Hacker News Reader

Comment by beachy | Hacker News Reader

> Google [ ...] could spend about a year and build a vastly superior product to both Oracle and SAP that everybody would want. So why don't they?

This is laughable to anyone who has ever worked on development of an ERP product.

Here's a thought experiment. Let's take one part - one part only - of the whole ERP / enterprise software spectrum.

Let's pick HR for an example.

Now let's pick one part - only one part - of the whole HR product spectrum.

Let's pick recruitment. How about of Google did what you say, but just in this tiny space? How much overkill would that be?

It turns out, Google did launch a product in this space. It was called Google Hire.

By your measure, Google should have smashed the life out of this little part of the bigger enterprise software space.

Instead, they shut down within 2 years in a burst of failure.

View on HN

I agree they won't do it, but it isn't laughable. I've worked at another FANG and have used a lot of the recruitment software you reference. That FANG rebuilt the software from the ground up internally and have successfully made their process much more effective to the point of it being a competitive advantage.

There's a tradeoff with building software internally and I agree is almost always isn't worth it. But the advantage is that the software has only 1 customer: yourself. That allows a level of product market fit that just isn't achievable with off the shelf software.

ERP Implementations go well for 2 types of companies. Type 1 companies have well documented, logical consistent processes and procedures that server as clear spec for necessary ERP modifications.

Type 2 companies admit that everything they do is wrong and they are willing to change all of their processes and procedures to match those of the ERP system.

Everyone else is deluded into thinking they can modify the ERP to adapt to their business. 10 months into the development cycle, team members realize that the firm requirement that A customer has One customer number is incorrect, and it happens to be a major customer, and there is not a workaround. Sales insists that it is a one to one relationship most of the time and refuses to believe this a big deal. Accounts Receivable thinks they could develop a manual workaround. Modification proceed until the money runs out, and the consultant move on to the next victim. 10 years later the new CEO want to embark on a modernization project for internal backend systems and the cycle repeats.

Google Hire was an external acquisition: it used to be called Bebop, and it came along for the ride when Google hired its CEO Diane Greene to run its Cloud operation. After Diane got the axe, Hire was promptly nuked as well.

Assuming what you're saying is even true, do you think I (or other regular people) actually care about that explanation?

All I see is one day Google pushing for people to use Google hire and the next day shutting it down.

This feels like an unnecessarily harsh reply. It is easy to find that Diane Greene came onboard with a product called Bebop: https://venturebeat.com/2015/11/19/google-buys-diane-greenes.... It is easy to find articles as well about how bebop turned into Google Hire.

Personally I think this was fascinating. We got a glimpse at how Google decides to create products and later kill them. In this case - executive pet project that died with the executives departure.

I find it harsh too when google first tries to on-board me on a new product and shortly after that discontinues it. Most (all?) of the work from transitioning from an other product is onto me, and now I have to do it all over again.

Excuse me if I'm a bit harsh over that.

> This is laughable to anyone who has ever worked on development of an ERP product.

I was going to write a very similar sentence in another reply.

This is just a case of people not having the slightest clue of what's involved with large enterprise financials. Google has never developed any system that even resembles that.

To be fair, I don't think Google Hire's target customers were considering Oracle or SAP for hiring. They seemed to be more in the Greenhouse / Lever market.

But the whole reason they failed is because nobody wants that integration burden.

Reply on news.ycombinator.com