Trends I'm currently observing in B2B SaaS: 1. Everyone and their mother is rushing to get their product AI-powered. Or is it AI-solution? Or AI-platform? Pick your poison. 2. AI features often require changes in customer behavior to be fully effective. Self-serve experiences struggle to accommodate these shifts, making human involvement frequently necessary. Sales and Support teams - it is your time to shine! Is PLG dead then (again)? Nah. But it's not always the preferred path. 3. There’s little understanding of how much AI actually deteriorates margins because... AI is so expensive and SaaS businesses are not used to deal with such costly 'things'. Move over astronomical AWS bills; there's a new, costly AI kid in town. 4. Freemium and free trials are hard to justify with AI costs. Paid and credit card–required trials are making a big comeback. 5. As AI costs change, companies will need to adjust their pricing... a lot. Having a platform that allows pricing experimentation will be key to success. 6. Product teams will need to get hands-on in owning pricing models for their features. SaaS is no longer 80%-margin candyland. Product teams will need to control costs and play a much more active role in pricing and packaging. 7. The rush to AI means many products will offer overlapping features, making it difficult to stand out. PMM-I feel your pain already... 8. Too many products are rushing to ride the AI wave by acting as simple ChatGPT wrappers with no proprietary functionality. This positions OpenAI (and others) to use these products as proof of concept, allowing them to observe user behavior (while getting paid!). Eventually, they can build competing functionality themselves and shut off API access whenever they choose. OpenAI giveth, and it can taketh away. 9. AI definitely feels reminiscent of the dot-com boom: exciting, but inevitably likely heading for a crash. #b2b #ai
Pooja Kashyap
8mo
AI has to align with business goals and most importantly, it should enhance customer experiences. Yes, pricing models are going to be critical, and it’s going to require a level of agility that many companies aren’t prepared for. Strategic, not just trendy, if we want to capitalize on AI's true potential - 100% spot on!
While there is great unknown looming over general market at the core a genuine application of AI can do wonders for your business, not some wrapper on top a foundational model. Few weeks ago I spent a good amount time explaining a client who demanded AI implementation for a problem which was no more than a statistical issue. Knowledge is key, so you need to find a specialist on AI before jumping head on any implementation. Just because it can be automated it necessarily does not need AI.
Charlotte Morgan
8mo
Absolutely agree. And beyond the aforementioned complexities, legal ownership in AI-generated content—is a minefield! Businesses must thoroughly understand the specifics of the platforms they’re using. Each platform has its own unique terms and conditions, making it anything but straightforward. Moreover, if an AI engine generates content based on existing material, businesses can still be held liable for using it under their name. It’s critical to navigate this space with care and clarity.
Dan Apps
8mo
Prices will come down with innovation in new models, as will the cost of compute. The question will be, how easy will it be to port to a cheaper model. Putting up prices for customers only makes sense if it's viable for the customer, otherwise they'll find a cheaper alternative, of which there will be one somewhere.
Srinivas Mahesh
8mo
Great insights! Completely agree that AI is shaking up the traditional SaaS playbook, especially around costs and margins. The shift in pricing experimentation and the growing role of product teams will be critical for staying competitive. It’s definitely a challenging, yet exciting, time for the industry! Elena Verna
Julia Ankudinova
8mo
Elena Verna from what I could observe many Saas for Sales added AI. And this happened very organically. Conversations transcription and analysis, for example. I wouldn’t say it changed behaviour significantly, but it added so much value… from what I can also observe AI is included into price but it is usually limited. And you need to pay more if you go beyond limits. What I wanted to say: SaaS and AI is logical and natural combination if it AI helps to achieve the goal faster and in a more efficient way. I don’t have SaaS product yet) but I’m trying AI and experimenting a lot. I have built very simple and very primitive ai coach for my personal soft skills growth. It took me no more than one hour)) And it helped me so much ( I paid only 40 dollars instead of 1000)). So my feelings are more positive )
Matthew Seeger
8mo
On 3, this thought is a little stale already, no? Prices are coming down fast and when you hold these new costs up against the costs of the status quo it seems that a lot of this is starting to make sense commercially. On 7, for sure! I think we're a bit past the "AI for Everything" mindset, some problems are well solved just the way they are. I think we'll see more of that and less overlap On 8, I see it differently. I see what openAI (and other LLM makers) doing what products like Oracle DB and MySQL did way back when. They provided something very core and fundamental and while they could hypothetically do it all themselves, they realistically and pragmatically didn't and many other products were conceived and grew using this technology (Salesforce, LinkedIn, Amazon). Salesforce was once "Just a Wrapper to a database" if you distill it enough. On 9.... hope not 🤣
Fantastic points! I think we need to wait for the end of hype to see the true value because it’s extremely hard to cut through all the noise. Everyone is rushing to get AI in the house because of FOMO without realising that it’s a ticking bomb on itself
Nick Malekos
8mo
I did not know about #4 its pretty interesting. #8 that's what I have been expecting since the first wrappers, and already the first wave of everbody offering a chatbot for others is failing, because chatbots are so easy to build. And, #5 is also a big expectation. Like Uber/Lyft and food delivery startups, they are being underpriced by VC funds. When those end and profits need to be made ... it's price apocalypse for smaller companies and users.
Chief Marketing & Product Officer / Founder @ Mopinion - The #1 Feedback Software for Websites, Apps & Email
8moThanks for sharing these insights! This is a great summary, and I agree; these trends are resonating in our space (user feedback/martech) as well. On point #8, one thing we’ve observed is the importance of factoring in privacy and security when using API wrappers for providers like ChatGPT. For example, in our case with user feedback, open comments often contain sensitive information. This is something that may not raise as much concern in the US, but for us here in Europe, data privacy and compliance are critical considerations. To address this, we opted for an open-source LLM, hosting it within our own ecosystem. This approach has two big advantages: we maintain full control over our data and manage costs effectively. As a result, our setup is quite cost-efficient, and we didn’t have to gate our free trials.