Ask HN: How to Beat SAP?

I would not focus on trying to replace SAP (which, to be honest, is nearly impossible for a startup) but to find ways to improve it or profit from its ecosystem.

For example, Celonis is a German startup with a >1BN€ valuation that started by mining business processes from SAP systems to tell their customers how they could improve their practices and save money. SAP could've done this themselves of course but they ignored this market entirely, giving Celonis a chance to grab it (now they're partnering with them). I'm sure there are many other areas where you can create tremendous business value from the data in SAP, which SAP doesn't pursue due to them being too "niche" for them or due to ignorance. So I'd say find one of these cases and build an MVP around it, which is more realistic than writing a full-blown ERP anyway given your probably constrained resources.

You seem to be working under the assumption that SAP is the only ERP game in town. This is not the case. There are, in fact, a ridiculous number of ERP solutions, most of them operating within a specific niche. Do a search for [industry of interest] ERP and take a look at the results.

Also, as a side note: CTOs buy Oracle and SAP because they are large, established companies with a whole suite of offerings, a service ecosystem, integration partners, an enterprise sales cycle, etc. etc. Not because they are necessarily the best possible fit for a particular use case.

I have worked as SAP consultant for most of my working life. Now lately I have changed carees and doing programming full time.

I think from techology point of view SAP definitely can be challenged and better product could be produced.

But all those business processes, edge cases, compliances with regulations. Thats what makes beating SAP difficult.

If somebody is thinking of creating ERP related products I would be interested in talking about it.

SAP, Oracle, IBM are purchased by CTOs because they have comfort that they are dealing with "big enough companies to understand their enterprise needs" - which translates to "the sales reps have big expense accounts and wine and dine me". I have worked with all of the above vendors and I have seen that so much of the marketing budget goes towards "entertainment", junkets, conferences, etc.

Besides buying from a big vendor makes them look good in the eyes of their credulous board. There is too much risk in being blamed for buying from an "unproven vendor".

This is a highly complex subject that - among other aspects - involves enterprise sales, the peculiarities that come with it and the inherent complexity and variation within larger companies and different industries.

If you really want to do this the key word for a solution here is "industry-specific". A huge part of why SAP ERP is such a complex beast is that it needs to accommodate each industry and each edge case.

An industry-specific ERP system could ignore that and focus on the particular requirements of a specific industry, which if properly executed should result in a much better UX.

Great Question. Would love to know the answer.

A major hurdle is providing 10+ of the typical ERP modules. E.g. production module (PLM) needs access to sales/CRM module (so that it proceed with ordering necessary materials etc. or plan for it in case of CRM); to do so it needs access to the procurement module, and accounting one; HR (so you can schedule workers), and things just keeps adding up. E.g. here are some of the typical ERP modules: Human Resource, Inventory, Sales & Marketing, Purchase, Finance & Accounting, Customer Relationship Management(CRM), Engineering/ Production, Supply Chain Management (SCM), etc. (e.g. https://en.wikipedia.org/wiki/Enterprise_resource_planning#/... )

The tricky part is that there aren't many standard modules used; probably with the exception of the accounting (Quickbooks, Xero); e.g. some companies bootstrap on top of accounting modules: * https://www.mrpeasy.com/xero/ * https://www.mrpeasy.com/quickbooks/

I'd love to hear any ideas of how the hurdle of needing to provide large number of integrated modules could be overcome.