CTO vs VP of Engineering

I have discussed this topic before but felt that Fred Wilson really captured the essence of CTO vs VP of Engineering with his write up here http://www.avc.com/a_vc/2011/10/vp-engineering-vs-cto.html

I am solidly in the VP of Engineering category.

A few tidbits on the VP of Engineering category from my experience doing that job for the last 8-10 years.

First its really important for the VP of Engineering to really understand and even help shape the business goals. A huge part of the job is mapping an engineering roadmap to the business roadmap and there are frequently critical tradeoffs between the two that can make or break the business.

Second I think its good for a VP of Engineering to have one particular technical focus that they “go deep on”. This is hard as they get pulled further away from code but important as a component of their leadership for a technical team (and obviously helps good technical decision making).

I personally try and stay as up to date as I can on databases as the heart of nearly all software. I have been managing Ruby On Rails teams now and I make sure I do some coding but would not hire myself as a RoR coder. My DB background and ongoing learning and playing gives me a technical area I can contribute on and frankly its one not many RoR devs enjoy!

More to the power of Kahn Academy

Kahn Academy is obviously a truly amazing internet property that captures the web at its purest and idealistic best.
(Along with Wikipedia and a short list of others.)

I found this article about how they determine “mastery” by one of their lead developers really interesting.

http://david-hu.com/2011/11/02/how-khan-academy-is-using-machine-learning-to-assess-student-mastery.html

It caught my attention partially because I have been reading up on Ruby powered AI lately.
(Its become SO easy to get some basic AI into your app. Just go do it!)

But more interesting is the amazing dataset that they are assembling about learning effectiveness.
By recording the progress of millions of students in a centralized and systematic format they are creating the richest dataset on how people learn that has ever been created. Based on their values and the article I can only assume its public data.

I look forward with anticipation to the magic that will come out of this data set. I really need to start my own search.

Digital to Meatspace

Implied in User Driven Design is software.
Some sort of tool to help the user find something they want through interactions.
Of course once that interaction is complete you still have to produce it and get it to them.

Here is a cool short video of a talk by Mint Digital (whom I have been working with a ton in the last year) on just that step. Get the physical into the customers hands.
In this case it’s instagram pics to fridge magnets.

Cool take on User Driven Design

The wide open space of UDD (User Driven Design) is still fascinating to me but its been a while since I have seen something new. (I need to start creating my own since Rails apps are so easy to build and Heroku makes hosting a true snap.)

But I digress. A key challenge in UDD is defining the space of options the user can make so that is both wide, but structured enough to ensure good results with nearly any selections. A team at Cornell has brought genetic algorithms to bear on this problem.

Start with a 3D shape. Genetic Algorithms pick a successive generation. Pick the ones you like and boom a new generation is born selecting for those.

I am not sure the results are so sexy but for me the process is.

http://endlessforms.com/

Yin and Yang? SQL and NoSQL?

There was an interesting writeup on InnoDB adding an integrated memcached component which brings along some of the benefits of a document/NoSQL DB. http://blogs.innodb.com/wp/2011/04/nosql-to-innodb-with-memcached/

This actually seems inevitable and like a step upon the “one DB, pick your SQL/NoSQL depending on the problem” path.

I hope someone (wish I had time to make it me) does a real hard core analysis of the performance differences.
There are applications for SQL based databases and there are cases better suited to document DBs.
Can a SQL DB be retrofitted to get 80% of the document DBs advantages? If so it feels inevitable that that will dominate.
There is simply too many things straight SQL is better for and too much expertise in the world for documents to be the predominate solution.

Maybe its not that easy though.

POTS is the new POTS?

Back when I was helping to build our spin out from Lucent I loved the term “POTS” short for “plain old telephone service”. Partly I liked it because the word itself sounded like what it was. Reliable. Plain. Understood. Boring.

Well old copper telephone service POTS is still on its long slow decline but its talked about even less. I think its time to repurpose the word for this decades POTS.
“Plain old TELEVISION service”.

The days of the cable company provided, coax, monolithic channel subscription (most of which you have literally never watched) POTS are not over… but they are without a doubt in decline. This decline will happen faster.

Management is …

Old but brilliant article from the Atlantic on management consulting: http://www.theatlantic.com/magazine/archive/2006/06/the-management-myth/4883/

My favorite quote, “Consultants’ recommendations have the same semantic properties as campaign promises: it’s almost freakish if they are remembered in the following year.”

Generalists and Specialists

An article on marketing I read recently got me thinking ~again~ about generalists and specialists.
http://startup-marketing.com/where-are-all-the-growth-hackers/. The article makes the case for a “growth hacker” rather than a CMO for most startups.
Makes a lot of sense to me. I have been deep in an organization with a CMO that was successful elsewhere but was ill-equipped for the marketing hacking we really needed.

The concept applies to all aspects of a small nimble organization, not just the marketing. Technical, biz dev, community outreach, support. I specifically say “small nimble organization” because even progressive bigger companies are starting to organize around flat smaller nimble groups under relentless market forces.

I would argue its more about the maturity of the product than the size of the organization. Early in a products lifecycle you need generalists who can try lots of things, look for options far and wide and take smart risks. For very mature products and markets you need people who go very deep, optimize relentlessly and protect the status quo.

The thing is, product lifecycles are getting shorter every day. It appears to me this will reward the creative generalist (hacker) though you could also argue the product builders need more unique skills to differentiate.

I have a half written “Why do (Tech) Managers Suck” which applies some of these questions to technical management in particular. The closer to the work you are, the better job you can do advising your team and making specific technical and architectural decisions. On the other hand, if you are focused and shoulder deep in the technical details its really easy to take a massive wrong turn organizationally or simply build the wrong thing. Technical management is particularly sensitive to this effect as the details of software technology seem particularly deep and fast moving.

So how long is the lifecycle of the product you are building?
Are you going deep or wide?

Online Consumer Protest

I came across info an interesting online consumer protest today about the pricing of e-books.

http://reviews.cnet.com/8301-18438_7-20051201-82.html?part=rss&subj=news&tag=2547-1_3-0-20

Its easy and cheap to organize these sort of things so I am surprised there are not a lot more of them. There probably will be soon. How long until the first company is completely put out of business by a consumer guerrilla protest?

The end of the article brings up a separate matter. Do online ratings matter?
I think they are incredibly important on the way up the food-chain (a few bad ratings in the app store can doom you to obscurity forever) but once you have arrived it matters much, much less.

So is this another reinforcement of the age old “rich get richer”?

App Store Exhaustion

After reading in the last 20 minutes about all the breathless excitement over the launch of Amazon Android Store
and then how Google has dropped the ball and there have been almost no sales on the Chrome Web Store, and keeping in mind that Apple is about to launch their own MacOS Appstore to supplement their mobile one… I have SERIOUS APP STORE FATIGUE.

My prediction is 90% of the app stores launched in the next year will be dead before their first birthday. There just isnt consumer interest and the iphone appstore has taken too much of “oh neat” wind out of their sails already. There will be vibrant marketplaces and there will be more than one. Thats good for software developers… the thing is there won’t be more than two or three really viable ones. Building something? You are probably placing a bet on one.