Multi-Tenant SAAS Is The Future
I have mentioned multi-tenant as a core feature of the best of breed SAAS applications in previous blog posts, and likewise the stocks we need to focus on. However, I feel the need to write a separate blog post just to ram home the point: Multi-tenant SAAS will dominate the future of software applications and any other architecture will be left behind. Artificial Intelligence will bring in the next wave of software and multi-tenancy is required for good AI.
Single-tenant is when you install the software on a single server, and your company is the only company who uses that installation/server. This is how software applications were designed in the 1980s. Oracle, SAP, Sage, Microsoft, Fiserv, FIS, Epicor, Syspro, Intuit, Infor…all of these were designed before the internet and use single tenant as it’s core architecture. 1 server, 1 application, 1 customer.
Multi-tenant is a server, or an entire data center of servers, running 1 application, but serving hundreds, or thousands, or millions of customers all at once. Salesforce, Slack, Amazon AWS, Microsoft Azure, GMail, Facebook, Twitter, TikTok, Shopify, DocuSign, Zoom, HubSpot, Mailchimp and RingCentral are just a few who use multi-tenant architecture.
Not only does multi-tenancy provide for a better overall customer experience, multi-tenancy is the foundation for the next generation of software.
Artificial intelligence, in a very simple way, can be thought of as not much more than statistics or odds. When you flip a coin, it might come up heads 5 times in a row. It’s rare, but it can happen. You might assume that a coin always lands on heads. But we all know that’s not right. In order for you to figure out that a coin has a 50/50 chance of landing on heads or tails, you need to do a lot of coin flips. After a few thousand coin flips, you’ll see the 50/50 chance show up.
In other words, you need a lot of data to see what the chances are of one thing happening over another.
When a software company uses single-tenancy for it’s architecture, the data is separated from each customer.
You can’t combine the data, or at least it’s not easy. When the software company uses multi-tenancy, all the data the customers are using is in one database. It’s already combined, and the software company can use all that data for statistical analysis, ultimately AI. And AI is what will make the software 100x more useful. Imagine:
You live in San Francisco where your company is based, and you are speaking to investors in Japan who don’t speak English. But that’s fine, because your Zoom call is being translated in real-time on your screen so you can see what they are saying in English, and they can see what you are saying in Japanese.
Next, your AI assistant pops up on your Zoom call to let you know you are talking too much and you’re losing the interest of your investors on the call. It gives you some questions to ask to get them talking and participating again.
The AI assistant wins the day and your investors decide they want to do another funding round with you. You use DocuSign to draw up the terms of the agreement, and the AI in DocuSign looks over the agreement to let you know there’s a section that’s not legal, and proposes some terminology for you to use in order to stay compliant and consistent with your other agreements.
Seamless.ai gives you the best companies to reach out to that would both be a good fit for your product - and based on their recent online history - are most likely going to make a buying decision soon. Now your list of prospects is 10 extremely high quality ones, instead of thousands of companies that have no interest in you.
Crowdstrike can use the data from a cyber attack on one customers’ environment, to automatically update the software in all customers’ environments to protect them in practically real time.
Bill.com can use the invoice layout from your vendor - Ace Plumbing Supply - and automatically be able to read the invoice and input the data for any other bill.com customer who uses Ace Plumbing Supply as a vendor. From now on, nobody else will have to input data when they receive an invoice from Ace, the AI bot will do it for you.
Peloton can see the time, frequency and intensity of your workouts and match you up with other people around the world to work out with.
Upstart Holdings can use data from millions of customers to give more accurate credit ratings to smaller banks, credit unions and financial companies, allowing them to offer more loans - as well as more accurate loans - resulting in increased revenue for lenders.
All of this is possible because of AI, which uses tons of data. And the way you get that data? Having a multi-tenant architecture in your software. There are more than a few software companies out there still selling a crap-ton of single tenant stuff, and in the not-too-distant future they will be obsolete. They will not be able to build AI on top of their single tenant system, and will be left behind.