4 ms·
This is about selling a less than $500 per month software to the enterprise. I have experience selling $100k per year software to the enterprise and it barely l
by polote 5y ago
This is about selling a less than $500 per month software to the enterprise. I have experience selling $100k per year software to the enterprise and it barely look likes what is described in the article.
- rmccue 5y agoThis is effectively How To Sell to Customers at Large Businesses, not How To Sell Enterprise Software. Useful if you want to sell your consumery software in, but agreed, there’s significant differences when selling bigger budget software.
- Aeolun 5y agoWhat does it look like? I think our software procurement process looks pretty much like this.
- micromacrofoot 5y agotwo lawyers arguing about indemnity for a year
- treis 5y agoIt's not a nurse or even lead doctor making a decision. It's the CXO or VP or Director depending on how big the company is. They're also price conscious. Everyone has a budget and the more of that budget you take up the worse it is for them. Enterprises are good because they have big problems that cost lots and lots of money. They're willing to give you a significant fraction of that money if you solve the problem.
- treis 5y agoForgot to add, but the politics. Especially for software where there's a perpetual struggle between IT and the business side for power. Have had one client where IT left a call when they released people from the business side were there.
- tptacek 5y ago(I was PM for a product that piloted just around $100k and that had a median full phase 1 deployment price in the 7 figures, and this was back in the early 2000s). Real enterprise sales are much more relationshippy and less transactional --- just in my experience --- than this post talks about. Making these sales is a lot about taking and managing meetings, understanding RFP processes, structuring bakeoffs and trial deployments, and, probably most importantly, staffing sales prospects with SEs to handhold the deployment and keep the relationship with the technical decision-markers warm. Knowing who the economic buyer is, and who the real decision-maker is, is another important skill not mentioned here. Patrick isn't really talking about enterprise sales; he's talking about how to hack the reflexes of enterprise purchasers to bootstrap a self-service product. Serious enterprise deals often take multiple quarters to close, and they can be staffed almost the whole time; it's a whole big thing. None of this should be construed as me knocking the post. The post is fantastic for what it's actually about.