The Bloomberg Terminal of Startup Ideas

Friday, November 21, 2025 | Daily Edition

The only newsletter that shows you what 250,000 founders are building before those companies exist. We crunch internal data to spot early trends as we explore opportunities in the startup world together.

📊 IDEA FUTURES INDEX

24HR MOVERS:

manufacturing ████████░░░░░+129%🔥

e-commerce - online retail ███████░░░░░░NEW🔥

brick and mortar ██████░░░░░░░+92%🔥

consulting █████░░░░░░░░+75%🔥

services █████░░░░░░░░+75%🔥

VOLUME: 416 new ideas submitted (↑ 129% vs last period)

• SaaS is still the volume king (84 ideas) but just posted a brutal -69% slide in submissions, even as Google search interest for “SaaS” is spiking (+5.9%), founders are clearly doomscrolling SaaS content but actually submitting offline-heavy plays instead.

• Brick-and-mortar (+92%), consulting and generic “services” (both +75%), and especially manufacturing (+129%) are the fastest risers, signaling a sharp rotation into real-world, cashflow-first businesses.

• AI/“tech” ideas cratered (-75% each), and healthtech is retrenching hard (-40%), suggesting fatigue with long-regulation and me-too AI tools.

• The surprising move: manufacturing and small physical franchises are suddenly hot niches inside the “unsexy” economy. If you can wrap software or AI around real-world ops, the timing window just opened while everyone else is still optimizing their landing pages.

Let’s dive into the data…

🔬 VALIDATION REALITY CHECK

This week's hard truth:

Of the 1,184 ideas submitted in the last 7 days:

• 88.6% still have no MVP

• 11.4% have progressed to MVP stage

• 46.5% still need a landing page (550 founders flagged a website need)

Here’s what founders are actually doing vs what they say they want to do:

Idea volume is up but MVP momentum is basically flat, and the gap between “I’m building” and “I shipped something that works” is the real graveyard. The encouraging signal: the tiny group that does reach MVP is heavily clustered in focused, narrow problems (automation tools and AI workflows), not broad “platform” plays. 

Translation: generalist, buzzword-heavy ideas are all talk; the people quietly moving to MVP are those who picked a sharp niche and ignored everything else.

🏆 TOP IDEA SCORES OF THE WEEK

(These are anonymized and summarized to protect the founders’ ideas)

1. Renal Safe Pantry Brand · E-commerce · Score: 78

2. Online Courses with Progressive Levels · EdTech · Score: 75

3. Commerce Electronic Electrical Products · E-commerce · Score: 75

4. All-in-One Education SaaS · SaaS (Software as a Service) · Score: 75

5. Lean Flat Priced Booking · SaaS (Software as a Service) · Score: 75

6. Digital Boundaries Business · Consulting/Services · Score: 75

7. Storytelling Horror Business · SaaS (Software as a Service) · Score: 75

⭐ CLUSTER OF THE WEEK

brick and mortar

n=23 submissions · +92% WoW

This week’s dominant cluster is “offline-first founders”: brick-and-mortar (+91% WoW), consulting (+75%), services (+75%), and manufacturing (+129%) ideas all spiked together while AI/SaaS submissions fell sharply.

What unifies this cluster is operators trying to modernize very real-world workflows, often layering lightweight software or AI on top of physical or service-heavy businesses. It’s emerging now because AI infra and SaaS are saturated and noisy (funding still flows there, but founder attention is shifting), and we found Reddit and TechCrunch are full of stories about regulation, trust, and “agentic” systems breaking things, pushing new founders toward tangible, defensible, cash-flowing niches. 

The opportunity: build “AI-in-the-loop” tools, playbooks, and financing products specifically for these non-digital-native sectors (micro-manufacturers, local services, informal retailers) where willingness to pay is high, competition is low, and speed + execution matter more than model benchmarks.

👤 FOUNDER OF THE WEEK

Founder of the Week: Don Lo – RenalLife

Don is building RenalLife. He stands out because he’s already done the hard validation work most consumer founders avoid: testing 500+ formulations, locking in an advisory panel of 3 doctors and 5 dietitians, and collecting 100+ reviews and 85% positive feedback before scaling. He’s also solving his own problem as a CKD patient, which shows in the precision of the positioning and the willingness-to-pay signals.

If you want to meet Don: Email us if you want to help his journey.

💰 DEAL RADAR: WHAT MONEY IS CHASING

Money is chasing ultra-specific AI infrastructure and compliance, not another generic copilot:

• Kaaj’s $3.8M seed for credit risk automation

• Sphere’s $21M for AI-native tax compliance

• Coverbase’s $16M for AI-powered procurement

all map cleanly onto ValidatorAI’s fast-growing “vertical AI back-office” cluster (finance ops, tax, vendor management). These rounds validate that founders obsessing over narrow, regulation-heavy workflows with clear ROI and existing budgets are directionally right, while the flood of “horizontal AI assistant” ideas in our pipeline is getting no corresponding capital signal.

Investors are rewarding products that plug directly into legacy systems (ERP, banking cores, procurement suites) and remove headcount or audit risk on day one. The niche inside the niche: founders who pick a single painful spreadsheet workflow in a regulated function and automate it end-to-end are getting funded earlier and at higher velocity than broader “AI for SMB back office” plays.

A few interesting funding rounds:

Founded By Ex-Nvidia Researchers, Flexion Lands $50M To Build The ‘Brain’ for Humanoid Robots
Source: crunchbase-news

With Customers Like Okta And Coinbase, Coverbase Raises $16M To Grow AI-Powered Procurement Platform
Source: crunchbase-news

Credit risk automation platform Kaaj raises $3.8M seed from Kindred Ventures
Source: techcrunch-venture

🎯 HIGH-CONFIDENCE OPPORTUNITIES

1. AI-native CRM and sales systems that actually replace legacy CRMs, not bolt onto them.

Lightfield’s pivot from a 20M‑user presentation app into an AI‑first CRM is working: 100+ customers, >1 hour/day usage, and YC founders explicitly skipping Salesforce/HubSpot for AI‑native tools.

Internally, we’re seeing a surge in “AI agency,” “agentic automation,” and “DFY sales funnels” ideas, plus Product Hunt is showing traction for GTM data/agent tools (Floqer, Guideflow, Oskar, TrustMRR). This is the same pattern as the early HubSpot era, but with LLMs doing the grunt work instead of SDRs and RevOps.

If you’re building here, niche down hard (e.g. “AI CRM for 5–50 seat B2B agencies” or “agentic revops for PLG SaaS”) and ship an opinionated workflow that auto‑captures, summarizes, and acts on customer interactions—not another generic “AI notes” layer.

2. Operational AI for expensive infrastructure: GPU, data centers, and manufacturing cost control.

External capital is pouring into “AI about AI infra”:

Lambda’s $1.5B, ScaleOps claiming 50–70% GPU cost cuts,

Cavela’s $6.6M to slash pre‑tariff manufacturing costs by ~35%.

Internally, manufacturing ideas are up 128% WoW and consulting/services around “AI audits” and “agentic automation” are spiking, while AI category idea volume is down 75%—classic sign that the gold rush has shifted from model-building to picks‑and‑shovels optimization. 

The niche inside the niche is vertical, high‑ticket infra: self‑hosted LLM clusters, GPU‑heavy analytics, robotics, and multi‑factory brands where a 10–30% efficiency gain is millions. If you can productize what ScaleOps is doing (auto‑rightsizing, workload‑aware scaling, procurement/tariff arbitrage) for a specific segment, start with a “done‑with‑you” consulting + internal tool, then crystallize the patterns into SaaS.

3. AI-powered customer intelligence layers for under‑served, high‑complexity segments (HNW, regulated, informal markets).

VCs and media are obsessing over “agentic era” orchestration and “customer intelligence” (Lightfield, Kaaj, Sphere), while B2B sales trust is eroding under AI‑sprayed outreach.

The pattern: segments with fragmented data, high stakes, and zero modern tooling (HNW insurance, cross‑border informal sellers, regulated services) are ready for a vertical “brain” that consolidates policies/contracts/transactions, flags risk, and suggests next actions.

Founders should pick one messy segment, embed with 5–10 real users, and build a narrow “single pane of glass” that starts as a high‑touch service plus simple UI, then layers in agents for reconciliation, alerts, and workflows once you’re wired into their data.

Top Category by Score: SaaS (Software as a Service) (Total Score: 5598, Count: 84)

🏁 THE WEEKEND CHALLENGE

Weekend challenge: Turn one vague idea into 5 real buying signals.

Pick a single idea you’re sitting on (or one from this email), and by Sunday night: (1) write a 1-sentence problem statement, (2) message 10 target customers (LinkedIn, WhatsApp, Reddit, cold email) with that sentence plus one question:

“How are you handling this today?” and (3) log how many reply and what tools/brands they mention. This matters because right now ~46% of founders stop at a landing page and <4% see revenue… 10 real conversations will tell you faster than any MVP whether you’re building another “nice SaaS” or something people will actually switch to and pay for.

Ship a landing page in the next 48 hours for brick and mortar. There are 23 fresh submissions this week and momentum is +92% WoW… reach out to 5 people in this niche before Monday.

🔮 PREVIEW OF MONDAY

On Monday, we’re unpacking a weird split-screen: AI startup ideas are collapsing (-74% WoW) just as AI funding and infra (Lambda, ScaleOps, Grok, Nano Banana) go parabolic and founders quietly pivot from “AI product” to “AI-powered consulting and services.”

We’ll map where the heat actually moved this week and show the niche-inside-the-niche plays that are getting built while everyone else argues about models.

.

That's it for Startup Trends Friday.

Take care!

See you Monday for Market Trends Monday!

Aron Meystedt


Chief data nerd at ValidatorAI.com

Your ideas matter! Check out ValidatorAI.com and validate your idea today!

Follow on social media below. Cool things are coming.