The Economist: Deflating IT | Hacker News Reader

The Economist: Deflating IT

economist.com

43 pointsstartingup12 comments

I'm a manager in an in-house corporate IT dept. Far from running scared from this kind of thing I'm constantly trying to get the rest of the business to except off the self software for all the little tools that they ask for so that we can focus on the key business differentiators.

Problem is, they're a big company and they want slick corporate branded stuff that integrates with our ActiveDirectory and where the drop downs have sucked values in from other apps so it all just works. Which is fine, only, it costs loads and its in most cases it's not worth it.

I think it'll be hard for a lot of big corporates to accept that it's better (for example) to adjust their business to Zoho CRM than to hire people to adjust SAP CRM to their business. I think a few startups will have to adopt things like Zoho and then grow big, before the incumbents change.

I think businesses have to pick and choose, but: if inhouse CRM is part of the competitive advantage, then it makes sense. If it isn't, and the only goal is for it to be acceptable, then it makes more sense to license an existing product.

Except for state/government institutions, which very often have many other regulatory hurdles that commercial software doesn't handle well. That's one of the ways in which regulation can be very expensive.

While I am a bit worried about how scatterbrained AdventNet seems to be - I do think that they have a chance to disrupt a very inefficient and parasitic software eco-system that exists in Business Software. They only need to figure out the integration piece and the fact that they have played with some Microsoft Active Directory tools and embrace outside identity systems may help them get there.

I have found that the biggest draw to "integrated" software is not having a new set of credentials to login into it.

I am finding some success with technologies that can play nice with various directories. I am currently playing with Novell's EDirectory and its Identity Management solutions and it surprises me how few companies focus on what I see as one of the central problems in business IT: the seamless integration between the company's CRM, the team's contacts and the directories that power each business solution.

There is a huge need for a business class 'identity-in-the-cloud' solution that speak to many internal systems using software hooks/drivers that are very transparent, have good auditing and can be controlled by the company/customer.

I bet a large percentage of software projects at companies have overlapping user management sub-systems that could be better addressed by an identity-in-the-cloud system like the ones emerging from Microsoft, Google et al.

And for those who say that your company has too many secrets to possibly use an outside solution for identity - get over it, it is not true. You have many people forwarding things to their Gmail accounts and many mobile devices around the country without even a pin code to secure them. Your sense of security is an illusion, embrace solutions that can provide complete audits of what happens with them, then you will know how insecure you really are and plan around the fact that total security is impossible. And your internal development team will never reach the self-auditing stage of the solution because as soon as it half works they are pulled to something else.

"Microsoft, having grown up making money selling traditional software, will have a hard time being successful online, he argues. He thinks Google is unlikely to invest much in enterprise applications, because profit per employee is much lower than in its advertisement-financed consumer business."

I hope for his own good that he's not staking the future of his business on those basic assumptions. If Google and Microsoft were to ramp up their development of online enterprise software (which IMHO will undoubtedly happen), he'll have very little room to maneuver and no time to react.

Google & Microsoft will both do very well, indeed. Even then we will have a decent business if we execute well.

First, we are much smaller, so it takes a lot less to be successful. Second, even in the face of the most successful business on the planet, Intuit and Adobe succeeded well. Third, monopolies are exceptionally rare, and fairly transient - so Microsoft's Windows/Office monopoly is likely to prove to be the exception than the rule, even for Microsoft itself. Linux on the server side, Mac on the client, Firefox/Chrome in browsers ... all point to that.

Dell is an outstanding success, and still only holds somewhere under 20% of the global PC market. That's what I expect to be true in cloud computing.

Bottomline: there are going to be many players with many successful strategies.

IT needs are a moving target. I don't see how you can commoditize custom software.

You refactor your product to need as little customization as possible, and make it easy to extend and modify. The more OSS you can build on, the more money you save. One IT employee is easily $100K/yr after benefits, so if you can eliminate that much custom work, that's what you can charge. I'm not saying that this is easy to do.

I was thinking that understanding the problem (the business's "IT needs") is hard to commoditize, and can't be done with 100% automation (unless all businesses were cookie-cutter, automatically formally describable - perhaps almost true for some franchise-style businesses).

But much can be done without reaching perfection.

He thinks Google is unlikely to invest much in enterprise applications, because profit per employee is much lower than in its advertisement-financed consumer business.

I know this isn't the article's centre of gravity, but...

It bugs me to hear this thrown around so easily. There is no existing business model that produces that type of cash from online advertising. Search is the exception. Not the rule. Google's hedging bets, tying their hands but at the end of the day, these free web services will either need to be:

1. subsidised by a giant (like google does now) for whatever reasons they can get together

2. very cheap to run. IE web-services will replace companies many times their size, taking only a fraction over the revenues. You can always get some revenue from ads.

3. paid for by users

4. a business model will emerge that will do for apps what overture/adwords did for search.

I wouldn't be putting all my shells on 4. Unless i'm missing something.

Gmail is probably one of the 'best-bet' ad-supported apps. If it was a yearbook, they might get 'most likely to succeed' Anyone have an idea how much they make?

> “I was convinced that something big was about to happen with software in India,” he explains. “I wanted to be part of it.”

> Customers must be convinced that they can entrust their business to the software.

The first kind of convincing is harder.

disruptive adoption: http://news.ycombinator.com/item?id=297597

Deflating IT? Oversimplifying IT, rather.

Reply on news.ycombinator.com