Interesting question. If you mean enterprise software/saas offerings with a lot of stickiness, there are clearly a lot of them in the data space. Palantir, Snowflake, maybe Splunk etc are the sorts of fundamental pieces of architecture which people build an “enterprise data strategy” around, and those sorts of projects take a long time to roll out and are therefore intrinsically hard to roll back. There’s a long way to go to become as big as Salesforce though clearly so we’ll have to see how the next few years play out for them. It looks likely to me that some of the “enterprise AI” players (openAI, scale.ai etc) may well end up in the same position also as large companies build LLMs and other models into their workflows.

One thing about Salesforce (anecdotally) is that CRM systems are painful to put in place because they require people to change. They promise that once the processes/workflows are redesigned this “digital transformation” of the business will result in various benefits. This gives Salesforce an army of free sales people at the big consulting firms who essentially sell this transformation project and salesforce as an side-effect.

However people don’t want to change. For this reason at every org where I’ve worked where there has been a big project to put in a CRM system (not just Salesforce - I’ve seen it with Siebel as well) there is a huge painful project to get the CRM system in and afterwards there is massive management pressure to try to force people to use the CRM system in place of existing processes (typically using email to pass around spreadsheets). This pressure is because the benefits of the CRM don’t typically accrue to the people who have to put data into the system. Typically they accrue to the managers of those people. So CRMs often end up out of date, full of duplicate records, bad data etc rather than the perfect repository of institutional knowledge they are billed as.

The promise of the new gen I listed above is because they operate on the data sources (rather than requiring the people to change their existing flows) the analytical insights can come from existing systems and processes without as much organisational pain. It remains to be seen whether they end up as sticky in the long term.