3.  Customers have information about competitors.

Even though my biggest competitor at the time was run by guys I really respected, that didn't stop me from agressively asking my customers if they used them and how much they paid and what they did that I should do or what I did that they don't do.

I bought eLandUSA.com because of a customer told me it was for sale.

4. Customers have information about pricing.

Pricing is key in most businesses. Too cheap and folks think it's not worth anything. Too expensive and everyone flies off to the competition. 

I always asked customers if my price was fair.  And I was the most expensive site in my business for a long time because rarely did customers say it was too expensive.  Of course, some did (and I could give a discount on the phone immediately to those that asked...)

My crazy recommendations:


1. Everyone in the company from the CEO to the janitor should listen in on a few customer service calls or read a few custom service threads every month. Calls and live chat are better for obvious reasons.

2. And senior executives should actually sub-in for customer service every so often (quarterly at least) to stay in touch with customers.  I answered CS calls until the day I left Loopnet.

3. Customer service staff should prepare regular weekly reports that summarize major CS issues, particularly repeat issues.  These reports should be no more than 1 page and cover the 5 most common CS issues and one unusual one (to keep the mood light).  These issues should go into your bug/work tracking system you are using company-wide and should be re-summarized in a monthly report that explains whether the issue is ongoing (we are too expensive but we have not made a decision on it) or resolved (customers no longer get logged out randomly, thanks engineering!).

Friday, October 15, 2010

Rise of the Superstar Designer

For entrepreneurs, the "team" is of utmost importance. As uber-tweeter Dave McClure has pointed out out in his many presentations, including the not-so-tasteful ones, the team is the secret sauce.

The product might be killer, the market might be huge, the timing might be perfect, but without the "team" it will be an uphill battle for cash, customers and cachet.

And while attending the FutureM Bootcamp Marketing last week, I heard one of the best presentations on design I have ever heard.  Leslie Bradshaw of Jess3 gave a slide presentation that simply blew me away.  Now I'll admit that I cried at the end of E.T. so maybe making me emotional isn't so difficult, but if anyone could have shown me that slide presentation in 2001 when LandAndFarm.com was a free listing service maybe I would have bought Loopnet.com instead of selling to them, er, probably not, but who knows?

And her point, which I asked her about from the audience was my own naive question, "Are you serious that you should have a designer as a co-founder of your start-up?"  And her answer was an emphatic yes.

And so I was having lunch with a developer/friend who builds mobile apps for brands like Tribe Hummus and we got to talking about our next startups.  I pitched my two ideas for local cash-register-based loyalty systems for the smallest stores and car-based social networking and he pitched back his idea for an enterprise level location-based content platform for enterprises.

And we both agreed that the missing 3rd person was a designer.  And then the realization struck that even though I've got some street cred from selling my company and I have a pretty strong understanding of UX and he has some major mojo/guru tech skills because he can whip up mobile apps like I can scramble eggs, a real designer (and I mean a UI/UX whiz with artistic talents) who has some tech knowledge would be a goldmine for our team.

And how much would that person be worth?  Answer:  a lot.

And how much would that person charge? Not enough.

So this is my prediction:  soon we are going to see some status inflation among the best designers.  Like Michael Lewis so correctly described in the Blind Side, the left tackle evolved when it's value protecting the quarterback emerged.  The designer is the left tackle of start-ups. And maybe the analogy isn't perfect. The designer isn't really in the background, the designer isn't hidden, her work is right on the front page.  But it seems the designer is still regarded as something you outsource, something you can "do later."  But in our visually active world, where "engagement" and "interaction" are so important, the dedicated designer who doesn't write code, doesn't market, doesn't sell and doesn't deal with the numbers, but can talk with all those folks, is like a keystone for a team.

And there are some superstar designers. Those who are "in the know" already agree with this and have already stopped reading my blog post and are probably giggling that I'm just now figuring this out, but what can I do?

In terms of value, the designer is now where the left tackle was when Bill Walsh and the 49ers were coming into their own.  There was a blog post by Ryan Junee that valued engineers in a start at between $1 million and $2 million.  There will be a time and it may be soon when the designers and the UI gurus and the UX whizzes will be valued similarly and perhaps even more.

Because isn't design as important as anything else? Hasn't Apple proven that design is elemental in importance?

I remember buying my first iPod nano and thinking how beautiful it was, forgetting that it was just a device to suck my cash out through itunes.  It really fit in my hand beautifully. It was sleek and easy to dial, even easier than my phone at the time.  I was impressed.  And then looking at competing MP3 players that had nearly the same look and feel as theWalkman from 20 years early, only smaller, well, we know who bought what.

So my next startup I hope to have a designer on board and I will be lucky to have one. Until then, my presentations will still be sucky and be in powerpoint. I deserve what I get.

Sunday, October 10, 2010

Waking up in the new world order of entrepreneurship

Skip,

I know by now you are thinking I'm a bit crazy as a loon, but it's all part of my learning curve as I try to get a grip on everything that's happened on the web in the last four years. I feel like I've been asleep and when I went to sleep it was "The web" and all you had to do was build a great site and drive traffic and put advertising on it and have great customer service and you'd make money. And that was the case and that's what I did with landandfarm.com and I sold it and made a little money and gave most of that away to my ex-wife, the government and my brother, in that order.

Everything was changed when I "woke up" 6 weeks ago when I realized that I had to get a job or start a company and it was not "the web". I don't know what "the web" is anymore. There is the "social network". There is the "web of physical things". There is the "mobile web". Al Gore's joke about inventing the internet is ancient history. Like, REALLY. ANCIENT. HISTORY. This is a world where your local doctor doesn't even have a smart phone, but the drop-out punk on the street corner panhandling is the Foursquare Mayor of the most expensive eatery in Davis Square. There is the world where rural Realtors(tm) whom I taught how to use AOL email now have automated twitter spam-bots that send out property listings to 100 different social networks automatically every night. I sometimes wonder if some of those brokers are dead and their web services continue to operate automatically, forever spewing out discounted lot deals onto the free social web services, blocked and filtered at every end-point and thus never being traced back to their origin to discover that Elber Potts, connected-Realtor(tm), died of a coronary in his garage in 2009 while fixing his riding lawnmower.

Everything changed. The local dog groomer no longer needs to buy Adwords because Yelp is all she needs for marketing. The local coffee shop owner has 15,000 followers on Twitter (and doesn't have even 500 real customers) while the local convenience store sells flash drives at the counter. The wine store has a SCVNGR QR code on the front door and an automated Group-on, Twilio-enabled phone messaging system so that you can pre-order Kazakh brandy before it shows up in product marketing on Entourage and watch it on your mobile device as it leaves Astana, tracked by an RFID chip and heads to you, arriving in 3 months (estimated) and the bottle of brandy has it's own fan-page on Facebook with 3,000 "likes" and if the brandy doesn't arrive before the new season of Entourage starts the case will be automatically auctioned on eBay and the funds converted to Farmville currency which the liquor store will buy from you and give you back two cases of your favorite beer.

I have had so many ideas, a new one every few minutes it seems as I've been doing a crash course on the "new web" that I feel like my brain will explode, spewing bits of meta tags and html anchor links around my home-office. and these ideas are so revolutionary that I am high like I've had cough syrup in a cup of espresso for hours after as I furiously research each one. And each idea I have, when googled, reveals a University professor in the UK who has a free version running that he and his class built in 1989 that's still running on the university mainframe and has it's own Facebook page with 100,000 "likes". And there's a venture back seed fund around the technology and everyone not only knows all about it, they are all onto the next big thing. And the new services I think of and the new ideas I have aren't really so new. They are old. Old, like more than 12 months old. Old, like more than 5 months old. The phrase "that's so last year" fits well and the tween version "that's so 5 seconds ago" isn't so ironic; it's actually quite unnerving.

For years, I was ignoring folks who were trying to tell me about twitter and facebook and social networking and I was asleep and arrogant, thinking that I knew what was what. I knew how to make money on the web. My customers paid me REAL. CASH. MONEY. And it flowed in like a broken ATM machine. I was well-off, I traveled to Asia, to Europe. People returned my calls. But I had nothing really to stay since I was being deaf to the changing world. And so I stopped calling and so nobody called back. I was basking in the glory of my successfully exited 8 year old startup. Hah. EIGHT. YEARS. OLD. It's like a bad joke. My startup was built in VBScript, using my knowledge of Applesoft basic, which I learned in 1981 on an Apple ][+ computer in my parents basement. My sell-by date is so in the past, it's rubbed completely off and is unreadable. Now startups aren't startups past 6 months. Twitter is 4 years old. FOUR. That's younger than my kid. My child is out of date. My future children will be born out of date. Who the fuck can keep up with what's going on? AND. I. THINK. ALL. CAPS. WITH. A. PERIOD. IS. GOING. TO. DRIVE. ME. INSANE. AND. I CANT. STOP. DOING. IT.

Web design, as well, has changed. It used to be all tables and easy shit. Now it's CSS and XHTML and object-oriented, DOM model AJAX enabled bullshit that I can't understand. And by the time you read this there will be some new, awesome standard that allows more connectivity, probably with time travel so it will be out of date before it's developed.

I built one of the first blogging sites (diarypro - failure) and one of the first rent payment systems (electricrent - failure) and bought and dis-improved an early social network (bloty.com - failure) and bought and fucked up an early photo sharing site (xaqa.com - failure) finally hit it small with a property listing service. But now I can barely hack out how to link my Twitter and LinkedIn accounts in a way that won't make me look like a noob. I spent 20 minutes recently trying to undo a style change I made to my Twitter page that made the links the same color as the background so I couldn't actually see them. And I have a blogging site or rather I have about twenty because I keep registering at various sites like blogger and tumblr and myspace and I can't remember which is which and if I should have a personal blog and a business blog so that the content doesn't get mixed up or two personal ones, one for my friends and one for my family or maybe three personal ones so I can separate my friends who knew me when I was young and carefree and those that know me now or four personal sites so I can be friends with all my ex-girlfriends and keep them from my awesome current girlfriend who doesn't want my exes sending me flowers on facebook and who may or may not still have a thing for me even though I'm pretty sure they all are married and have kids and have 5,000 followers on Twitter and have crowd-sourced their own blog page design so it looks totally awesome and feels all web 3.0.

I feel like I'm so late to the party that the hosts cleaned up, got older, had kids, left town and sold the house to a developer who tore it down and put up a condo development that's half-built, half-sold and sitting abandoned at the end of a cul-de-sac in a gated community and when I come by with my bottle of wine to give to the host, I'm greeted with a QR Code that I can scan for loyalty points that can be redeemed when I visit the next open house and after I download the real estate developer QR code reader app I have to register and while I'm trying to enter the same password i use for EVERY. SINGLE. SITE. I drop the bottle of wine and then drop my phone into the puddle of broken glass an wine and cut my hand trying to fish it out.

And holy shit there are SO. MANY. STARTUPS. It used to take balls and gumption and a fearless desire to work for yourself to start your own company. Now it's a required course at the local community college. And to make my brain sizzle and burn at the edges just a little more some much-more-successful-than-me entrepreneur started 500startups.com which is jammed packed full of awesomely design hot hot hot super api-enabled, networked sites with twit/face-linkage a-go-go and seed-round funding and A and B round funding and Omega/Zeta/Tau round funding and a board full of people who sold their first web startups in 1965 and repeated that feat every year until last week when they won the "80 under 80" lifetime achievement award for entrepreneurial achievement from the Never Heard of It Foundation Of Super Successful Folks in Warm Weather Climates. And I should know who these people are. And I don't.

Starting-up is now a viral phenomenon. Everyone has a startup. Because every blog is essentially a publishing startup. Every twitter account is a mobile startup. Everything is a startup. You can get seed round funding for choosing the right hashtag and going viral with a video-enabled tweet.

In 2006, Twitter launched. 2006! Can you believe it? My youngest son is older than that!

And this entire rant was supposed to be the pre-amble to describe what I want to do with hyper-local marketing and propose a new start-up using some of the new technology that's out there, but I'm a bit fried....

Thursday, October 7, 2010

Bloty.com launched - QR code fun

I recently launched Version 1 of http://www.bloty.com. This is a QR code generator that shortens urls and creates a QR code for the url.

When scan a QR code (using your smart phone - download a QR Code reader like Shop Savvy for Android), Bloty will store the time of the scan and send you an email if you want.

You can create unlimited QR codes.

I'll be posting some videos of how to use QR codes in the next few weeks, plus launching some more features.

Thursday, September 16, 2010

Financial Modeling Tutorial

After ignoring my financial modeling tutorial for nearly a decade, I have done a spit-shine and moved the spreadsheets to Google Docs.

It's actually pretty good at explaining the basic differences between an income statement, a cashflow statement and a balance sheet and even delves a little bit into depreciation.

I may start another site called "WTF is a bond?" but still mulling the idea.

Steve's Financial Modeling Tutorial