YC SUS: Eric Migicovsky hosts founder office hours
These notes were generated by AI and may contain inaccuracies.
Eric Migicovsky, course facilitator at Startup School, hosts online office hours with Startup School companies. Office hours are 20-30 minute sessions where YC partners dive into a founder's specific problem and brainstorm solutions. These are accelerated sessions with companies signed up via Startup School forums, conducted publicly. Advice is kept focused for broad value. Office hours help founders escape isolation, talk through problems, and solve them by articulating them aloud. Recommendation: founders talk founder-to-founder about problems.
Company: Boxscore (boxscore.live). Product: Beta futures gaming platform with trivia about televised future events (mostly sports like NFL playoffs, entertainment like Oscars). Users predict outcomes (e.g., who wins match, best picture) and play against friends in groups, not the house. Eventually gambling/betting, but not monetizing yet.
Stage: Live beta with 240+ users last month for NFL playoffs, 20 active users now, 300+ signups (half organic, half promo). Proof of concept as second screen during live events (e.g., users requested scoring for Oscars). Solo founder since June, launched beta, hit wall end of January due to burnout. Accepting $50k investment, hiring engineering students from alma mater in Mexico at $5/hour (one working, couple more potential). Plans to raise more.
Future plans: Add utility (live scores, previews, recaps) better than existing platforms using real-time user preferences (e.g., from event selections). Covers multiple sports (Grand Prix, World Cup, Super Bowl) as sports fans follow multiple year-round, targeting Mexico (Liga MX, Champions League, European leagues). Aims to be comprehensive second screen for all followed sports, showing performance vs friends, stats during match.
Advice from Eric:
- Test smallest thing users want that existing apps lack (e.g., trivia for specific team/sport).
- Focus on one sport/team initially for virality (e.g., soccer in Mexico) rather than all sports.
- Don't wait for Euro 2020 or Olympics (August); use existing 20 users. Talk to them weekly for addictive features.
- Iterate weekly: talk to users, identify problems blocking activity, assign <1 week engineering batches, test fixes.
- Sunday meetings to plan next week and measure prior week.
Jose confirms weekly user talks (500+ predictions from some), regular contact.
Company: IQ (hard tech, new consumer tech category: smartphone as smart binoculars). Inspired by Star Wars macrobinoculars/Harry Potter Omnioculars. Plastic holder with proprietary touchpad (maps to app for camera control), lens magnifies phone camera. Features: zoom, view, capture, live stream events (e.g., nosebleed seats at games). App: IQ Space Live (iOS App Store, eye-ball-Q not letter I). AR app in development (e.g., yellow line at football games).
Stage: Prototype ready. Kickstarter raised $27k goal, now $30k+ (200+ backers worldwide). Half seed round closing soon. Tooling done, first article ready. Pilot production to ship to backers and seed market before 2020 Olympics (July timeline). Major leagues interested.
Team: Founders (Mary, CTO mechanical/optical engineer/inventor), sports community folks, software engineer for AR app.
Challenges: Supply chain risks (China manufacturing amid coronavirus). Factories for first plastics.
Advice from Eric:
- 3D print parts instead of plastic injection molding; customers fine with it (YC companies sold thousands 3D printed). Avoids China factories.
- With Kickstarter funds, order components now for 300 units (expect 2 months shipping); weekly WeChat check-ins with factories.
- Focus on fulfilling 300 pre-orders first, not scaling to thousands (delays risk backer happiness/word-of-mouth).
- Run small pre-product-market fit (just founders ideal; low burn for breakeven/profitability). Use contractors for short needs (e.g., electrical engineer 2 weeks), not full-time.
- For developer interest: Email 300 Kickstarter backers asking for hackers/builders. Ship first; users will hack (phone-based).
Company: Cargo Labs Incorporated (fintech for truckers). Product: Trackport mobile app (launched December). Tracks time trucks spend at shipper/receiver during loading/unloading. Runs on phone.
Problem: Trucks wait long at shippers/receivers. Times unclear to brokers paying them; hard to invoice for wait rates.
Primary user: Starts with trucker (and...? transcript cuts off)
The app primarily helps truckers and trucking companies get paid for wait times at shipper/receiver docks due to inefficient loading/unloading processes. It calculates detention time using GPS on the trucker's phone, applies a detention rate, and generates an invoice for the owed amount. Truckers run it on their own phones. Example: At a Walmart Distribution Center, truckers wait 30 minutes to multiple hours without pay. This detention pay is typically outside the main contract, handled verbally as a gentleman's agreement.
They launched in December with 7-8 truckers using it. Summerfield Trucking added it to their website; they run an accounting firm for 25 other trucking companies. Truckers use it selectively for problematic locations, encountering issues 2-3 times a week (e.g., 14 stops tested last week across 5-7 orders, though not all consistent due to bug fixes). They have contact info (phone/email) for all seven, including RL, an owner-operator who has been most helpful.
Currently working on a dollar-backed cryptocurrency processing system with smart contracts to instantly move money from broker to trucker upon order completion, reducing 30-day payment delays. Eric questions necessity: Truckers already get paid in 30 days for main loads; app adds extra $10 for detention on top of $100 load pay. Suggests adding to existing invoice via simple processing like PayPal/Venmo without crypto.
"I would highly recommend keeping it as simple as possible and potentially just reconciling it in cash or Venmo or PayPal or even the same contract."
Eric advises focusing on core pain point (detention pay) with minimal solution to test hypothesis quickly. Adding smart contracts delays product-market fit validation. Signs of fit: Users tell friends, use for every load, request features/bugs proactively.
Bootstrapping. Spoke to first broker yesterday for partnerships to take processing fee on funds movement (PayPal allows marketplace fees). Truckers sometimes denied main payments due to unverifiable detention time. Eric pushes simplicity: Use check/PayPal/Venmo without broker deals initially.
Next week goal: Double users to 14 total, get them using on majority of loads. So far, app logs times accurately; auto-invoicing ready but not activated due to discomfort. Can manually send invoices.
Nate from Meme Party (meparty.game.com): Mobile party game on iOS/Android for up to 8 friends. Each round, players edit the same meme image macro with text; everyone votes on the best. Like Apples to Apples or Cards Against Humanity, but competitive meme creation. No other app focuses on competing to make memes.
56 weekly active users last week: 20 from brewery event with friends/strangers; also an in-person event. Primarily weekend use with friends in-person (everyone opens app together, like Space Team).
iOS released 3 months ago, Android a week ago. Transitioning to marketing: Going to breweries several nights/week, approaching groups (Lucy handles outreach). Social media/flyers ineffective since it's not solo-use. In-person conversion ~100%: Ask groups to test app for feedback, they play 10-15 minutes.
Unclear if brewery users return later (no prior tracking; adding Mixpanel now). Plan more brewery visits this week. Eric suggests in-app messaging/push notifications (e.g., Friday meet, Saturday reminder: "Hanging with friends? Try again.").
Other ideas:
- Post user-created memes on social media.
- Record short "how to get started" video.
- Target card shops (Magic, Warhammer), trivia nights, pubs.
- Offer to run trivia night via app if bar patrons download.
- Observe first 30 seconds: Setup, connecting friends—fix bugs live in-person.
Main metric: Weekly users (party game, not daily). Goal: 100 WAU next week (from 56). No revenue yet (free app); focus growth/retention/churn first before ads/in-app purchases/upgrades.
Tournament idea like Players Lounge (YC company: bet on self, win money from others). Eric: Standard gaming models suffice; innovate on game, not business model. Crack retention first.
Cheffy from WP Clover: Software engineer running lifestyle company transitioning to platform/service for WordPress development, hosting, maintenance targeting small businesses/startups building MVPs.
Founder previously built WordPress sites for individual companies, now transitioning to a hosting platform for small businesses to start websites easily on WordPress. Recommendation: Focus messaging on helping small businesses launch their first or good website, with WordPress as the underlying platform rather than the main emphasis.
They've always hosted client sites but previous method wasn't scalable. Now automating processes to minimize hands-on work while delivering value. Customer onboarding: Send ticket for fixes, repairs, content additions; service provides 1 hour/day to fulfill tasks. Customers don't need to be technical or know it's WordPress—e.g., "change the menu on this restaurant site" or "add a holiday to this calendar."
Currently 3 customers, onboarding 3 more next week (doubling to 6). Easy transition from previous customers. Biggest challenge: Finding target market. Lifestyle company targeted women, but market research and ads show audience is largely men. Changed website to women-focused feel, but converting more men. Website: wpclover.com—tagline "Be great with WordPress hosting, development, maintenance on one service."
Not necessarily for people who understand WordPress. Targets: (1) Existing WordPress users who don't know how to work with it, (2) Brand new websites, (3) Migrating WordPress sites to avoid management.
Consider different landing pages for segments. Current page good for existing WordPress users frustrated with management. For new businesses wanting a site handled: separate page needed. They have "office hours" for WordPress discussions and web development page with solutions for freelancers, startups, businesses—but lacks appeal for businesses. Inhabit small business mind: Problems like getting listed on Google, Google Maps. Use customer conversations to identify pains and address on site (e.g., "We're going to solve your problems").
Broaden beyond WordPress as transitioning to solutions provider.
Started October; initial info too technical, not closing deals (transitioning from enterprise open source to B2B/B2C). Simpler content: Blog post "How to get a yes" most popular; simple YouTube snippets.
Advice to be less technical while technical: Talk to friends with the problem (trust factor). Examples: Restaurants, small physical businesses, internet consulting freelancers, startups needing web presence.
Ask friends: Recall setting up first website—what did you want, who did you consult, search terms used? Post-setup problems with WordPress/Wix/Squarespace. Enumerate specific pain points. Helps break from daily specifics.
With friends: Ask many questions, have them open dashboard, observe buttons clicked, recent changes. Hyperfocus on LinkedIn distracting; friends want to help. Do in person or focused video call (e.g., "In neighborhood, hang out 30 min, bring lunch"; share screen).
Product built: Dashboard + ticketing system. Have staff. Challenge: Getting buy-in; higher cost than Wix but cheaper than $3,000 WordPress developer.
Advice: Segment user base in spreadsheet—e.g., no site, Wix users, WordPress users, hired WP developer. Log conversation data to shift from anecdotal to data-driven (e.g., "Consulting businesses in US with Wix seeking upgrade"). Talk to 5 people per 5-6 categories (1-2 weeks), pick best (e.g., category B). Then find more: Facebook groups, Reddit, Twitter hashtags, meetups/conferences.
No need to hyperfocus immediately—do legwork first, then experiment finding 50 more in top category.
Great meeting; follow up via tweet after doubling to 6 users and segment experiments. wpclover.com.
Audio issues resolved (AirPods, YouTube stream closed).
Wusu (wusu.io): Platform for rotational savings groups. Low-income example: Save $1,000 in 10 months solo vs. 3 months in micro-group. Like small-group crowdfunding.
4 people need $100 at different times: Each puts $25/month into pot; one takes $100, cycle repeats until all get $100.
Pre-launch: Interactive demo (shared individually), landing page. Open platform for anyone (not just known people).
Concern: Feels pyramid-like (put in before getting out). Response: Circular, not pyramid; 800+ years history, cultures worldwide. But public groups risky vs. family/small private groups (like small crowdfunding).
Founder ran 3 groups personally. App looks slick but build slower—hack MVP now: Use Venmo/Square Cash + Google Sheets. Founder arranges, tracks payments, enforces (ban non-payers). Launch groups tomorrow. Share idea with family/friends for sign-ups, phone numbers; use iMessage group, new sheet per group for tracking commitments/payouts.
Test hypothesis manually first, then minimum app features. Not a developer—use no-code tools (Startup School threads).
Fintech, regulated (money service business, money transmitter license?). For app MVP with platform cut: Use PayPal/Stripe Marketplace features (they handle transmission, take 3% cut). Or check WePay (YC company, similar chipping in for events).
Goal: Manual group next month. Follow up via tweet.
Phabs: Software moonshot to predict genetic evolution, e.g., better flu vaccines (relevant now).
Flu vaccine historically <50% efficacy. Scientists decide next year's now via educated guess on DNA.
Use ML on public data (tens of thousands influenza sequences yearly, government databases, last decade+) to predict better.
Data not hard; bottleneck: Computational analysis. Can't analyze all—even reconstructing last epidemic: Scientists process ~500 samples, 1 month compute time.
Stage: Prototyping on small real datasets. Founder: Programmer, computational biologist, statistician (founder-market fit). Backtesting on past vaccines/sequences.
Discussion on milestones since predicting flu vaccines is big milestone and so of this
Arman explains the smaller milestone: analyze more than 500 sequences in less than one month, ideally one day or one hour. Eric asks for clarification on "analyze 500 sequences." Arman says the big goal is predicting next year's flu, but they can't even understand last year's evolution because they can't analyze all the data in a reasonable timeframe. They can't answer which mutations happened last year or how flu spread.
Public datasets of past years' flu DNA exist, with tens of thousands of samples, some duplicates but many different because flu evolves quickly. Each sequence represents one randomly sampled person, providing frequency data.
First milestone: take a scientist's 10,000 or 1,000 DNA sequences and build a picture of the epidemic in one hour—diagram showing how influenza mutated and spread.
Eric questions why one hour is important. Current software takes at least one month. Eric suggests starting with less than a month, like two weeks. Arman agrees that's fine but says it's easier to aim for one hour because current software is a serial algorithm that doesn't scale with big computers; his does.
Revised first milestone: be twice as fast as current software (e.g., two weeks vs. four weeks). Hypothesis: once hitting 2x, can easily scale to 10x or 20x.
Arman works with epidemiologists and flu vaccine scientists who could switch. First usable software (less than four weeks) is a few months away; has been working full-time for a couple years, with setbacks on user data.
Users dump data passively. Next, coordinate supercomputers. Eric advises focusing on primary problem: talk to scientists using slower software, ask how much they'd pay to switch. Start with crunching numbers from one sample in three weeks vs. four.
Next month goal: benchmark current build, soft launch on landing page claiming 2x faster than existing software. Eric offers to tweet the site; has a friend who may be interested.
Amy from Curved Wear (curvedwear.com): provides pants for curvy women for travel/outdoor adventure with technical features—moisture-wicking, quick-dry material, deep functional pockets. Comfort, style, function not currently available.
Struggles with sample prototype from LA manufacturer (she's in San Francisco); has visited, brought samples, but latest pre-Christmas sample was opposite of vision—fit for size 2 mannequin-like builds.
Eric suggests: schedule full weekend workshop with manufacturer to outline motivation and deliverables clearly, or switch suppliers. Don't get committed to sunk costs.
Brainstorm cheaper prototyping: learn pattern-making/sewing herself via course (policy researcher background, new to apparel), local college teacher/interns, sewing clubs, Facebook/Instagram groups for home sewers (e.g., children's clothing patterns), potential co-founder.
Goal: make one pair for herself to photo/video for crowdfunding, test if it solves problem, regain motivation after setbacks.
Build iteratively: make one, then two, five, ten—handmade one-offs sold at higher price to women with the problem. Has 40 sign-ups; make five for them at double price.
Don't focus on social/digital marketing yet; main problem is make one and sell one to friends/high-need people first. Hardware advice: don't aim for 1,000 via crowdfunding like Shark Tank; iterate small batches.
Lindy from Work Life: provides mental health care as workplace benefit—every employee gets regular company-funded confidential therapy, regardless of issues.
Eric advises pausing after core pitch to let it sink in, confirm understanding.
Difference from competitors: others have barriers (must have issue, tech first, unlicensed therapist). Work Life proactively signs up all employees for therapy, normalizing it, preventing issues.
Opt-out model (like organ donation: US opt-in ~15% signup; opt-out higher). Confidential.
ROI: therapy increases productivity 73%, reduces turnover 67%, sick days 62%. Tech company can make $50-200K per employee via therapy benefits.
7-8 weeks past idea stage. Cold emailing/calling/networking friends for first B2B sale.
Eric recommends: start at previous workplace (where she experienced problem); they know/trust her. If no sale, diagnostic. Then use as reference for intros via HR.
Don't mass-email personal contacts; call for coffee, specific ask: "Introduce me to HR Director so I can call." In-person harder to refuse than email; gather feedback from reactions.
Do user interviews per The Mom Test: talk to HR about top 3 employee problems (e.g., absenteeism from seasonal disorders/sickness). Let them bubble up mental health; then position solution.
Pitch 50-60 people in-person (shared rides). In-person for warm intros too—show up at office for 20 mins.
You're only eight weeks in, so you still have a lot to learn about the problems HR departments are running into. My last job was actually interviewing HR departments about how amazing they had, which is where the idea came from. I also have those connections where I can follow up with them and say, "Hey, I talked to you in my last job."
Sounds like you're in the right business.
All right, awesome. Thank you so much. I really appreciate it.
Happy to help. Send me a tweet in a couple weeks. I want to catch up and hear how things are going.
Okay, thanks. Thanks.
Lindy, cool. Well, hope you lasted the whole way through those two hours with us to get a sample of how office hours go down. If you'd like to do these, we'll probably run a couple over the next few weeks. We'll post it on the Start School forums—that will be where you'll hear about these for setting up. Thanks very much.
Keep Y Combinator in your library
Save the videos and channels worth coming back to, and find them again in one place.
