June 2nd #norocketships 2026
Want a Data Lakehouse? one careful owner?
Hey people - Another issue from the hive mind here - written by me (Jamie) but contributed to by all those of you i’ve been speaking with these last few weeks.
#Gossip innit
Vile Rumour & Gossip
“Shrugs Shoulders” - As we enter the re-ordering of SaaS, some vendors are simply shrugging their shoulders as clients say “we’re only using 20%, we’re gonna build it ourselves”. A case in point are Splunk (let’s face it nobody ever knew what they did anyway), who are simply standing up, shaking hands and saying “if it doesn’t work let us know”.
Contentful have stepped up (before stepping down) - Not what we’d have predicted, but as the game starts to play out these guys have been moving beyond the standard CSM playbook. Clients are applauding quietly from the sidelines… so they sold the business to Salesforce, which will kill it. Nice work People!
SaaS billing will change - Duh! We know it will don’t we. The questions customers are asking are: Now v renewal? Consumption v % of features used? The first offers are coming out from deflated Search & Merch platforms looking to save clients…
Projects are out there we promise - Despite the doom in SaaS, clients are invigorated. Yes - online is proving tricky to profit from as the socials “gouge” again, but we speak to merchants and vendors all day long who have plans.
When you least expect it - Optimizely seem to have a doozy on their hands with Opal. Very good feedback from customers who have done the POC’s, and seemingly product / market fit - large orgs who had the crappy old stuff from Optimizely are delirious. And - it’s good enough for those with a much more modern stack…
Build v Buy (all over again)
So much to say here, but check out the substack article tomorrow on AI careering (sign up and you’ll get it inbox-stylee). Build v Buy is now about the ambition of the stakeholder inside the business.
This isn’t about “Vibe coding” or some term that vendors find convenient to use derisively about home-grown code. Instead it’s about gradually re-thinking the stack and removing half of the tools. In most cases that’s about 15 SaaS products.
Which ones are for the chop the soonest? Here’s our countdown
1 - All things BI (Business intelligence). Very pricey and “Claudable”.
2 - Search & Merchandising. Not that hard and if you’re a forward thinking company with a feed strategy this is a no-brainer
3 - CRM. Sound odd? A stripped-out system with no “send” capability is on the cards for many.
There’s more coming…
SaaS or Infrastructure?
In amongst the chatter about SaaS is a continuing conversation about valuation, funding and (frankly) finding an exit.
The irony I’m finding is that if you’re a SaaS vendor the multiple might be 3x revenue or worse, whereas if you’re classed as Infrastructure the world is your lobster right now. AI infrastructure? 30x revenue or more.
Vendors such as Hightouch are therefore in an interesting position. Massive growth (almost a rocket if we allowed that term), $100m ARR in just 6 years. Most importantly, the platform stores no data, and is growing because everyone wins… data warehouses see more incoming and outgoing data, so they love it. Customers aren’t paying for yet another storage platform which they manage poorly anyway, and they do not appear to be a SaaS vendor to PE or VC companies.
Fascinating times - if you’re a data centre owner you’re suddenly worth 30x more than you were 2yrs ago when you couldn’t give the business away 🙂
Data Warehouse and its time to shine
Turns out you can’t really “do” AI unless you’ve cleaned house. If you do start AI’ing without good housekeeping, your outcomes will be non-deterministic and you’ll be fired. In fact, they’ll be unpredictable anyway :)
Whether you have a Data Lake, Warehouse or Lakehouse - that’s the thing that’s gotta be clean. As the Commerce Futures group joked last week, Lakehouses come in all shapes and sizes, but many are located in swamps, covered in flies and are prone to flooding with poor foundations…
Seriously though - get it right and the Data Warehouse is where AI comes alive for customers.
And finally - our insult of the week. Please feel free to use freely. Don’t forget to find us on substack... we’ll carry on emailing every three weeks or so, but there’s a richer conversation to be had over there we hope.
Finally finally - Our mates at UNRVLD are doing a thing with the Opal people - all about Semantic layers and stuff... wanna go?
Jamie