1000x an idea

MEMORANDUM

TO: Marcus Webb (CEO) & Executive Team FROM: Staff Product Manager, Tally DATE: October 24, 2023 SUBJECT: From Tip Calculator to Network Effect: The $1B Path for Tally

---

1. The Mechanism and the Dimension

Kenji’s beta proved that the tip-out calculator saves managers 25 minutes a night and drives worker app engagement from 34% to 81%. But the real mechanism that makes this 1000x bigger isn't automation—it is multihoming labor coupled with a viral consumer loop.

Right now, 41% of workers at beta restaurants work at another restaurant. Because the tip calculator demands they check the Tally app at close to verify earnings, they are carrying our brand into competitor venues. When they ask their other managers to switch, those inbound deals close in 9 days (vs. our 41-day average) at one-third of the acquisition cost.

The single dimension we must push is Cross-Restaurant Worker Identity. Right now, a worker with two jobs has two siloed accounts. By unifying them into a single identity, we transform Tally from isolated B2B restaurant software into a consumer-grade network where workers organically pull non-Tally restaurants onto our platform.

---

2. The Most Ambitious Version & Opportunity Sizing

The 1000x version is Tally Financial Network: a closed-loop financial operating system for hourly hospitality labor.

We stop selling software top-down to stressed operators alone and start growing bottom-up through the financial urgency of 190,000+ hourly workers. Tally becomes the infrastructure where workers instantly access, manage, and move their daily earnings across the fragmented restaurant landscape.

#### Opportunity Sizing (Range) * Conservative Case ($35M–$50M ARR added in 24 months): Driven purely by viral acquisition velocity. If we reduce sales cycles from 41 days to 9 days across a larger cohort of worker-referred restaurants, our organic acquisition rate triples, lowering CAC and accelerating ARR past $80M without expanding headcount. * Aggressive Case ($120M–$180M ARR added in 24 months): Driven by Same-Night Tip Payouts. Our payments partner allows us to pay out tips already earned the same night for $0.99, netting Tally $0.60 per payout. With 190,000 workers (plus viral additions), assuming an average of 3 shifts a week and a 44% adoption rate (validated by our survey), this creates a massive, high-margin transactional revenue stream that dwarfs our current SaaS subscription model.

---

3. The Execution Sequence

We have two engineering squads allocated for next quarter. Here is the strict sequence, prerequisites, and hard stop-conditions.

#### Squad 1: Unified Worker Identity & Viral Loops (Q1) * What they build: The underlying database architecture to link a worker’s identity across multiple restaurant profiles (with explicit worker consent), alongside the frictionless "Invite Your Other Job" referral and onboarding flow inside the worker app. * Prerequisite: Legal review confirming that linking accounts via worker consent does not violate data-sharing agreements with restaurant operators. * Stop Condition: If fewer than 15% of multi-job workers consent to linking accounts, or if viral referral conversion drops below 10%, we abort and keep accounts siloed.

#### Squad 2: Same-Night Tip Payouts (Q1) * What they build: Integration with our payments partner to enable same-night, post-close tip payouts for $0.99, ensuring strict compliance with state tip-pool rules (keeping managers/owners out of the pool) and avoiding early-wage-access licensing traps. * Prerequisite: Final sign-off from Compliance confirming that paying out already-earned tips post-close requires zero new state licenses (avoiding the 9-to-12-month licensing delay). * Stop Condition: If payment partner error rates on same-night liquidity exceed 0.1%, or if take-rate margins are squeezed below $0.40 per payout, we halt.

---

4. What We Should Do with Marcus’s Idea

Marcus’s vision—“Tally becomes the AI operating system for restaurants: inventory, menu pricing, marketing, reservations, all of it... That's how we become a $1B company”—is the right long-term destination, but the wrong sequence for next quarter.

If we try to build inventory, menu pricing, and reservations right now, we dilute our engineering focus and walk straight into a bloody feature-parity war with entrenched incumbents, while simultaneously losing deals to Crewbook on price (11% loss rate last year).

Recommendation: Politely shelve the "AI Operating System" breadth for 12 months. Instead, weaponize Kenji’s tip calculator.

By unifying worker identity and launching same-night tip payouts, we leverage our unfair advantage: the workers. We use consumer-grade financial utility to bypass the restaurant owner's sales resistance entirely. Once we own the liquidity of the hourly workforce across 50,000+ restaurants, then operators will have no choice but to adopt our broader operational suite—because their staff demands it. That is how Tally becomes a $1B company, starting with a 99-cent tip payout.

To: Marcus Webb and Executive Team From: Staff PM Subject: Make tip-out a worker-led distribution network—not an everything app

Recommendation

Give the two squads next quarter to test a bigger thesis: Tally can become the portable earnings account that workers bring from restaurant to restaurant, making workers a distribution channel for our employer business.

The dimension to push is reach across workplaces, not breadth of restaurant software. Kenji’s calculator may be the entry point to that network. The evidence does not yet establish a literal 1,000× opportunity; it establishes a mechanism worth testing before we commit to a broader platform.

1. The mechanism: one worker can recruit the next workplace

The calculator solves a nightly manager problem, but its more consequential effect may be on workers:

• Worker-app weekly activity rose from 34% to 81% at beta restaurants.
• 41% of beta workers also work at another restaurant.
• 29% of surveyed workers say they have asked another manager to adopt Tally.
• We recorded 57 worker-requested restaurant signups last quarter. Those deals closed in nine days versus our usual 41, with roughly one-third the sales spending per deal.

The potential loop is:

Accurate nightly tips → workers return at close → workers request Tally at another job → another restaurant adopts → more workers encounter Tally.

This could change acquisition economics, not just improve a feature. It also creates a reason to choose Tally that Crewbook’s free scheduling does not directly answer.

But this is not yet a demonstrated network effect. The beta was self-selected, the survey measures stated behavior, and we have not established how many of the 57 signups were incremental or originated in the beta.

2. The ambitious destination—and a defensible size

The most ambitious version is a worker-consented earnings network spanning restaurants:

• One identity connects a worker’s jobs, shifts, and earned tips.
• Workers see verified nightly earnings and can optionally receive already-earned tips immediately.
• A worker can ask a second employer to join through a lightweight invitation and onboarding path.
• Restaurants join first to deliver trusted tip accounting and settlement; scheduling and payroll expansion follow.

Eventually, the network could extend beyond restaurants already using Tally’s full suite. That requires proving we can verify earnings and onboard an employer without compromising compliance. It is a destination, not next quarter’s scope.

We can size a beachhead from our current footprint, not a national TAM:

Subscription sizing uses today’s average of approximately $4,844 ARR per restaurant. The acquisition scenarios are sensitivity cases, not forecasts: we must validate the referral denominator, incremental lift, saturation, retention, and pricing. They are not automatically additive to our existing 380 quarterly signups.

Payout adoption likewise discounts—or at its upper end accepts—the survey’s 44% stated willingness to pay; transaction frequency is an assumption. Not every worker will have eligible tips.

These are separate revenue measures, not a combined ARR claim. They show a potentially material business on our existing footprint. The larger, cross-platform opportunity remains unquantified.

3. Work backward: earn each expansion

Next quarter: establish the worker connection and test causality

Squad 1: Portable identity. Build consent-based linking of existing restaurant accounts into one worker account, with clear employer boundaries, revocation, and no leakage of one employer’s information to another. This is the stated one-squad-quarter investment; do not bury unrelated work in it.

Squad 2: Prove the distribution loop. Harden tip correctness and dispute handling, instrument worker invitations through restaurant activation, and run a controlled expansion beyond the volunteer beta. Randomize rollout or invitations at an employer/network-cluster level where practical, accounting for workers who connect clusters. Distinguish an invitation, a signed contract, and a restaurant actually running payroll.

Test referrals using existing accounts while identity work proceeds. Do not wait a quarter to start learning, and do not promise advances or build a broad financial product.

Gate to broader rollout: After at least two normal sales cycles, demonstrate incremental, retained restaurant acquisition versus a comparison group—not merely faster attribution of deals we would have won anyway. A provisional investment hurdle is 0.05 incremental activated restaurants per participating restaurant per quarter, roughly one-quarter of the provisional beta rate, with lower fully loaded acquisition cost than ordinary sales. Set retention and tip-accuracy guardrails before launching.

Next: monetize earned tips without weakening the loop

Once verified tip accounting and payout reconciliation pass Compliance review, Squad 2 can pilot the partner’s $0.99 optional same-night payout; Squad 1 extends proven identity and invitation flows. Preserve free tip visibility.

Gate to expansion: Observed repeat purchases—not survey intent—produce positive contribution after support and losses, with no material increase in disputes or deterioration in worker engagement. Failure here means dropping paid payouts, not automatically abandoning referrals.

Then: open the network beyond full-suite customers

Only after acquisition and retention replicate outside the beta should we test a lightweight tip-and-earnings entry product for employers not ready to switch payroll.

Gate: Verified earnings, enforceable pool rules, economical onboarding, and incremental customer value without unacceptable full-suite cannibalization.

Stop result: If an adequately powered test over at least two sales cycles shows no meaningful incremental retained restaurant acquisition—or economics below our agreed hurdle—we stop funding the network expansion. Keep the useful calculator. Pause immediately for material tip-allocation or privacy failures.

4. What to do with Marcus’s idea

Keep Marcus’s ambition; reject the proposed scope for this planning cycle.

Inventory, menu pricing, marketing, and reservations each require new data, workflows, and distribution advantages that this pack does not establish. Calling them an AI operating system does not connect them to our strongest evidence.

Reframe the vision as “Tally becomes the earnings network workers bring to every restaurant.” AI can support that strategy where it demonstrably improves outcomes; it is not the strategy itself. Defer advances and their licensing burden. Return to broader platform bets only after we prove the mechanism that earns us the right to expand.