The place where Value Propositions go to die
“Not another solution looking for a problem”
Welcome to Autumn and approaching Q3 for most. Somewhere in the midst of an explosion in data centres, AI adoption is proving patchy and SaaS is feeling worried about its future in an Agentic world… The wheels of commerce keep turning and we’ve all got a number to hit.
You’ll have seen by now we’ve added a podcast to the #norocketships offering, and we reckon episode 2 is a doozy. The slightly reluctant interviewee Ditlev Bredahl (founder of Hosted AI, his 7th startup) talks about why he can’t stop doing startups, and what true entrepreneurship takes out of you. He’s also a man with high EQ, unusual for a high achiever..
Let’s go shall we?
Vile Rumour & Gossip
The race for an AEO feed is on… and it’s not fair. OpenAI has published the form for brands (or anyone) to complete so that GPT can find your products, but they’ve not opened it up yet. In the background ALL of the eCom platforms have done deals with the LLM’s to feed them - without clients being in the know.
Beware the bill for your AI-created content - We have a few brands in our community who have adopted a pretty advanced landing pages strategy - creating 1000’s of landing pages using OpenAI. They are all being slightly screwed by the Search bill that’s coming their way - API calls maxed out and an escalated invoice. Much to unpick in the new world.
Code quality of LLM’s (discussion points) - For all the SI’s who read this email, a very interesting trend is appearing in LLM’s writing code. Front end? Perfect - fine - excellent. But - the further you go down the stack the LLM’s are less effective, bloated even. Code isn’t fixed… it’s rewritten in order to bug fix. Customers are not interested in this.
Is Demand gen finally growing up?
To BDR or NOT to BDR, that’s the question. In the end everyone does, it’s too tempting and too baked in to funding models, but we’re seeing an explosion of newsletters, podcasts, owned conference series and EGC (employee generated content). The quality threshold is currently quite low, but like all things the hard yards will reap rewards… By issue 100 they’ll have some traction (they’ll all be binned by then right?)
Weirdly, at the same time as this is happening the quality of contact data in CRM’s across the SaaS market is declining as folks become reliant on the ABM machines like 6Sense. Highly dubious for GDPR btw but that ship has sailed now.
The last mile is the challenge still, that first meaningful conversation with a target account. We’re currently running a super-Bougie dinner series for several clients, with liberal amounts of restaurant temptation involved. We’re going back to the 80’s everyone!
The AI Solutions hunting a Problem
Of course it started with IBM Watson, great tech but no market fit. In the end clients couldn’t find a problem that it solved. Now we have 000’s of AI services launching into a market in which the customer is totally conflicted - let me explain.
The board sees “more with less”, cost savings and efficiency, but they don’t understand ANY of the tech. General staff are terrified of losing their role and so resist at every turn if possible. Engineering? Well they’re being told to automate, and internally are mostly a department under pressure for failing to deliver what the business wants 🙁 - so they’re under pressure to lose staff.
We are bound to see a whole ton of examples like Artifact, Ghost Autonomy and more who never get traction, but to the sales leaders out there tempted by the big VC cheque to go AI side… we’d be careful about anything before scaleup stage.
For clients - we’re hearing much more “wait and see”... unless the engineers are doing skunkworks to validate the tools and build trust internally.
eCom is fading
Is it just me, or is eCom on a sort of life support currently. While we wait for the LLM’s and google to decide how the transaction will change, physical encounters are increasing and stores in Europe are taking the investment once destined for digital.
Why? Well - when I speak to those operating retail brands, the revenue in store is currently more certain (and predicted to rise as people return to full days in an office), and tech projects typically do not fail, neither do they change mid-flight.
But the biggest factor right now is…. Tumbleweed : There’s real little advancement in eCom; attribution and tracking is all over the shop; the platforms aren’t innovating quickly in real terms, and so digital is being called out for being an unpredictable luxury channel in some brands.
Then of course there’s the specialists… BÆrskin, Swissklip, Snag, Pott’d and others are killing it because their focus is a super-tight coupling of product, creative and media choices… true test and learn with AI making everything faster and cheaper. But there are very very few of these.
Go where there is appetite… it’s tough to see much of it here.
