Cliff Notes from Founder Factory 2009

138View Commentshttp://warwickp.com/2009/11/cliff-notes-from-founder-factory-2009/Cliff+Notes+from+Founder+Factory+20092009-11-20+20%3A11%3A18warwickp

Nov 20 2009 (conferences)

Yesterday the Founder Factory took place at the World Cafe Live in Philadelphia. I love the venue and there were some interesting speakers and discussions. Here are the things I found most interesting from the talks:

Doug Alexander, President – Internet Capital Group

- If you have a B2B product or service, figure out which budget you come out of. Don’t come out of the IT budget, which is heavily pressurized and quickly trimmed. Come out of the marketing, or COGS budget, which are generally more sacrosanct.
- Again with a B2B product or service, you want to be solving a problem that keeps a senior executive (at your target client) up at night. By inference (if it were indeed a b2b product), I declare sildenafil citrate the greatest product of all time.
- Avoid missionary selling, which is to say, avoid spending money to educate influencers in your market that they need your product or service. This missionary work usually has the effect of educating the market on behalf of your competitors as well.
- Advice to a fishbowl company: If seeking outside funding, ensure you have a concrete vision (with detail) of what success looks like for your company. Instead of relying on the potential investor to create their own vision of this scenario, paint this picture in detail and sell this vision to the investor.

Dr Jeremy Siegel, Professor – Wharton School

- While the dollar has slumped recently, it slumped after a spectacular and unnatural rise, setting it’s current value still above the levels it was in 2007 (I believe, I need references, since I do not follow currencies at all). Essentially, view current value in a broader context before declaring disaster. I’ve had this sentiment expressed to me before about the stock market in general.
- Focusing on the deficit and foreign debt as threats to the US economy ignores far larger threats, such as social security and medicare, which represent buckets orders of magnitude larger than the deficit.

David Brussin, CEO – Monetate

- Use external QA resources. It’s a triple win of fresh eyes with more perspective and no emotional ties to the product.
- Cloud, cloud, cloud. And CDN. With cloud.
- Interesting tactic: don’t upgrade your software in the cloud. Deploy new infrastructure and move traffic over, your rollback being the previous setup. (IMHO this is only practical to a threshold. What happens where your ’setup’ is hundreds/thousands of nodes?)
- If your product imposes management overhead on your customer, becoming “full service”, ie: providing a product as well as management of the product as a combined offering, might be a solid commercial solution and an easier sell. “The work on your part will be limited to banking all this money we create for you”.

The Fishbowl

In the fishbowl, three companies in various phases of development, asked a panel of experts (and the audience) for advice on certain issues.

Kendra Gaeta of Kidzillions was unfortunate enough to be the person who taught some of the speakers (and me) a valuable lesson, which is: If you cannot do your presentation without your presentation, then you might want to rethink your appearance. Nothing cooperated with her, yet she soldiered on and won the support of the audience, sans slides. It may have been better without slides. The most valuable advice I took from her feedback is while certain products may lend themselves to white-labeling,  sometimes a strong brand is more valuable to the business, and should remain in tact through all markets.

Ryan Meinzer of PlaySay asked for advice, mainly on a strategy to deal with potential threat from the very large players in markets adjacent to his. The consensus among the panel seemed to (to me anyway) be that when you can profitably serve multiple smaller niches with a predictable formula, you don’t need to be overly concerned about the Walmarts of your space. I appreciate the candor of an entrepreneur who tells you that a business “landed in his lap” and he ran with it. Also, there is dogfooding going on here where PlaySay evolved out of his own need to learn Japanese. There was discussion of patents, and all I could hear was Paul Graham talking about how the value of patents for a startup really is “as an element of the mating dance with acquirers”. I was hoping one of the panelists would say (warning, my $0.02 ahead) that should PlaySay’s immediate strategy be acquisition by one of those neighboring 800 pound gorillas, then by all means invest in those patents, which can be line items in the inventory to sell an acquirer. Alternatively, focus resources on product development.

The three guys from RevZilla have built an impressive business. They bootstrapped heavily, are passionate about their products and the space, and were seeking advice on growth. I was (pleasantly) surprised to hear a panel of VCs suggest to these guys not take take external funding, and thereby introduce a whole new set of circumstances into their business, when they have such demonstrable growth. It seems that the powersports industry suffers from a common problem (is Ludditry a word?) which is that equipment distributors and manufacturers generally will not sell to Internet-only retailers. I found this in the golf equipment industry too, when I looked into it. The Revzilla solution is an impressive looking store front/warehouse in South Philly. In general, it seemed that the advice to RevZilla was that the curve they are on seems very promising, so ride (oh no a motorcycle pun) it as 3 founders to a great life, rather than try to engineer it as 3 founders + VC money and complex priorities. This may be a serious over simplification, of course.

During the course of the day, the following diagram began to form in my mind:

How VCs, engineers and customers view a business

At every conference I have been to, the best speakers are the ones who provide a few select nuggets of advice, for problems they assume the audience has. If, as a speaker, you need to turn around and read a bullet point off your own PowerPoint slide, then you may as well be part of the audience yourself. Also, if you are simply detailing the a story about something you have done, then a book or a blog post is a better format, since you hold a different type of attention from the audience when up on stage. It’s generally not a very patient type of attention, either. Unscripted-ness, or even the illusion of it, goes a long way in winning attention. Speakers are getting better and better at this, but we are still not there. If you’ve ever delivered a presentation which included more than 6 PowerPoint slides, each with more than 2 bullet points per page (that group includes me, by the way), I can’t recommend an evening at The Moth enough.

A big thank you to Philly Startup Leaders for doing this important work in the Philly area. I personally enjoyed the day, and took home some interesting advice and thoughts to write rambling blog posts about.

WP Memcached Manager

130View Commentshttp://warwickp.com/2009/11/wp-memcached-manager/WP+Memcached+Manager+2009-11-18+16%3A04%3A57warwickp

Nov 18 2009 (WordPress)

If you use Memcached and WordPress you may find this little plugin handy. Version 0.1 of wp-memcached-manager allows you to test your connection to a Memcached server, see some basic Memcached server stats and perform a cache flush. There is also a very rudimentary tool to see some of the cached data stored, however, it is super basic and not massively useful. In future releases you will be able to walk through data and perform updates on specific keys etc.

Plugin home: http://warwickp.com/projects/wp-memcached-manager/

On WordPress.org repository: http://wordpress.org/extend/plugins/wp-memcached-manager/

Getting to know our Amazon Kindle

53View Commentshttp://warwickp.com/2009/11/getting-to-know-our-amazon-kindle/Getting+to+know+our+Amazon+Kindle2009-11-12+14%3A00%3A06warwickp

Nov 12 2009 (gadgets)

I’m always sneaking cool gadgets in as birthday gifts to my wife. “Happy birthday, honey, here is your electronic grill scrubber.” But this time she actually asked for a Kindle, so I didn’t need to be too sneaky. The idea is that she can read while trying to hold junior. She reads a lot, so theoretically this thing should pay for itself through the lower book costs. Right?

First thoughts, as I unwrap this thing.

• So Amazon hired someone from the Apple packaging team then? Nicely done.
• I thought it had a paper screen covering on in the box. Turns out that’s the screen. Amazing display. Looks like paper. Amazing! I said that.
• There is significant lag on most operations, even typing on the keypad. The screen refresh rate is low. I guess that’s the very idea of this device.
• The keypad is a little finicky (used that word in two consecutive posts. Need a prize). Particularly the enter/return key which needs to be bigger IMHO.
• There is a browser, and most pages render like it’s 1995 and you just fired up Mosaic. Nostalgia.
• The Amazon wireless network is called Whispernet. I immediately thought of Blue Thunder.
• The first book I wanted to buy is Atlas Shrugged by Ayn Rand. Doesn’t appear to be in the Kindle Store.
• How about Neil Stephenson’s Anathem? Found it. OK, let that be my first Kindle purchase then.
• OK, I can subscribe to TechCrunch. Oh, wait, that will cost me $1.99/month? Arrington, you’ll stay on my Mac.
• Great, there is a blog called “Hey, It’s Free!” so I’ll subscribe to that, whatever it is. Oh, Hey It’s Free costs $0.99/month? Hmm, OK.

The Kindle User Agent string is:


"Mozilla/4.0 (compatible; Linux 2.6.22) NetFront/3.4 Kindle/2.1 (screen 600x800)"

The main challenge this Kindle is going to face is drool and repeated punching by my son, which will no doubt occur while my wife is trying to read “Patient Parenting Tips” on it. I’ll let you know how it holds up.

CSS Sprites, minify, compression and 95 from YSlow.

6View Commentshttp://warwickp.com/2009/11/css-sprites-minify-compression-and-95-from-yslow/CSS+Sprites%2C+minify%2C+compression+and+95+from+YSlow.2009-11-11+18%3A02%3A49warwickp

Nov 11 2009 (optimization)

My consulting company does enough WordPress work that I thought I would create a separate brand for our WordPress services. WP Icon was born. Do not laugh at my stock photos! In creating a quick site for this service, I thought I’d have some fun and try to get this new site to be as fast as possible with an hour or so’s work.

YSlow screen capture showing 95% A score for WPIcon.com

Back end performance

Briefly, I am using the Nginx Memcache module to fetch cached page content directly from Memcache without touching WordPress, or even the PHP interpreter. More on this setup in a future post.

Front end performance

I’ll focus on this here. I have met and heard Steve Souders, author of YSlow, give a talk on front end site performance. While some view YSlow as promoting either flawed, or at least premature, optimization techniques, I find the tool useful and thought I’d try to give my new site a 100% score,  for fun. I really haven’t done much front end work recently. Here are the optimizations I made, some suggested by YSlow, some not.

CSS Sprites

I had not used the CSS sprite technique so I gave it a try. The basic idea is to put all images into a single image file and use CSS to assign visible portions of this image to various objects on the page. Theoretically you eliminate the file format / meta data overhead of each image, plus the network and protocol overhead of negotiating and loading each image. It’s pretty finicky work, and I’m not sure if I actually achieved optimization here. The numbers:

• Combined size of individual images (24 bit non-interlaced PNG): 295K
• Size of (unoptimized) combined sprite: 405K (+138%)

Additionally if I look at the page load size using FireBug, it tells me that Firefox has loaded multiple copies of the sprite image, bringing the total page size to 2.43MB. Is FireBug mistinterpreting what Firefox is doing?

PNG Optmization

I used Yahoo’s Smush.it tool to reduce the filesize of my sprite PNG.

Content Compression

Technically a backend optimization, but the optimization is delivered on the front end, I am using the Nginx gzip module to compress static content between the server and the client (browser). The numbers:

Minify JavaScript and CSS

The idea of minification is to optimize the delivery of JS and CSS content by combining the files into a single file and then stripping out whitespace and superfluous content (such as comments, etc). Since I am using WordPress, I used the handy WP Minify plugin. The numbers:

Move JavaScript to the bottom and keep CSS at the top

The CSS is minified and delivered in the head of the document by the plugin mentioned above. As for external JavaScripts, I am using the Google Analytics WordPress plugin which puts the GA JavaScript in the WordPress footer call.

HTTP Expires Headers

Because of the way my caching system works on the server, between Nginx and Varnish, all static and cacheable content types are given future Expires headers. This effectively tells the browser not to reload each object on each page load, making the site seem faster to the user.

Problems with YSlow

YSlow makes the suggestion that you use a Content Delivery Network to serve static objects. This is a valid suggestion for a large site, which wpicon.com is not. Regardless, YSlow has a flaw here, in that it does not recognize mapped/CNAME CDN URLs. So sometimes even if you are using a CDN, but on a URL scheme which is a subdomain of your site’s own domain, YSlow does not pick this up.

Conclusion

A quick excuse to play with sprites and YSlow, I got a YSlow score of 95 for WPIcon.com with about 30 minutes work.