Refocus & Game Plan

CaterZen

A 30-day reset on what you actually need. Less sourcing. More delivery. Built around what you said on our calls — not what we assumed you wanted.
April 30, 2026 · Prepared by Ricki / Cynthia
Cynthia.

You don't need more leads. You need to close the ones already in front of you, faster.

When we did our growth call on April 14, we agreed on a big go-to-market motion: cold email, FB ads, an outbound funnel hitting EZ Cater users. Smart on paper. But on the April 23 follow-up, you brought Avi and the entire conversation was about delivery — vibe-coding restaurant sites, restricted UIs for end-customers, AI voice receptionists, your CaterZen lead dashboard. Not a single question about lead volume.

That tells the truth. Your bottleneck isn't finding caterers to talk to. It's turning the ones you're already talking to into closed customers fast enough. Brooks Grocery is the proof — you mystery-shopped 3 delis, ran a ChatGPT phone audit, built them a catering page, and they're meeting with you next week. That motion works. But each one takes you days of personal effort.

So this plan stops trying to fill the top of the funnel and starts compressing the middle. Make every prospect take 2 hours instead of 2 days, and you 10x your close rate without changing anything else.

Three quotes that reframed everything.

"I've mystery shopped three of their delis. They only have three, you know, three stores. Took the transcript and ChatGPT, and did a phone audit. So we're basically gonna do a catering growth audit. We're gonna incorporate some things we know, like how to use the phones because, ultimately, we want AI voice to be able to do leads." — Michael, April 23, on Brooks Grocery

Translation: your sales motion is already proven. Mystery-shop → audit → pitch → close. You don't need a new sales motion. You need that motion to take 2 hours per prospect, not 20.

"I've coded a dashboard for taking all leads, whether they're forms online, or they create a proposal online or a sales rep has a lead that they're talking to on the phone. We've vibe coded — it's not gonna be done anytime soon because there's a lot of stuff to get done." — Michael, April 23, on his lead dashboard

Translation: your most-wanted product is half-built. Stuck because you're solo on the build side. Avi is helping but you're his only project owner. You'd take real help shipping it.

"You're dealing with restaurant owners that aren't super tech savvy. So what's gonna be the most efficient way for them to be able to make changes?" — Michael, April 23, on the end-customer experience

Translation: your customers are Bubba BBQ. They will not log in to Lovable. They will not type a prompt into Claude. The product surface they touch has to be radically simpler than the tools you and Avi use to build it.

An honest accounting.

Since you started paying $2K/month on April 10, we've delivered the following. Some of it serves what you actually need. Some of it doesn't — not yet.

SHIPPED
Full growth playbook (April 15)

The 18-section handoff doc covering FB ad funnel, 9-persona AgentMail outreach, 24-hour Next.js sites, programmatic SEO engine, AI voice via Vapi, lead enrichment via Apify + FireCrawl, SDR scripts, ICP targeting, pricing model, demo URLs, API keys.

Useful? Reference material — yes. Operational right now — no. Most of it requires you (or Avi) to stand it up on your own infrastructure first. Sitting in vault waiting.

SHIPPED
Reach platform tenant + 2,615 caterers seeded (April 30)

Provisioned your account at reach.cynthiaconcierge.com (login: michael@caterzen.com / cynthia2026). Loaded 2,615 unique catering businesses scraped from Google Maps across 24 metros. Classified by 14 catering verticals (BBQ, mobile food truck, corporate drop-off, wedding, Italian, etc.). Scored 0–100 on ICP fit.

2,615
caterers in graph
1,389
A-band (≥80, ready)
145
BBQ caterers
31
corporate drop-off

Useful? Yes — if and when you want to outbound. Today, you don't. So this is fuel waiting for a fire that's not lit.

DEFERRED
8–12k national expansion scrape

Designed an 810-query Apify expansion to push the cohort from 2,615 to 8–12k caterers nationwide. Hit a network blocker (Hetzner can't reach api.apify.com right now). Holding off on retrying until the sourcing motion is actually needed — per the rest of this doc.

The honest read: we've built a great sourcing engine you didn't ask for. It's there if you change your mind. But three weeks in, the right move is to stop building outbound infra and start building the thing that actually closes prospects you already have.

The math: where time goes per closed deal.

Brooks Grocery is your template close. Let's break down where the hours actually went.

StageToday (manual)If templatedTime saved
Mystery-shop the prospect's locations~3 hrs (call each, take notes)30 min (auto-call via Vapi, GPT transcript)2.5 hrs
Run phone audit through ChatGPT~2 hrs (manual prompts, edits)10 min (templated prompt + scoring)~2 hrs
Audit the prospect's website~2 hrs15 min (auto-pipeline already built)~2 hrs
Build site mockup / catering page~4 hrs (Avi in Lovable)30 min (Claude Design one-shot)~3.5 hrs
Compose pitch / proposal deck~3 hrs30 min (templated, auto-populated)~2.5 hrs
Run the sales call1 hr1 hr (no change — this is your edge)
Total per prospect~15 hrs~3 hrs12 hrs
At 15 hrs/prospect, you can pitch ~3 per week. Even with a perfect close rate, that caps your monthly closes at ~12.
At 3 hrs/prospect, you can pitch ~13 per week. Same close rate, same operator. 4× more closed deals per month with no extra people.

Build the Audit-to-Pitch Toolkit. One thing. Two weeks.

One small piece of internal tooling. You feed it a prospect URL and a phone number. It does the work that takes you 12 hours today, in under an hour. You spend the rest of the week on calls, not building decks.

PIECE 1
Auto mystery-shop (Vapi + GPT)

Trigger from a single form: enter a prospect's name + number of locations. The system auto-calls each location via Vapi voice agent posing as a catering customer ("Hi, I'm planning a 75-person event next month, do you do catering?"). Records, transcribes, scores against your phone-audit rubric. Outputs a 1-page PDF per location.

Why this works for you: the Brooks Grocery prospect was sold on what your mystery-shop revealed. Make that the first thing you can hand a prospect on call #1.

Build cost: ~6 hours. Vapi infrastructure exists already.

PIECE 2
SEO + Web Audit auto-generator

Already 70% built. You give it a URL; it crawls the site, runs Google Places, scrapes 3 nearest competitors, and produces a branded 7-page CaterZen growth audit PDF in 5 minutes. Just needs Caterzen branding and the new mystery-shop section bolted on.

Build cost: ~4 hours.

PIECE 3
Site mockup — one-click

You showed Avi the Claude Design flow on April 23. He saw it can one-shot a polished restaurant site in 5 minutes. We package that into a single internal command: enter the prospect's name + colors + photos, get back a deployed Vercel preview URL the prospect can click. This is the wow-factor moment.

Build cost: ~6 hours of templating.

PIECE 4
Pitch deck auto-fill

Five-slide template. Auto-pulls the prospect's name, mystery-shop scores, audit findings, mockup link, pricing tier. You add 2–3 sentences of personal context, send.

Build cost: ~4 hours.

Total build: ~20 hours over two weeks. When it ships, every new prospect goes from 15 hours of your time to ~3. The remaining 12 hours go into running more sales calls. Same you. Same close rate. ~4× the closed deals.

Three things sitting in reserve.

RESERVE
2,615-caterer outbound list

The prospect graph is loaded, scored, segmented. The day you decide outbound matters, we flip a switch on cold email or LinkedIn. Until then it sits, and that's fine. It's not blocking anything.

RESERVE
Restricted edit UI for end-customers ("Bubba BBQ panel")

The thing you specifically asked Avi about — a restaurant-owner-friendly admin where Bubba can request "change the heading to BBQ Catering Specials" without touching code. Real product feature. Adds value to your CaterZen platform. But it's a customer-success unlock, not a deal-closer. Build after the toolkit is shipping closes.

RESERVE
CaterZen lead dashboard (the one Avi half-built)

Your "all leads in one place — web forms, proposals, sales rep calls" dashboard. We can help finish it. But again — this is a CUSTOMER-facing feature for your platform. It doesn't close more CaterZen deals. Sequence it after the deal-closing toolkit.

Four decisions. Each takes a minute.

01
Confirm the bottleneck is real

Read this whole doc. If we got the diagnosis wrong — if you'd actually rather have us pour gas on the outbound side — tell us and we'll pivot. But the calls strongly suggest delivery is the wedge.

02
Which prospect do we mystery-shop first?

Pick a prospect you're already in conversation with. We use them as the proof case for the toolkit. The first end-to-end run is for them.

03
CaterZen branding assets

Logo (PNG/SVG), brand colors (hex), 2–3 hero photos you'd want on a generic site mockup. So the Claude Design one-shot gives back something CaterZen-branded, not generic.

04
Avi's involvement

Should Avi be the maintainer of this toolkit on your side, or do we own it as your service? Either works. Different costs. Different speed-to-iterate. Want a quick call to decide.

Here's what changes by May 30.

3 hrs
per new prospect (vs 15 today)
12–15
prospects pitched / month
3–5
target closes / month
~$15K
new MRR added per month
And the entire 2,615-caterer outbound list stays in reserve for the day you decide top-of-funnel volume is the next constraint. We'll be ready when that day comes. But it isn't today.