Watching Larry Ellison Become Larry Ellison | Hacker News Reader

Watching Larry Ellison Become Larry Ellison

steveblank.com

148 pointssajid77 comments

For anyone interested in learning more about Larry Ellison and Oracle, I'd highly suggest "The difference between god and Larry Ellison (god doesn't think he's Larry Ellison)."

Oracle is a 150 billion dollar company, similar in market cap to Intel. Once you've read histories on these people, it suddenly makes sense that Warren Buffet and Bill Gates hang out, while Steve Jobs and Larry Ellison were BFFs. Like with Jobs and Apple, Ellison and Oracle are aggressively polarizing.

I see a future of declines for Oracle; Salesforce.com is redefining and taking their market (corporate IT budgets). There is simply no way other than lock-in and being dicks that they'll be able to see the profitability they once had, let alone sales growth (unless by acquisition). You can see it already with their patent suits; they're out of ideas. But if you work in the business of software, Oracle's history is worth knowing.

Every enterprise I have worked at/with had Salesforce, but I've never seen or heard about anyone use it. (What do they even do?) I feel like it's one of those companies that's good at selling but not really building product(s) people actually use. Not sure same can be said about Oracle, I worked at many corporations where the tech teams were using Oracle DBs for simple 1-page microsites.

Mark Benioff, the CEO of Salesforce, started out at Oracle. Larry Ellison was his mentor. I think he was even at Larry Ellison's (3rd?) wedding. He left Oracle to start Salesforce.

Salesforce is a software layer on top of databases, often called CRM. It's the answer for "how do you sell a relational database to a guy that just needs a rolodex?" Salesforce actually uses Oracle databases to power their infrastructure, to this day.

Over the past 10 years, Salesforce has built out its own platform that basically turns Salesforce into a web-based version of visual fox pro. Which sounds bad, but consider that many "Salesforce.com administrators" are NOT programmers. Then to be taken more seriously as a custom development platform, to be able to add more advanced logic, they built their own scripting language (Apex) and bought Heroku.

So it started as (pretty much) a fancy rolodex but has evolved into a platform through which you can develop certain types of software and workflows that would be difficult with alternatives. "Difficult with alternatives" becomes less true every day (which is why I got away from pitching businesses on custom development with Salesforce and into pitching businesses on custom development with python and purely open solutions.) Despite having legitimate competition from open source, their sales and marketing has done a very good job at convincing their users that those don't exist, or that Salesforce invented them (cough heroku).

Maybe you've heard of Oracle or SAP consultants in the 90s making a very good living advising businesses how to set up and use these products? Salesforce.com is now that. And despite being "enterprise" they actually have very talented engineers. Also, critical mass. They have managed to build this ecosystem and everyone working in Corporate IT is drinking the kool aid. Google "Dreamforce 2016" if you want to see just how cult like it's all become. Like I say, they are going to replace Oracle. It's funny how much shared DNA they have though.

Well put. There are an amazing amount of opportunities in selling custom solutions not tied in to enterprise behemoths.

I noticed in the last couple of years as geographically and social circle unrelated friends who were in sales suddenly were working for SalesForce. I imagined they just ramped up their own sales force.

From a business standpoint they are big enough to just buy companies. From a longer term tech standpoint it is harder to guess what could happen. The analogy of drug companies doing leveraged buyout of pharmaceuticals with patents nearing a sunset date seems to apply here.

ApEx (Application Express) is an Oracle product, more like a visual RAD tool for the web. Did Salesforce really call their scripting language the same thing??

Yep. That's one of the reasons why searching for "ApEx" on Stack Overflow isn't fruitful.

The other is that nobody, if they can help it, uses ApEx.

> The other is that nobody, if they can help it, uses ApEx.

Oh oh. I worked for more than two years with a team that continued to expand their investment in Oracle and Apex.

When everything you know is Oracle, everything looks like a task for Oracle ;-)

That was my experience too, I worked on ApEx stuff for around 2 years.

It came free with the database license, so it got used.

I don't miss it.

I've had really positive experiences with APEX. If you already have an Oracle DB, it's hard to beat for quickly standing up data centric business apps.

I'm curious - what made your experience negative?

It's fine if you paint inside the lines. For quick CRUD stuff it was fine.

As soon as you need something that doesn't quite fit the limited palette, you need to go underneath and write PL/SQL.

Not as much fun.

In my experience, nothing stays pure CRUD forever.

Right, but it's not for developers. It's for users who would otherwise have knocked something together in Access. At least it gets them onto managed infrastructure rather than running it on their own PC for their workgroup...

I actually used to describe it as "Access, for the 90s web, for Oracle".

That said, I'd be surprised if many places have it as easily available to end users as Access.

Disclaimer: APEX dev team member here. Independent of APEX and even the Oracle database, if you are not using stored procedures, you are not using the database efficiently. See e.g. this presentation for a performance analysis: https://www.youtube.com/watch?v=jfkph4lXmKU

Yes, but you do have PL/SQL to fall back on. It can be very useful.

That sounds like a reference to the Azadian intermediate sex with a reversible vagina and ovum, from Ian M Bank's "Player of Games". [1]

[1] http://tvtropes.org/pmwiki/pmwiki.php/Literature/ThePlayerOf...

I was once burdened with building a complex software using a SaaS CRM as the framework. It was a terrible experience. Everything was slow. The database didn't allow complex schemas. I quit.

Now I build everything with Python on Bottle.

What are the good open source alternatives to CRM, or to ESP (ExactTarget)?

Last I checked CoreBOS (forked from vTiger, who forked from SugarCRM a few years earlier) seemed very promising, having cleaned up a lot of old php mess.

Also Odoo, formerly openERP, is something I would like to take a closer look at. Edit: especially now that they have updated their license terms.

Some of my colleagued evaluated Odoo recently and found it to be a slow, buggy mess. It has everything but even the hosted version did not work well at all when they used it with a few people. Anyone have a better experience?

Thanks for the late reply, thats nice to know.

My reason so far has been the AGPL license that I have shunned so far but I was really hoping for the next release that I understood should be under a liberal license.

AWS, postgres, and django.

Given a company that might buy licenses for 50 salesforce users, I make this case: That company can expect to pay $75,000 per year simply for licenses, that's before bringing in a salesforce developer for $75-250 per hour for setup and support.

So instead, give me the $75,000 to build a django web app that does everything you wanted salesforce for and then some + get better support than you would from Salesforce.

The trouble with solutions like Sugar and VTiger is they try to do what Salesforce does but they do it worse, and in the end just make a bad product. All of the complaints about Salesforce rollouts people talk about, 10x them for the open CRM alternatives.

If what you want is a fully integrated CRM platform (for whatever reason) Salesforce.com is the best choice. Otherwise build custom. Most companies are using 10% of the features of Salesforce, hence why I argue they could get away with something a lot simpler.

>Mark Benioff, the CEO of Salesforce, started out at Oracle. Larry Ellison was his mentor.

Mark Benioff worked at Apple also. He is very open about admitting Steve Jobs as his mentor. Although, nothing wrong in having two mentors.

Yeah, especially with those two! "My mentors were Steve Jobs and Larry Ellison." Who wouldn't want to be able to say that?

Foxpro! http://e.lvme.me/kdzw08x.jpg

Do you work in sales? Probably not I think.

Salesforce automates sales as well as customer support and was the first to do this effectively on the SaaS model. It's dominant among sales teams in mid-sized companies. You can be up and running in a few days including training. I find the interfaces a bit "enterprisey" (i.e., complex and clunky) but would not dream of having a sales team that did not track sales through SF or something similar. Personally I like the visibility it gives you into the sales pipeline--at any given time you can tell exactly what your current revenue is for the quarter and project into the future as well. It's very helpful when running a company.

I'm not in sales, however in my current role I get to talk to a lot of sales people and it seems that most of them don't (and don't have time to) use any fancy tools at all, especially the successful ones.

Also the last two companies I worked with were in the process of switching from SF to MS Dynamics. Not exactly sure to why, but seemed like one of those cases where the people at the top made the 'deal' and later people who actually had to use realized it didn't work for them.

My last few companies have been very heavy Salesforce users (mostly large companies). (I'm in sales). I've used a few CRM suites over the years and Salesforce is basically the "least worst" and with a bit of customization is actually quite usable and tweakable without requiring coding skills.

They also have a great non-profit policy where it's free to use, so a large-ish student exchange organization I volunteer for uses SFDC for their contact management, marketing , conference registration, and order accounting.

You think a company can host the largest tech-conference on earth, Dreamforce, with 200k attendees in SF while noone uses their software?

Whole companies have been built on top of the SFDC stack - Veeva, FinancialForce, Velocity, ServiceMax, CloudCraze, etc. That's how large the ecosystem of force.com is.

SFDC is for sales, marketing and service people. Call Centers use ServiceCloud all over the globe. Amazon, Bank of America, Japan Post, etc all use Salesforce CRM. Practically all big pharma companies use the SFDC stack through Veeva for their field forces. etc etc etc

The enterprise is a big world. All these consumer unicorns are neat, but there is much more to be done.

I take back everything I said including my explanation of why and how so many people use salesforce.

You are an awesome human being and thinking that because you don't see someone using something it must mean that no one uses it is a perfectly valid way of thinking.

Good on you and have a great day.

Your comment would be immensely better without the personal attack at the beginning. The personal attack is totally unnecessary. It is also why you are being down voted.

You didn't offend anyone because they haven't heard of Salesforce. You acted like a jerk, so no one wants you around. This is normal on HN and in any other typical human company.

I consulted (extremely briefly) with a startup whilst they were going through a Salesforce integration. I believe the end goal was to be able to get accurate data about... stuff... from which to generate reports, though this was never entirely clear. What it mostly seemed to involve was several developers writing vast swathes of code to map from the existing system to various fields in Salesforce via some sort of horrendous SOAP API, and a business analyst spending a lot of time deciding what those fields were. It seems absolutely impossible, to me, that the benefits gained from all of this in any way outweighed the amount of work and expense it took to do it (I left before they finished, if, indeed, they ever finished).

Based mostly on this story, I have successfully dissuaded several startups since from going anywhere near Salesforce, though in one case the exorbitant costs Salesforce quoted did a good job too.

Just because salesforce integration stinks, does not mean it is a lousy product.

The very basic premise is simple: it's a contact DB for sales teams, and helps manage relationships.

We replaced our old CRM (Peoplesoft) with Salesforce.com. Management did some surveys of users about how they liked it, the average score went from 2 (terrible) to 4 (good).

Some of this was because our Peoplesoft was slow, buggy and downright painful to use.

> What do they even do?

At the time they launched I worked for a CRM reseller.

The market was roughly segmented into two sorts of solutions: contact management and automation for individual sales people to run on a personal computer like the extremely popular ACT!, and more enterprise style client server CRM systems like Saleslogix, Peoplesoft, Sage, Pivotal, etc. These latter products were relatively expensive and complex to implement and integrate.

It was often remarked at the time that the majority of CRM implementation failed to deliver value. That matched my personal experience as well. Despite this abysmal success rate, companies kept adopting the systems because the alternative was individal sales agents keeping personal records, leading to very fragmented and inconsistent interactions with contacts and customers.

The initial Salesforce offering was quite limited, but it gave everyone another upgrade path away from individual contact managers like ACT!. My coworkers and I were initially dismissive due to data ownership/confidentiality concerns. Boy were we wrong. Salesforce blew up because it solved the problems with contact managers and simpler sales force automation software, but didn't impose the costly implementation and integration efforts the enterprise client server systems did.

And of course as tends to happen, now decades later the product is a massive offering that is frequently just as massive a headache to implement and integrate with.

This comment really surprised me. I'm not in sales either but half my company operates a hundred percent on salesforce and I genuinely believe it's incredibly valuable. It's essentially like jira + GitHub for sales and marketing people.

And when your audience isn't capable of building their own tools that's all the more valuable especially if you want them to quickly ramp onto as they already know. Once they attained Total Domination, there blossomed a large amount of add-ons and consulting firms that specialize in implementing on their platform which locked in their network effect.

At the most basic level they just killed creating an opinionated ticketing system for the sales motion which is about as Mission critical as it gets to any B2B company's bottom line.

> What do they even do?

Good question! The easy way to think about this is that it's a system of record or database for sales and marketing teams. The idea is you can keep an extensive and detailed history of every lead, contact, opportunity, and account you interact with. You can also report on and analyze this data to see how much you've sold over the last year, or how many leads turned into sales from a conference.

Salesforce is also a platform that you can build applications for. There's a really big ecosystem of developers who do nothing but build custom tooling and applications on top of Salesforce.

For example, my company recently started using a commission tool that is also a Salesforce application. So when you want to use it, you log into Salesforce and get a little custom "Commissions" tab.

The great part is that Salesforce's developer tooling has been hilariously inadequate for years. Many developers are not using version control at all. It's the wild west. Part of this is on Salesforce and the poor tooling they've provided. Google "Salesforce DX", and then realize that many things you might think of as standard are just being introduced.

"Every enterprise I have worked at/with had Salesforce, but I've never seen or heard about anyone use it. (What do they even do?) I feel like it's one of those companies that's good at selling but not really building product(s) people actually use."

Salesforce is a webapp. It's a big, giant, monster database driven webapp that should be immediately familiar to anyone who has built any kind of business Intranet.

The actual salesforce product is (IMO) not that interesting.

However it turns out that lots of huge companies were buying Oracle database deployments almost solely for the purpose of building webapps that do what salesforce does.

So, nobody is buying salesforce instead of Oracle - that's silly, since salesforce isn't a RDBMS. Instead, they are using salesforce instead of building their own internal webapp - which just happened to run on Oracle. Which means fewer Oracle licenses get sold.

Do I have that right ?

Unless you are in sales/marketing/customer support - why would you expect to working with people who use a CRM?

If you want a different perspective, "Softwar" is a good read. The author (Matthew Symonds) worked directly with Ellison to write the book (as opposed to other books which were written based on third party accounts and public information). He made a deal with Ellison that neither would alter the words of the other. Instead, Ellison could provide a reply or commentary wherever he disagreeed with Symonds. And Ellison did that, throughout the book.

I was expecting to see a lot of Oracle hate here, and looks like the train's already started. Regardless of your perspective of Oracle today, it's hard to dispute that Larry Ellison was a successful entrepreneur and businessman. He co-founded a $150B company, remained in the senior leadership position for decades, and guided the company towards extremely profitable outcomes for shareholders. You may not like him personally, you may not like his products, you may not like his company's work environment, but I'm sure any aspiring entrepreneur can learn a lot from Larry Ellison's example.

Completely agree. Oracle's success did not happen by chance. They have a great track record of navigating successive technology waves. Their strategy is often "bring a gun to a knife fight" like when they bought InnoDB out from under MySQL AB. Not to everyone's taste but they have outmaneuvered an amazing number of competitors over the years.

Ellison in the flesh is quite personable.

> Ellison in the flesh is quite personable.

Sounds oddly familiar...

I was in a couple of meetings with Larry Ellison when I worked for Oracle back last century. I found him to be very pleasant and pretty considerate. Couldn't reconcile this with the fear with which people regarded him.

I suppose the same could be said about Al Capone.

Nope. Al's business was short lived and he did not remain in command

That made me smile but as someone noted Capone didn't last that long.

Citizen Kane: "It's no trick to make a lot of money...if all you want to is make a lot of money."

A good book about Larry Ellison is “Softwar: An Intimate Portrait of Larry Ellison and Oracle” from 2004. It looks like Larry himself proofread the book as he offers comments and corrections at the bottom of some of the pages.

Oracle is unlike any other technology company in the world. The majority of their software offerings are not aimed at consumers, so many people (including programmers and IT-savvy folks) have a hard time understanding the company unless they’ve spent weeks studying one piece of their product stack.

I started working for Oracle in 1998. I thought I would be able to get up to speed quickly on everything Oracle. I was wrong. It took a long time to get a feeling of confidence. The main product, the Oracle database, is a vast monolith of technology and features that is constantly being updated. It has a documentation set that could easily fill a bookshelf. Taker a look: http://docs.oracle.com

Every so often someone takes at shot at Larry at a press conference or online, implying that Oracle’s time has passed, that some new hotshot company was going to overtake them in some way.

Spend some time studying Oracle and Larry Ellison, and you’ll understand why Larry always smiles at these questions.

I find myself thinking that recent years of change in the Linux world is partially to blame on Oracle and Red Hat having a falling out.

Those two used to have a fruitful partnership where RH suppiled the distro, and Oracle supplied the software that ran on top.

But right around the time RH released RHEL 6.0, Oracle up and forked RHEL into Oracle Linux and started offering support contracts for the whole stack. Since then RH seems to be on the warpath, tightening their control over the various sub-projects that RHEL is made of.

But didn't RedHat enter middleware first, snatching jboss under Oracle's nose?

Well neither was established in that market at the time, afaik. But Oracle Linux was not only Oracle entering onto RH turf, but using the very FOSS that was RH life blood against them.

Oracle was basically offering existing RH customers to consolidate their support contracts with Oracle rather than have the OS with RH and the database with Oracle.

I have a policy when assessing opportunities. Use of Oracle products is a deal breaker.

I have a similar policy, if I'm working on something and looking for solutions, the solution that doesn't involve Oracle wins by default.

I try to not be ideological about technology; oracle is a pretty decent database and for some cases is still superior to my favorite RDBMS (Postgres). But while the technology is fine, the licensing leads me to believe that the decision to use oracle was made on a golf course by a guy that doesn't even know how to set up his own email client, and doesn't care how anything works as long as he's next in line for CEO. And so yeah, I agree with you...Oracle is a deal breaker.

A golf course or an entirely different establishment?

Why? Is it the money?

The DBMS technology is excellent. I worked at Sybase from 1990 on and much as I loved the company Oracle beat us fair and square with features like Row Level Locking and Multi-version Concurrency Control. Oracle also had a healthier attitude about data corruption. Oracle fixed the bugs or didn't let them get in in the first place. Sybase on the other hand developed DBCC, which repaired the problems on live databases. I lost numerous tables on Sybase while developing products. This never happened after switching to Oracle in the late 90s.

One of the things I still like about Oracle is that no matter what data management problem you are solving there is generally some feature that helps you do it.

VPD (virtual private database) is a great example. I was working at a SaaS in the late 90s and we ran into a problem where we absolutely needed data from multiple tenants in a single database but still fully isolated between each tenant. Oracle had a feature called VPD that did exactly that. It saved our bacon on that project.

I have dealt with the Oracle license police and yes they can be jerks. But you can't fault their technology. The company deserves respect for what they have accomplished.

Edit: fixed typo.

Just one reason, among many.

From the article:

  He often said, “It’s not enough to win—all others must fail.”

I think capitalism taken to its extreme is counter-productive to society.

One of the big lies of capitalism is that competition is desirable.

Competition is desirable for the consumer, not for the producer.

In addition to this, I would add that competition can take many different forms. I think it's very useful when it inspires everyone to do better and everyone can shake hands afterwards because they enjoyed the journey. It's less than useless when people use it as an excuse to cut each other down without improving anyone's performance, which is disappointingly often.

I think it's very useful when it inspires everyone to do better and everyone can shake hands afterwards because they enjoyed the journey.

Do you have any examples of this happening in real life?

If you meant crony capitalism, sure. Then you'll use enough old boy tactics to ban your competition or give you a state-enforced monopoly or deal (effectively skewing the market in your favor and against your competition).

Really? I sure as hell can fault them - getting them to even acknowledge bugs is a gigantic PITA. Unless you are huge, getting their support people to respond or even understand issues is impossible. The CBO is a gigantic black box and can be incredibly hard to troubleshoot.

They have a CSO who has publicly stated they don't want bug reports or anyone to give them security exploits, and who once gave a talk "How Joshua DoSed Jericho: Cybersecrets of the Bible". [1]

There's plenty more to make you stay away from them at all costs, not the least of which are completely unethical business practices.

1. https://www.rsaconference.com/videos/how-joshua-dosed-jerich...

Fair point on support and security. Mary Ann Davidson's rant about security on the Oracle blog was indefensible. [1]

While I'm impressed by some of the Oracle engineering there's no question they also know how to squeeze customers just up to the breaking point.

[1] https://www.wired.com/2015/08/oracle-deletes-csos-screed-hac...

Not just the money although that's part of it. But I've dealt with Oracle as a database before and I found it to be a miserable experience. Figuring out why something was slow or wasn't working the way you thought it should involved paying Oracle a whole bunch of money to help you figure out which undocumented switch you needed to flip.

It may be solid but it's complex enough under the hood that you are practically required to pay ungodly amounts of money to them to be able to effectively use it. I'll never willingly use them again.

Now come on. Oracle and PostgreSQL and MySQL and the others all overlap a lot, but there's one thing that Oracle does much better than any of the others.

It makes money for the Oracle Corporation!

I think people forget that the Oracle database is just one of the many products that oracle has. I think Weblogic is the best application server around.

Why is it best? The only time I encountered it was in conversation of people that were using it and hated it, everywhere else it was either JBoss or just Tomcat (with new frameworks one doesn't need all the EJBs).

Are there any comparison chart with e.g. JBoss?

You don't need application servers. They just being bloat. Their ship has sailed.

Seem s like Weblogic came under Oracle's control when they bought BEA Systems, so it's just another good product they bought with their db money.

Weblogic was bad in 2007. It's a crime against humanity now.

(2014)

I found this post completely uninspiring, though I'm biased as I thing Steve Blank often does more damage than good with his recipes for success. Would love to hear the broader HN community's sentiment towards Steve, and the broader entrepreneurial snake oil in a book/consultant movement.

Steve Blank's snake oil??? "The 4 steps to the epiphany" was a transforming book for me and other entrepreneurs and the popular "The Lean Startup" follow up from Eric Ries a direct result of working with Steve. He genuinely cares about entrepreneurs and the valley ecosystem go watch "The secret history of Silicon Valley"

That's a pretty solid ad hominem case you got going there..

Reply on news.ycombinator.com