Companies like Anthropic have one advantage in the world of SaaS. They’re developing a product that serves their own Industry sector. This is a very rare situation and it’s why most SaaS products suck. Things didn’t work out too well for Dr Jekyll when he tried to double up as Mr Hyde, and the same is generally true for SaaS providers.
I’ve posted before about the business evolution map I developed with David Cox. This map allows us to predict which attempts at diversification will succeed in the short term (< 3 years) and which will need a lot longer (> 10 years) and therefore have a high chance of failure. What the map shows is that a move into a non-neighbouring sector is a bad idea. It works because the map positions sectors based on the activities and knowledge needed to succeed. Sectors with a high degree of overlap are placed together, and as the differences grow, so does the separation. The map can be used like a physical map, and diversification like mapping a journey.
So what does this mean for SaaS companies?
Any company that builds and runs a SaaS product, is by nature a software development company. That is its primary business sector. The product it creates, however, has to sit in the business sector it serves. This means the SaaS provider has to also operate in that sector. If the sectors are separated by any distance on the map, the SaaS company needs two different mindsets, two different sets of skills, and ultimately two different management styles.
Amazon managed this issue with retail, logistics and technology by separating the companies and running each one differently, with different management control. AWS is run very differently to Amazon Retail, and Amazon Logistics. If you look at the time taken for AWS to go from initial formation to initial presence in the market, it fits nicely with the +10 year timescale dictated by the map. All that money didn’t make it go any faster. People and skills take time to acquire.
This is why most SaaS offerings are bad.
If the SaaS company is primarily a software house it will have a poor understanding of the market it serves. Its products will be created with a strong bias towards what works for the software development minded person. The people trying to use them will complain of unintuitive interfaces and confusing user experiences. Even if they work with customers to develop the product, either the software mindset will dominate in an effort to standardise the offering, or the product will be a duck-billed platypus of different customer’s needs, squashed together.
If the SaaS company is primarily native to the industry it serves, then they’ll struggle to be good at creating software products. Their management styles and mindsets will fight against the approaches needed to build great software. Consequently, the product may well be a good fit for it’s market, but it will be an unreliable, unmaintainable mess that causes as many problems as it solves. That’s why your HR systems, your finance systems and your warehousing systems are so awful.
And that’s the way they’ll probably stay.