Armand - Analyst

Last week, Dorus and I went to SaaStock 2022 in Dublin. It is one of the bigger fall events for software ventures in Europe. The focus lies on knowledge sharing. What follows are some key takeaways from my side after listening to the lectures and talking with founders and learning about their propositions:

1. Go-to-market: PLG PLG PLG. I think I'm not exaggerating when saying that in 50% of all lectures, people discussed PLG. The concept is being pulled very broadly, however. The moment you're driving decisions on metrics derived from product usage, you're a PLG company, according to some of the speeches. Whereas we at Newion still define product-led growth as growth that is initiated by product usage. ie. the product has a natural need or want to expand. Without the capability of sharing music, Spotify would not be a product-led growth proposition. I would therefore want to make a case for using the concept of PLS instead of PLG, Product-Led-Sales, a term that covers much more of the actual content.
2. Proposition: The great suite decomposition in multiple tools is over but also very present. Microservices have allowed products to become feature sets. The API-first ideology has taken this a step further. It enables use cases where tools can interact directly with each other. Customers can compose their own solutions by stitching features together in a new feature composition layer, creating bespoke suites. Companies present that were surfing this trend are for example Open Integration Hub and Captain Data. Zapier automated pipelining for events, these propositions facilitate flexible pipelining for data. SaaS 1.0 was on collecting data, SaaS 2.0 was on activating the data, and SaaS 3.0 is on managing that data.
3. Proposition/Go-to-market: Now we come to the title of my text: SaaS is Dead, Long Live SaaS! SaaS vendors have been played by their own game and without proper action, it will bite them back right in the face. SaaS is successful because it enables vast process improvements without steep Capex expenses. Services crawl up on the P&L of their customers up until the holy grail of all cost obfuscators: cost of goods sold! Once there, your customers will be able to pass these costs on to their customers, and poof, your product all of a sudden becomes “free”.

Usage-based billing: the holy grail for AI startups

This game has also been played by suppliers of SaaSvendors, naming AWS/Azure/… for example. Compute has moved from a CAPEX expense on servers and OPEX expense on electricity towards a monthly bill from a cloud computing provider.

These bills are often structured at a cost-per-fetched instruction level ie. [EUR 0.0000001 per CPU instruction]*[# fetched CPU instructions during that month]. This moves (cloud)compute costs entirely up on the P&L into the beloved CoGS line, and it’s free again! Add a markup on your pricing and your customers will pay for the compute.

Hold your horses, it’s not that easy! SaaS vendors offer services at a tiered pricing scheme with often 2 or 3 tiers. This is okay when your software is not compute-intensive but can become very costly if your product is compute-intensive and you fail to precisely measure the used resources by every individual user. Your markup on computing cost will fail and your gross margin will tumble (auwtch).

The complexity of this problem becomes even greater when thinking of the modern decomposition of product suites where customers can combine their preferred features into a personalized product. How to price every feature and allocate the real costs to the feature usage for every individual user and this at a billion+ requests per month?

a16z going all-in

a16z has apparently signaled and identified this issue and is now investing on this thesis. Early this year, a16z led a $30M round in Metronome, and in September, the fund invested in London-based company Sequence. Sequence wants to become the one-stop shop for every compute-intensive startup with regard to pricing, billing, and usage. Cool! Here is a link to the a16z article on why they invested in Metronome: https://a16z.com/2022/02/01/investing-in-metronome/

M3ter, another UK-based SaaS pricing optimization platform present at SaaStock, has also recently raised a $17M seed round. Their proposition differs from the above-mentioned ones because of their ecosystem thinking, with open access to other shackles of the value chain.

Why are these propositions particularly cool (for VCs)? Because these products are very sticky! Everywhere in the codebase where a compute request is made, another line of code is added with a request toward Metronome or Sequence. Try to demolish that out of your product at a later stage. (Just like a proposition like Twilio for example).

What's next?

I am very excited to further investigate this space and interact with new innovators that are taking this to the next level. SaaS is not dead, but tiered pricing maybe is. As with everything in nature, pricing follows the path of the least friction/unnecessary fat. Usage-based pricing allows for this optimized pricing efficiency, and now because of these innovations, it is finally attainable. Will it achieve the same disruptive effect as what SaaS did 20 years ago? That’s to be seen but it will certainly enable new business models and innovative propositions to be economically viable.

I am looking forward to what the future will bring and how we will be able to surf the usage-based wave!! SaaS is dead, Long Live SaaS

If you want to further discuss this topic, please reach out!