The General Grant method

I referred to the General Grant method, with no explaination. It is not a completely accurate analogy because we are not at war, but I will have to cover Money Ball later.

Oh, and I guess it is on now, I made this a favorite. I must admit I am terrible at analyzing individual companies, but I am familiar with telecom. When I get off topic feel free to invite me to go back to METAR.…

He overwhelmed the south with more men, not necessarily good men, or good tactics, just lots and lots of men.


"Grant responded to this recognition by becoming a more ruthless and more grimly determined strategic leader. With his remarkable ability to see the larger picture he understood that the war would be won not by a few clever battlefield maneuvers, but by prolonged and bloody pressure on the enemy. The north had many more men than the south and its casualties could be made up fairly readily while those of the Confederacy could not. Grant saw the long term implications of this situation. He took the battle to the Confederates whenever he saw the chance to gain a strategic advantage and fought them stubbornly until they withdrew or surrendered.

It was a heartbreaking strategy and a terrible one for a man who so hated bloodshed that he refused even go game hunting. Rationals, whether directive or nondirective, see neither honor nor glory in bloodshed and find no satisfaction in it. They are never thrilled, as are their utilitarian cousins the Artisans, by skirmishes or warfare. But there was iron behind Grant’s casual and sloppy dress and manner. He insisted on fighting almost every battle to the bitter and bloody end, grinding down the enemy forces until they were helpless or exhausted. It was Ulysses Grant who became famous for having said “I propose to fight it out on this line if it takes all summer.” The statement has become a symbol of grim, unyielding determination, and as such it characterizes Grant extremely well. His usual facial expression fit equally well, for Grant was said to wear an expression on his face which indicated that he was considering driving his head through a brick wall, and that he did not intend to be prevented from doing so."

Google has about 50,000 employees, very good ones, led well, AT&T has 250,000 employees, above average ones, not led so well. The Google employees have better training and more experience with software, the AT&T employees are technical people, trained to solve problems, but with little software training.

There is one difference, the AT&T employees understand, what was is gone. Nobody believes the past matters,either they have a plan to retire or adapt, for them it is a critical battle, to lose is to lose the life that they have built.

AT&T can throw massive amounts of people at a problem, but Google, and Apple can throw massive amounts of money at a problem. In fact Apple could buy AT&T cash.

For AT&T the game is big data. AT&T has the advantage of holding huge lakes of data that no one else has, and has the network to move it. What AT &T does not have is the programming talent to mine it. I am sure the executives lay awake at night worrying about it because this and the NPS (Net Promoter Scores) are what keep AT&T under valued. From the inside, I can tell you that the software systems within AT&T are horrible, and they are the cause of many ills. This is not lost on the executive leadership and they are moving to correct it. However, they cannot go into a bidding war for talent with Apple, Google, and Amazon, they must use the troops they have.


Welcome to the board Qazu. I can’t wait to hear more of your thoughts.



If AT&T really wanted to solve their software problems there is a way to “buy” into a solution without getting into a bidding war.

As you stated, the talent behind the Apples and Googles of the world are a lot more experienced. In fact, they can outperform your average “C” level employee by a factor of 10. But, those “A” level employees are not exactly 10x as expensive - really only 2-3x as much, tops. However, you are correct in your assertion that your firm may not be able to hire them, but for reasons unrelated to cost.

An “A” level person wants to work in an environment where they can thrive and engage in creative, cutting edge-work that will thrill a user-base appreciative of their work. But most important of all, “A” level people want to work with other “A” level people. If your firm has scores of average so-so talent the "A"s just aren’t going to want to work there - no how matter how much you’ll say you’ll pay them.

While you may not be able to hire those "A"s directly, you can find a consulting partner who has them. I speak to this from experience. I work for a technology consulting company that hires and retains a lot of "A"s. We do really great work - a team of 3-4 of our guys can outperform a team of 30 in an offshore model and I mean that as no joke.

My advise to AT&T would be to find one of those consulting companies and partner with them. Not the Accentures of the world - they’ll be happy to sign a nice, large contract with you and then proceed to the nearest high level academic institution where they can back up the bus, arm those new grads with a weeks worth of “professional service” training and deliver them to your doorstep ready to take on your project. Instead, find a smaller, more “boutique” firm that employees seasoned software veterans that live in a DevOps world and follow an agile project methodology. Don’t go too large or you’ll see “dilution” in their talent pool. A market cap of $40-70M will do. The "A"s that work in those organizations will be happy to take on your software project because 1) they can continue to work with other "A"s and 2) they’ll transition off your software project after the work is complete (i.e. they can move on to other exciting things).

Best of luck,


It is worse than that. In software, a well-integrated team of 10 talented people can outperform an average team of hundreds or even thousands. This is not just because of the talent differential, but because a small team communicates and responds in a way that a large one can’t. Large teams only exist because of the persistence of the notion that 9 women can have a baby in a month.


I forgot to mention there is another dynamic at play, and that is with your “B” level players. The above average (but unremarkable) “B” level people have a hard time recruiting or accepting “A” players into their culture. A “B” person will say a new candidate is overqualified or too expensive or have some other reason why the person won’t fit in. So, the culture is stymied and doesn’t progress from that standpoint either. It is a strange phenomenon I’ve seen in many business environments. I know this info may not be entirely helpful, but it is another observation/commentary on the topic.




I think AT&T has tried this, over and over again. What we have found is the company is dynamic. The software project that rolls out and is so cool today, will not work tomorrow.

Think of this, in 1999, AT&T was a long distance company.

That company is a small, a core, but small part of AT&T today. 4 years ago, AT&T was going all wireless, in June, July at the latest, AT&T will be the largest video content provider on the planet.

The software systems are not only changing, they are a core asset of the company. AT&T can no more outsource its software development than Google can outsource its hit ranking algorithm, or Apple can outsource its look and feel design.

Because of the speed of business, AT&T must use agile program management, but the fact is, it will never end. The software systems must be torn apart and rebuilt on the fly. We tend to think of upgrades like we would about rebuilding a car. We pull it into the shop and pull the engine out and drop in a new one.

Today’s environment isn’t so easy. It is more like we are on a road trip, we start out with a '78 Suburban, but we need to arrive in a Tesla, and no stopping in the way.

I understand about small teams and “A” level talent. However, I have seen teams come together adhoc and make some pretty amazing things happen in the last few months. Not software, some complicated stuff. (We added about 3.6 million to the bottom line with no capital outlay, no new staff, no over time. It was a true grass roots cooperative effort.) I will go into this a little later when I clean up my post on Money Ball.

Back to software. I am just now learning software, it is very late in my career, I only have 10 years left. However, I can already see what I will do with it. Personally I probably will only rough in a proto type, enough to pitch the idea then attempt to build a team to implement the idea. The fact is, only someone with intimate knowledge of the internal working of a telecom power plant and with an interst in green energy and grid scale storage would be interested enough to even try.

This type of thing is everywhere inside of the organization. It goes from Router Engineers who have outside interests that understand enough to creat new and novel API’s to sales reps who see and solve problems by building tools on the fly. (Anyone who has been in a AT&T store knows how long the process takes, most of the excess time is horrible software.)

The executives called for a skills pivot last year. The response from the front line people was quicker and more broad that they expected. They offered a generous retirement package with the warning that if enough didn’t go, the targets would be met, and the terms wouldn’t be as generous. (I hope so for your sake, the emperor is not as forgiving as I am.) on the other hand, more people, even late career people like myself took up the offer on training and are moving forward at a pretty good clip.

I suspect that there is a lot of undeveloped talent within AT&T, and more energy than the leadership expected. While there are still significant issues, both with software and culture, there is a lot of change going on, and growing sense of excitement about the changes happening.

The only big problem that gives me indigestion is this: I hate TV, additionally, Southwestern Bell worked really really hard to get into long distance right before it became worthless, and it’s new iteration AT&T managed to get wireless right, just about the time the margins start compressing. Now AT&T is getting into Entertainment distribution at a time when my son doesn’t have a TV and doesn’t want one. (I don’t have one either.)

I do understand the business case for Direct TV and I believe it will be a really big deal, and that it is just part of the whole, which is tied together by software.

Qazulight (Sorry for the ramble, maybe one day I will clean it up.)


It is easy to classify people as A, B or C. But in my experience it can take many months to discover that an apparent A is actually a C (or worse) and vice versa.
As to the issue of setting out in a 78 Suburban and arriving in a Tesla - it can be achieved more easily by applying a strong, long term orientated discipline to your development.
I wrote an accounting application in 1992 which is still in use today, although the software platform is radically different. Because the underlying design was logical and built to be extensible we were able to successfully transport it to new environments as technology advanced.
Most programmers do not write code in this way because they are under pressure to produce results quickly and cheaply. But this will nearly always work out long term expensive as the code is too poorly constructed to be upgraded and the whole thing has to be discarded.


By the way these days I do arrive in a Tesla, as a result of selling the software business two years ago. Fantastic car (and pretty rare in Wales, where I live).



Thank you for sharing the details of your journey! It is a reminder that we all have unique experiences to share from life’s rich history.

Reflecting on some of your comments, software is frequently never built to last and there are several good reasons to explain why. In fact, knowing you will be tossing your software platform out the door and rewriting it wholesale every 5-10 years can be an acceptable business strategy, especially for test market offerings or for start ups that lack sufficient resources.

But as you are now paying witness, when a software product becomes part of your core capability, throwing the whole thing out the door every couple of years looks a lot less appealing. You’ve got to build your software product to last. And when I say “last”, I mean build it so it can shrink, grow, adapt and morph itself not only from a dolphin into a tuna, but from a dolphin into a Lamborghini. Your analogy of replacing the engine in your car and morphing every other attribute along the way is a good one. We do this with software all the time, and we do it while the engine is still running. It really can be done.

I stand by my assertion that a good consulting partner can help a non-traditional delivery organization appreciate the differences between “building software to last” vs “building software to specification”. Having these insights will really help you raise the bar for the organization.

Frankly, the software practitioners who understood these concepts rarely existed prior to 2000 and the idea only started to get traction around 2005. In 2015 there is now more of a vocal minority, but finding professionals who understand these differences and exhibit the behaviors are still going to be a challenge. In the past I blamed Microsoft for “dumbing it down” for the software professionals, but even they are starting to adapt and absorb some of the engineering, design, testing and operations rational into its vernacular.

There is a book I recommend to anyone interested in becoming a better practitioner in the software field. That book is Clean Code by Robert “Uncle Bob” Martin. If you haven’t read it already, I highly recommend checking it out:…

The book isn’t the be-all end-all book you’ll ever need, but it is a great place to get started.




You give me too much credit. I am a tech, a user, not a provider. I can see the big picture but cannot participate in the building. Yet.

I will take and I believe pass “Programming is for everyone” in the next 10 weeks. After that, I will go back and try Udacity “Intro to Computer Science” (I learn poorly if at all on the Udacity platform, but it the chosen platform by my company at this time.) the goal is not to become a programmer, I am too old for that, however the rest of the training is directed at Big Data. My goal is financial analysts, specialinpzing in using software platforms to identify and exploit business opportunities. My hope is that I can then use that knowledge in my hobbies, The study of MACRO economics and equities trading and business investment.

By the way, if you have someone in your life that may be interested in programming, Programing is for everyone" is a free Coursera course, the book is free also. Although, I downloaded it via I-Books and the I-Book contains the videos. Also, the language is Python 2.7 not 3.0 and it is free along with the IDLE GUI.

The course starts tomorrow although if you miss it there will be another.


1 Like

For Big Data, I might also recommend John Bates Thingalytics.

1 Like