Why this quarter
Pacific Avenue closed the take-private on March 19, 2026 — 32 days ago. Fund II's first investment, so Year-One execution is the fund's thesis proof. Board deck being drafted now. After day 100 the story calcifies.
Ten consumer trust-handoff marketplaces where an AI agent routes a high-stakes request — a caregiver, a handyman, a tutor — to a human provider, and the context collapses on the way. The order and shape of the fifteen days is designed so each wave compounds the intel we use in the next.
One vertical per wave so reply evidence compounds — a Care.com reply sharpens UrbanSitter before send; Angi sharpens TaskRabbit. Cross-vertical signal lands by end of Wave 2.
| # | Account | Buyer | Vertical | Day | Why this, why now |
|---|---|---|---|---|---|
| 01 | Care.com | Brad Wilson CEO, narrative-first operator |
Care | T+1 | Pacific Avenue PE closed the acquisition 32 days ago; Year-One board deck being drafted. CareBenefits is the anchor line. Highest-leverage dated signal of the ten. |
| 02 | Rover | Brent Turner CEO, promoted from COO (Q4 '25) |
Care | T+2 | Blackstone-held; Q4 '25 earnings named "AI-powered personalization" as a 2026 roadmap pillar. Repeat-book rate decay after first mismatch is an open problem on the pro forums. |
| 03 | UrbanSitter | Lynn Perkins Founder-CEO, hands-on |
Care | T+4 | Trust-graph as wedge since 2011 — the category's only friend-of-friend referral model. Founder-led raise environment in 2026 rewards AI-native growth primitives. |
| 04 | Angi | Jeff Kip Public-company CEO, turnaround operator |
Home | T+6 | Post-IAC-spin Year-Two turnaround. Q4 '25 earnings call named "AI-enabled pro-matching" as a focus area. Lead-quality class-action history makes context-enrichment the cleanest pitch shape. |
| 05 | Thumbtack | Chris Patalano Head of Product |
Home | T+7 | Mid-execution on "home-management" repositioning; auction-based matching is in structural tension with a stateful-homeowner thesis. Open senior product role as of early 2026. |
| 06 | TaskRabbit | Ania Smith CEO, IKEA-owned platform |
Home | T+8 | IKEA 2025 AI-strategy update referenced agent-enabled customer experiences. Return-reduction is the cleanest IKEA-native KPI TaskRabbit can move — tasker briefing is the lever. |
| 07 | Handy | Umang Dua GM / Co-founder inside Angi |
Home | T+9 | Angi-owned vertical specialist — can move faster than the parent. Natural proof environment for the Angi-wide AI-matching thesis. Founder still deeply product-involved. |
| 08 | Preply | Dmytro Voloshyn Co-founder / CTO |
Tutoring | T+11 | $120M Series C late '24; infra-investment mode. AI Tutor Assistant launched early '25 — a first-party agent already handing off to tutors in production. The context-handoff problem is live, not theoretical. |
| 09 | Wyzant | Paul Mishkin CEO, IXL-owned portfolio |
Tutoring | T+12 | Less AI-forward in public signal — which is the thesis. Platforms that haven't declared strategy are where the handoff problem is most acute and least acknowledged. SAT-prep organic traffic is the leading indicator. |
| 10 | Outschool | Xueyan Wang Founder-CEO, mission-led |
Tutoring | T+14 | Live-class format makes context-passing higher-stakes than async tutoring. Strong parent-brand trust at risk without an AI story. Founder replies fast to sharp peer-tone. |
Pacific Avenue closed the take-private on March 19, 2026 — 32 days ago. Fund II's first investment, so Year-One execution is the fund's thesis proof. Board deck being drafted now. After day 100 the story calcifies.
Narrative-first CEO. Streaming pedigree (HBO Max, Disney+). Reads business as consumer-product story, not marketplace-ops. Owns the Year-One board narrative and employer-partnership strategy.
AI-originated demand (ChatGPT, Alexa+, Siri routing "find a caregiver") hits Care's front door in 12 months. The handoff is stateless — parent tells agent five things, caregiver sees three keywords. Match conversion gates CareBenefits retention.
90-day CareBenefits cohort pilot, outcome-tied on match-conversion uplift vs. control. 20 min next Tues / Thurs. Alt if no reply by T+10: Chris Sznewajs, Pacific Avenue MP — register shifts to PE Fund-II thesis.
Blackstone took Rover private at $2.3B in 2024. Brent promoted from COO in Q4 '25 — first full year as CEO of a PE-held company is the exact window where an AI-infra partner gets signed. Q4 '25 earnings flagged "AI-powered personalization" as a 2026 roadmap pillar.
Ops-first CEO promoted from COO. Thinks in unit economics and retention cohorts. Graded on repeat-book rate and marketing payback. Fears supply churn and a post-incident trust collapse.
Rover's repeat-book-rate decay after the first sitter mismatch is well-documented in Reddit and App Store reviews. The root cause is context loss between booking and sitter handoff. Foible carries the allergy, the routine, the anxious-dog history across.
20 min. Pilot scoped to boarding vertical, one metro. Alt: David Hayes, CPO — register shifts to roadmap-pillar execution.
Smaller than Care or Rover. Founder-led since 2011 — Lynn reads her own email and makes vendor decisions herself. 2026 raise environment rewards care companies with AI-native growth primitives. Trust-graph (friend-of-friend sitter referrals) has always been the wedge.
Founder-CEO, hands-on, product-led. Thinks in parent-trust and neighborhood density. Owns everything. Replies fast when the pitch is sharp.
Trust-graph is already a handoff model — friends vouching for friends. The natural evolution is AI-originated handoff that carries the same fidelity. UrbanSitter is positioned to own the "AI-plus-trust-graph" category before Care or Rover can.
20 min. Pilot sized to one metro. Alt: direct — Lynn is the decision.
Spun out of IAC March 2024. Jeff took CEO mid-2024 — Year-Two turnaround window is the critical one. Q4 '25 earnings call repeatedly referenced "AI-enabled pro-matching." Pro-side lead-quality history (class-action, documented on Reddit) is the inflammation point.
Public-company CEO, finance background, turnaround operator. Reads in EPS terms. Owns Street narrative and pro-side retention. EA-triaged mailbox — responds to commercial framing.
Angi's lead-quality problem is a context-pass-through problem — pros get a keyword job description instead of the homeowner's actual situation. Foible is the brief-enrichment layer; your ML team sits downstream of it.
25 min. Pilot sized against pro-side lead-acceptance rate. Alt: Umang Dua at Handy — which is exactly our account #07, and serves as the natural proof environment.
2025 repositioning around "home management" is mid-execution. Product is shifting from lead-marketplace to homeowner-platform. Open senior product-leader role as of early 2026 — decision-makers are looking to make a mark before the org firms up.
Product leader. Reads in user-journey + funnel-stage terms. Graded on repeat-use rate and time-to-first-quote. Reads for mechanism specificity.
Home-management requires stateful context about each homeowner — the previous plumbing call, the aging water heater, the HOA rules. Auction-based matching can't price for that. Foible is the stateful-context layer; profiles are static, this is live across jobs.
20 min. Pilot scoped to repeat-use cohort. Alt: Marco Zappacosta, CEO — register shifts to repositioning thesis.
IKEA-owned since 2017. IKEA's 2025 AI-strategy update publicly referenced agent-enabled customer experiences. Return-reduction is the cleanest IKEA-native KPI TaskRabbit can move — every unassembled bookshelf returned is a cross-subsidy audit risk.
Corporate-platform CEO. Thinks about parent-company P&L as much as platform P&L. Owns the IKEA cross-subsidy narrative. Fears: losing the subsidy, Tasker churn, a brand-damaging quality incident.
Tasker-briefing quality is the upstream lever on return-reduction. An IKEA customer who ordered an IVAR shelving system tells an agent about their ceiling height, pet in the apartment, lack of a drill. The Tasker currently sees "assemble IVAR." Foible carries the brief.
20 min. Pilot scoped to assembly vertical, one metro. Alt: Brian Leonard, CPO.
Angi-owned specialist for pre-vetted home-services pros. Parent-company AI commitment cascades — Handy can move faster because it's smaller. Handy is the demo environment for the Angi-wide AI-matching thesis. Umang is still deeply product-involved.
Founder inside a corporate parent. Runs Handy like a startup P&L inside Angi. Fears: losing autonomy to parent initiatives, getting absorbed into generic Angi product.
Handy is the right environment to prove the AI-enabled matching thesis before it rolls to Angi-wide. Pilot here; parent inherits the evidence. Scoped to cleaning category first — shortest conversion cycle.
20 min. Alt: swing the thread up to Jeff Kip (#04) as the parent-company sponsor once pilot shape is agreed.
$120M Series C closed late 2024. Post-Series-C, pre-IPO = infrastructure-investment mode. AI Tutor Assistant launched early 2025 — a first-party agent already in production handing off to human tutors. The context-handoff problem is live on Preply's system, not theoretical.
Technical founder-CTO. Reads in system-architecture terms. Makes vendor decisions on technical merit. Owns platform engineering, AI roadmap, agent-to-human handoff architecture.
Your AI Tutor Assistant knows more about a learner after 20 minutes of conversation than the tutor sees on session intake. The gap between agent-context and tutor-briefing is our pilot. We sit between them.
30 min technical deep-dive (CTO register). Alt: Kirill Bigai, CEO — commercial register.
IXL Learning-owned, US-focused, high-intent tutoring. Less AI-forward in public signaling — which is the thesis case. Platforms without a declared AI strategy are where the handoff problem is most acute. 2025 SAT-prep organic traffic is declining as queries route to ChatGPT.
Education-platform CEO. Outcome-first, parent-company-aligned. Owns Wyzant P&L and learner-outcome metrics. Fears: AI substitution eroding the funnel, losing ground to Preply on AI narrative.
Reframe: the AI-substitution threat is actually a context-redirect. Students who asked ChatGPT first arrive at Wyzant with richer context, not less need for a tutor. The tutor just doesn't see what ChatGPT already heard. Foible does.
25 min. Pilot scoped to SAT-prep vertical. Alt: Paul Mishkin direct — this is the direct.
K-12 enrichment and live-class tutoring. Strong parent-brand trust at risk post-pandemic-normalization. 2025 focused on retention and cohort quality. Live-class format makes context-passing higher-stakes — a teacher has 60 seconds to start well or the learner checks out.
Founder-CEO, mission-driven, education-native. Thinks in parent-trust and learner-safety terms. Founder mailbox — replies to sharp peer-tone.
The first five minutes of an Outschool class is when retention is decided. Teachers currently ask the learner who they are. That's the context-handoff gap. Foible pre-fills the first five minutes so the teacher starts in-session.
20 min. Pilot scoped to one live-class cohort type. Alt: Amir Nathoo, Co-founder.
Care.com is #1 because the Pacific Avenue board deck is the sharpest single dated signal of the ten. Wave 1 leads with the highest-density signals so voice and mechanism get calibrated against real replies.
Three accounts per vertical, one vertical per wave. A Care.com reply sharpens UrbanSitter's send. An Angi reply sharpens TaskRabbit's. By end of Wave 2 we have cross-vertical pattern signal.
Care → Home → Tutoring is also a tightening of handoff stakes. A tutor has 5 minutes. A caregiver has a 90-minute first-meet. The pitch matures with the audience.