Wunder Radar
May 12, 2026

AI Prospecting for Wunder Mobility

A quick overview of how the system works, what it's catching this week, and where v2 takes it.

Built for the AI-Native GTM Associate role by Can Avci

In 8 minutes
01 The thinking Architecture and agents.
02 The output Top 10, drill-in on the #1, v2 Pain Stack with a live Nextbike scan, and the stack underneath.
03 Why Wunder Three honest reasons.

The thesis

There are 150–200 shared mobility operators worldwide. Finding them is not hard. The hard part is timing — catching the moment they're ready to switch platforms. Wunder Radar watches for that moment.

53
active operators tracked
12
with switching signals this week
€3.7M–€8M
weighted pipeline (top 10)

How it works

Five Claude subagents · MCP for data sources · scoring + PVP run as separate orchestration steps

News & M&A
Job postings
App store reviews
LinkedIn signals
→
Score
Pain × Timing × Fit
→
PVP generator
3 WHYs audit
→
Weekly brief

This week's top 10

Ranked by Pain × Timing × Wunder Fit · top three highlighted

# Operator Score ACV est. Why now
1 Europcar On Demand 80 €400K–€700K App store 2.1/5 and not recovering since the 2024 rebrand. Platform pain disguised as a brand problem.
2 Cooltra Group 75 €800K–€1.5M Three-way merger (+Cityscoot, +felyx) 18 months in. Three legacy stacks. You already power Spain — this is an expansion play.
3 Nextbike 50 €1M–€3M 115K bikes. Star Capital PE acquired May 2024, "comprehensive rebrand" announced. PE platform reviews happen in months 6–12 — we're in that window.

How one operator becomes a reply

Walkthrough of the #1 ranked prospect this week

2.1 / 5
Europcar On Demand · The signal
App Store rating · no recovery since the 2024 rebrand

The first message

Subject: On Demand at 2.1 — pattern from your last 90 days of reviews

On Demand sits at 2.1. MILES, SHARE NOW, Free2move are all above 3.5. The gap isn't fleet quality.

I tagged the last 90 days of On Demand 1-star reviews. Three themes carry almost all of them:

  • Unlock failures on first booking attempt
  • Charge disputes from duplicate authorizations
  • App crashes mid-booking flow

All three are platform-level, not vehicle-level — which is why ops fixes haven't moved the number. The 2024 rebrand was a re-skin; the underlying stack is what's hitting the ratings.

For context: GreenMobility (1,600 EVs across 4 Nordic countries) and Forest (14K e-bikes in London) both run on our platform. Migration at On Demand's fleet size: city-by-city, ~9 months end-to-end, zero customer-facing downtime.

I packaged the review analysis + the migration mechanics into a 3-page brief. No marketing, no deck. Reply "send" and it's in your inbox today.

— Can, Wunder Mobility

Pain2.1 vs category baseline of 3.5+ (MILES, SHARE NOW, Free2move) · three platform themes I tagged from 90 days of their reviews — gives them external context they may not be tracking, plus the reframe of what the 2024 rebrand actually was
ValueThe message is the value. A 3-page review-pattern analysis + migration mechanics delivered in exchange for one word in reply. Plus the "ops can't fix this" reframe — political air cover for the next leadership review.
ProofGreenMobility (1,600 EVs, 4 countries) + Forest (14K e-bikes, London) — both running on Wunder, operator-shape match: same fleet scale, same multi-country footprint
AskOne word in reply ("send"). Zero meeting friction. The artifact ships today; the conversation starts after they've seen the analysis, not before.

v2 — Pain Stack

A hypothesis of pain-based segmentation. Three non-obvious signals + one compound pattern.

v1 watches the obvious signals. v2 watches what nobody else is pulling for shared mobility, then stacks the weak ones into a single strong segmentation pattern.

City permit cycle
Operators 6–12 months from permit renewal with complaints surfaced in council minutes are in strategic crisis, not just operational pain. Source: EU TED, city transport authority board minutes.
Review trajectory, not level
A falling 3.5 is a hotter prospect than a static 2.5. Direction beats level — the pain is rising in their org right now. Easy to layer on the existing review agent.
Fleet count gap
Operators advertise fleet sizes; cities publish actual deployment counts in open-data feeds. A widening gap signals scale-down or reliability collapse — both vendor-stack pain.
The Pain Stack hypothesis

Single signals are weak — easy to dismiss as noise. Stacked, they segment the universe by shape of pain, not firmographics.

Permit renewal in <12 months + Trustpilot falling 0.5+ points in 90 days + fleet count gap widening (claimed vs deployed)
= a five-alarm operator before any v1 signal fires

The cohort isn't "carshare in DACH at 1K–5K vehicles." It's "operators whose pain has this specific stacked shape." The segment is named for its pattern, not its category.

Live scan · Nextbike · May 2026
✓
Contract churn. Lost Berlin (6,500-bike contract, expired June 2025, court-ordered bike removal January 2026). Lost Glasgow to Voi in a competitive tender (October 2025). Cardiff scrapped before that. The pattern is current and public.
✓
Service quality collapse. Trustpilot sits at 1.3/5 ("Bad", 629 reviews). The April 2025 v5 app launch triggered a spike of 1-star complaints: GPS failures, wrongful charges, refunds denied.
✗
Fleet count gap. Doesn't fire. Nextbike added more cities than it lost in 2025 and reported 32% revenue growth. This signal doesn't apply to operators in growth-by-acquisition mode — and the system tells you that, instead of forcing a fit.
v1 score 50 Rank #3
→
v2 score 100 Rank #1

v1 ranked Nextbike #3 on the slow PE-rebrand signal. Pain Stack upgrades Pain from 2 to 4 — contracts being actively lost, service collapsed. Nextbike moves to #1, and the playbook flips: go direct to the cities Nextbike just lost this week, not PE-side intros in Q3.

Pain Stack → message in action
Subject: After Berlin

Sebastian,

Three contract losses in 18 months — Berlin (court-ordered bike removal Jan 2026), Glasgow to Voi (Oct 2025), Cardiff before that. The April 2025 v5 launch dropped Trustpilot to 1.3/5 ("Bad", 629 reviews). The pattern is the loudest signal in European bikeshare right now — and I'd bet your last six all-hands have all referenced it.

I ran the same scan across your remaining ~300-city footprint. A handful have public permit reviews in the next 12 months, with the same complaint themes already surfacing in council minutes. The shortlist is in a 3-page brief.

I'm at Wunder Mobility. We power Forest in London (14K e-bikes, fastest-growing share scheme in the city) and a few other operators in the same size band. The pattern when an operator is losing cities to competitors is consistent: the platform becomes the visible explanation, whether or not it's the cause. Replacing it gives leadership a clean answer for the next council meeting — and a different starting position in the next tender.

Reply "send" and the brief is in your inbox today. No deck.

— Can, Wunder Mobility

PainThree contract losses in 18 months named back to them (Berlin, Glasgow, Cardiff) · Trustpilot at 1.3 after the v5 launch. The recipient knows all of this — naming it back signals "I did the work" and earns the rest of the email.
ValueShortlist of the next at-risk cities (Pain Stack applied to the remaining 300-city footprint) + the political reframe: "the platform becomes the visible explanation." Both are things they'd commission externally to get.
ProofForest (14K e-bikes, London) — operator-shape match: same scale, same modality, came off an in-house stack.
AskOne word reply ("send"). Same low-friction ask as the Europcar message — the conversation starts after they see the shortlist, not before.

Timing + pain is everything.

The stack

What v1 runs on, and what v2 adds.

Brain & orchestration
  • Anthropic Claude — Sonnet 4.6 for signal harvesting, Opus 4.7 for scoring + PVP generation
  • Claude Code subagents — one per signal type, run in parallel
  • MCP (Model Context Protocol) — for external data connections
  • Python + Anthropic SDK — with prompt caching for the universe-wide weekly run
v1 signal sources
  • Tavily / Brave Search — news, M&A, exec changes, funding
  • Apify MCP — LinkedIn posts, job postings, founder activity
  • App Store + Play scrapers — reviews and ratings (Apify actors)
  • Apollo MCP — company + people enrichment
v2 Pain Stack additions
  • EU TED API — public procurement notices for permit-cycle scanning
  • City open-data APIs — NYC, Madrid, Berlin, London (and others) for deployed fleet counts
  • Wayback Machine API — historical App Store / Trustpilot trajectory
  • Trustpilot scraper — sentiment-trend tracking, 90-day window

Why Wunder, specifically

Three honest reasons

AI-first GTM is the bet I want to be on

The direction you've publicly committed to is the work I want. Most B2B SaaS is hedging on "AI as a feature." The agentic layer Bojan is building treats it as foundational. That's the bet.

Hamburg

My brother lives there. That's it.

The buyer's pain is visible

Shared mobility is one of the few B2B categories where the signals — App Store ratings, reviews, news, job postings — are public and structured. AI-based GTM is unusually well-suited to this vertical. PVP almost writes itself when the rating is 2.1.