Hacker News Reader

Comment by wayoutthere | Hacker News Reader

I don't think I've ever heard of an ERP implementation that wasn't an absolute clusterfuck. My impression is that while most folks from the integrators know their segment of the product pretty well, they don't have deep technical skills in a general sense or know much about other modules of the product. Also that the salespeople know almost nothing about the product and promise that it can do things it cannot. The customer eventually figures this out and ends up ripping out and re-implement tools from other vendors.

It's a common joke for senior technology leaders at industrial companies to ask eachother "so what was the first ERP implementation you got fired for?"

View on HN

I find it impossible to believe that an ERP implementation can ever "go smoothly".

At best, even if the system does what it does perfectly, it's still a classic "diffusion of innovation" (1) problem where you have multiple stakeholders with wildly varying levels of acceptance. This is always challenging. As those of you who make a living at this stuff recommend, it's better to go 100% with the ERP's way and not customize. Well, that can be extremely difficult or impossible in an environment filled with stubborn beancounter/battle-axe personalities.

Most importantly, however, the people who make the purchasing and sign-off decisions ARE NOT the ones who use the thing. There are going to be problems and those problems will not usually bubble back up to the consultant/vendor. People will just have their noses pushed to the grindstone until the job gets done-- never mind the tedium, never mind the countless little mistakes. New people don't get trained, they just get put in front of the the thing to hunt and peck through it, sometimes with the assistance of a surly veteran of the system, sometimes with no one to help them. There are thousands upon thousands of people sitting in cubicles right now using Oracle EBS with it's horrible grey java applet UI and ridiculous inscrutable query functionality. They've been sitting there suffering it since the 90's. Who listens to their screams? Not Oracle, not exec that bought the thing! Or maybe they're silent and they've resigned themselves to their fate?

(1) https://en.wikipedia.org/wiki/Diffusion_of_innovations

Totally agree!

My clients are all ripping that crap out because nobody uses them correctly anyway. Most of the data in those systems is so error-ridden they're almost dangerous to use. Newer cloud-based systems offer a much more intuitive UI, which ultimately reduces bad data.

Oracle and SAP are basically bad words at this point: you don't say them if you want to get hired.

This is not necessarily true. I find accounts people who have grown up with these systems find Netsuite (a cloud system) slow, cumbersome and unintuitive. Fast repetitive data entry always seems faster in desktop clients. We are talking about professional full- time users, the initial easier learning curve soon loses its advantage once you have learnt the keystrokes.

As for data integrity, I worked in very large SAP implementations without data integrity issues despite thousands of users. I have worked with cloud systems that are full of rubbish.

> I find it impossible to believe that an ERP implementation can ever "go smoothly".

ERP implementations are like kitchen remodels. Even when they go mostly right, you hate the people doing the work because they've messed up the thing that's at the center of your home, made you eat take-out for weeks, and dust has gone everywhere.

I have been on one that was pretty flawless. SAP at the time, with full costing/ manufacturing, multi-currency, multi- country. I should write a blog about that, shouldn't I? I only recall two minor faults.

Yes, please

I been on the system implementor side of the ERP for long time and done many ERP projects(SAP). I have also never seen ERP implementation that has gone smoothly and everybody involed has been 100% happy. I think implementing ERP is 98% change management project and 2% IT project or something like that.

I recently also wrote my thoughts on what ERP is. Maybe its interesting to someone. https://www.integrated.ee/posts/the-erp-software/

On the tech strategy side, we're moving everyone we can off of traditional ERPs and into cloud-based systems. Salesforce is largely taking their place as the CRM, with purpose-built solutions (often leveraging cloud-based ML models) for downstream functions. All the consulting companies have process maps, which are honestly where most of the value of traditional ERP comes in.

TCO factoring in implementation costs is a fraction of a traditional ERP and is far easier to keep up to date. Most ERP implementations still have a lag of a month or so before data is consistent.

You make some very broad statements here...

Cloud ERPs stay up to day whether you are ready for the change or not, it is not all rosy when they break important things at critical times of year. Traditional ERPs often don't get upgraded because there is no need. It is simply not worth the testing for the new 'features'. Lots of managers value the stability.

TCO is certainly not necessarily less, it is common for a company to move onto an ERP and stick with it for a decade or so. One off licensing can be much cheaper over this period than some cloud systems. I saw a system where TCO of was cheaper with the trad system after only 4 years.

Cloud based ML models are a bit buzz-wordy for my tastes

> "so what was the first ERP implementation you got fired for?"

At my last job we underwent a transition from one ERP to another. Halfway through it my boss resigned and we got somebody new, mostly on the basis of his experience doing the same at another company. I heard a rumor later that this other company had nearly gone bankrupt because of the roughness of their transition. Things weren't looking good for ours either when I left.

This is pretty much par for the course; corporate IT at the highest levels is a game of musical chairs where everyone fails up.

Having been through a few ERP projects (on the buying side) I'd say the scale runs from 'painful' to 'disastrous'.

One large ERP project I worked on barely made it past the kick off meeting.... a number of us escalated that we were not happy with the implementation partner so our boss (bravely) recommended that the entire project get canned - which it did.

Reply on news.ycombinator.com