A Framework for Investing in Enterprise Software

A lot of people have been asking me for some more context to the slides I recently posted. It's part of a much broader talk on enterprise software and is easier to converse about rather than put in writing. That being said, this post will serve to aggregate my slide deck, blog post and thoughts on the subject along with a tweetstorm I recently did on it in order to provide something easier to digest outside of a formal talk.

Tweetstorm Below (or click on link embedded above)

So let's unpack some of these thoughts: 1) Business model guides strategy - this basically is the premise that bottoms up, product led models tend to move horizontally quicker than top down, direct sales models; examples include Atlassian, Elastic, Datadog, etc

This is not to say that Salesforce or Workday or Anaplan with traditional top down models don't move horizontally, far from the case (see CRM vacuuming up the SaaS industry), but they go vertical for longer because of high CAC strong retention needed to make the model work

2) Neither top down nor bottoms up is a better business model, what matters is alignment of cost structure. The concept here is making sure the company's CAC and sales velocity matches the steady state retention and average ASPs that they're pursuing

The examples I used, ServiceNow and Twilio both have aligned cost structures but you can see NOW spends much more on upfront CAC because they have lower steady-state churn and higher ASPs vs Twilio which employs a more rapid sales velocity to close more deals w/ less CAC each

3) Switching costs take multiple forms - classic switching costs are things that are strategic top-level company decisions (i.e. adopting a CRM), they house a ton of data, typically have specific admins or users trained on them and take lots of time to implement, hard to rip out

New switching costs have to do with customer love, ease of use, and integrations - they enable the customer to do something much faster and easier and ingrain themselves in the customer workflow which produces a high cognitive load to then switch; these are harder to evaluate

An example would be productivity software, if you use Asana, a product must either be cheaper (which is hard if you're on free version) or must be 10x better to cause you to switch to say Notion or Trello, because you already understand the UI and everything just works

4) Distribution in enterprise software is one of the most powerful things - establishing a good distribution engine cements a product lead as that engine lowers CAC, brings in more customers, and spins the flywheel allowing for more R&D and S&M to further cement a lead

An example is Salesforce, their product has a release cadence of once per quarter while Dynamics365 is agile and daily, by many accounts the core CRM product of Dynamics is better; however CRM has an army of consultants & SIs that get professional services work and rev share

They are literally selling Salesforce's CRM for them because it benefits the SI as well; Salesforce then built out integrations & an app store to broaden the product horizontally to attack multiple use cases further augmenting its product in customer workflows, lowering churn

Once this distribution engine is built, Salesforce buys or builds new "Clouds" like Mulesoft and injects them into the distribution engine immediately increasing revenue with a decrease in CAC if Mulesoft was to pursue alone, this model is replicable for many other verticals

Atlassian does something similar except built the distribution engine up through primarily website traffic, user communities, and evangelism and then used that attention to buy Hipchat, Bitbucket, OpsGenie, Trello to throw into that distribution engine and grow

Effectively both bottoms up and top down models are utilizing distribution to expand the surface area with customers in a FCF efficient way which then spurs a flywheel effect going forward as well

End/ All this is why I think enterprise software companies have some of the best business models in the world but understanding these concepts mentioned above is necessary to tease out the quality from the pack

Blog Post

This is what kicked off the idea for the presentation where I dug a little deeper into these areas to flesh out my thinking more.

Presentation

Finally, here's a link to the slides.

Hopefully this is helpful to those who read it. It's the best way I have of unpacking my thinking on these subjects without talking through it live in a Q&A format.

As always, please feel free to reach out regarding any questions and please comment if you have any thoughts on the points I discuss!