I have personal experience with other ERPs and Salesforce is the best of the lot in my opinion. With my current company, we have employees in Singapore, Europe, USA, and South Africa etc. who all interact with Salesforce on a daily basis. We use it for CRM, service support (VOIP with call tracking), marketing, financials (questionable decision), and analytics (Tableau). Where it shines in my opinion is that does automate 75% of business processes and if you pay the right fee, it's all seamless. The flow automation low code tool is a step in the right direction. Where it falters, is (a) its own hype machine gets in the way -- your customers are already locked-in, please stop the bullshit, SF is not easy to learn -- its data model requires experience to understand and architect properly (b) Mulesoft sucks, please stop pushing it (c) its web portal (lightning experience communities) needs work -- the security permissions model is hopeless convoluted like the fabled Gordian knot.
SF doesn’t do ERP level financials natively, but there is an industry of apps built on top of SF. Some try to turn SF into an ERP. Examples I’ve worked with in the last year:
-Rootstock
-Accounting Seed
-FinancialForce
That said, I design ERP integrations for a living. SF based ERPs are consistently the lowest quality and most limited ERPs we deal with. I would highly recommend that any finance professional seriously considering implementing a SF based ERP take the following steps:
-Burn a large pile of money
-Tell 2-5 of your finance employees to waste 30% of their work hours for 3-6 months
-Implement an non-SF based ERP
I'd like to ride along on an ERP integration some day.
I've been a part of a scaling start-up and it just feels like the kludge of tools small teams buy would be better served by a more integrated system. So I'd like to see what a big implementation looks like and figure out what pieces might be usable for small teams.
Yeah, it’s a surprisingly interesting space. ERPs are basically as complex as CRUD apps get, and because they model businesses, you learn a lot about how companies work along the way.
Wish I could take people on a ride-along, but because of the financial/business data involved it’s just not possible.
It 100% depends on the business. The ERP market is vast, and there is at least 1 ERP for every industry. I even worked with a country club management ERP once.
That said, I’m a big believer in going with what is popular. You’ll get more features, better support, better docs, and a community to ask questions to. And most importantly, more software will integrate with it. This is crucial. The last thing you want is your finance team having to manually key data from your satellite systems into your ERP.
Never ever let someone sell you on a niche ERP.
Other than industry, the big deciding factor is how big you are. SAP targets the Enterprise market. These ERPs are absurdly complex systems, and you’d only want one if you are big enough to have dedicated ERP consultants/admins.
That said, for general small businesses I would recommend:
-Quickbooks Online
-Xero
When you outgrow those, you’ll have a better idea what you want. Oracle NetSuite and Sage Intacct are pretty common next steps for digital companies.
This is a big question though. I finish with this hard-earned wisdom: enterprise software is only as good as whoever is implementing and administrating it. A good VAR makes all the difference in the world.
Not the original poster, but I noticed two things about MuleSoft. First and foremost the developer tooling - especially MuleSoft's Anypoint Studio IDE - is horrible. It could compete for the worst IDE experience out there and easily win by a landslide. It requires obscene amounts of resources (CPU/memory) and barely works with Hello World type applications. For anything midly complex than 1+1 = 2, be prepared to face constant freezes, frequent restarts and loss of developer productivity. Mule code is an XML based proprietary DSL, learning which isn't useful outside MuleSoft ecosystem. And this lock in / proprietary way is everywhere. e.g. Mule uses RAML (instead of Open API for API design) or Mule uses its own data transformation language called DataWeave which has a steep learning curve. The second and more concerning issue is the cost. MuleSoft is incredibly expensive as a product, and hiring good MuleSoft consultants is hard. It's difficult to justify it's price tag when there are better options available at lower prices (e.g. apigee/kong for api management). Other reasons include bad documentation, history of miserable support for backward compatibility (aka Mule 3 and Mule 4), treating other programming languages (python, java etc) as second class citizens in Mule etc
Everything the other poster said is correct. Every thing about Mulesoft seems designed to be endless tedium and a time suck. For example, instead of a centralized keystore, this is how developers are supposed to manage secrets. https://developer.mulesoft.com/tutorials-and-howtos/getting-...
