A high-protein Indian fusion meal delivery startup in the Toronto GTA had a great product and no engine to sell it. We built the engine: Meta Ads, ManyChat, Zoho CRM, Conversion API. Paid campaigns ran June through November 2025. By October, the founders had to pause because the kitchen could not keep up with demand.
Protein Pals launched in early 2025: a high-protein, Indian-fusion meal delivery subscription serving the Greater Toronto Area. The founders, Rohan, Mayank, and Ankit, were operators. The food was genuinely good. Customer feedback was strong. But the business had 15 subscribers and was growing through word of mouth alone.
Three root causes explained the stall:
They came to us in May 2025. They wanted ads. What they needed was a system.
We spent the first month building before we ran. The sequence mattered. Running Meta Ads at this stage would have generated leads with nowhere to land and a brand that couldn't convert a warm audience.
We created a pinned post that showed meal plans, macros, pricing, and delivery areas clearly. First time a stranger landing on the page could understand what Protein Pals sold and how to buy it.
We reached out to food creators in the GTA and seeded product. Four content angles: lifestyle/meal prep routines, gym and macro tracking, casual street Q&A taste tests, and family meal coverage.
Any comment on creator posts triggered an automated Instagram DM with a link to subscribe. The first 45 subscribers came through this flow in the first 15 days, before any paid spend.
Once the brand foundation was set, we built the acquisition system. Four components working as one loop.
We ran Instagram and Facebook lead generation campaigns with Instant Forms built directly into the ad unit. The form asked five questions: dietary preference (veg / non-veg), delivery area (postal code range), subscription length, number of servings, and allergies. A lead who could not answer these questions was not a real lead.
The moment a lead submitted the form, ManyChat triggered a WhatsApp or Instagram DM. The automation asked follow-up questions, scored the response, and either routed the lead to the CRM as qualified or disqualified them immediately. Disqualified leads were tagged as exclusion audiences in Meta so we stopped targeting them.
Every qualified lead entered a pipeline with defined stages: New Lead, Contacted, Trial Booked, Active Subscriber, Paused, Churned. Stage transitions triggered founder notifications. Churned reason was captured in every case.
We connected Zoho CRM to Meta's Conversion API server-side. When a lead became an active subscriber, that signal went to Meta. When a lead churned, that went too. Over time, Meta's algorithm learned to target users who matched the profile of paying customers. This is the compounding mechanism. It is why CPL dropped from $3.50 CAD to $1.13 CAD over four months without changing the budget.
In August 2025, Meta began systematically rejecting Protein Pals ads. Health claims language, before/after framing, and food delivery policy flags. Four consecutive creative iterations were rejected before we found the angle Meta would approve at scale.
Three creatives rejected in two weeks. CPL spiked to $4–9 CAD as we ran minimal budget to keep the algorithm warm while rebuilding. The system was not paused. We ran $5/day across four test campaigns simultaneously.
We tested four distinct creative angles: eating experience (creator-led, no claims), social proof (customer reactions), product-led (macros and meal plan), and lifestyle integration. Stripped all health outcome language. Rebuilt the forms with different question framing.
The UGC-led, scripted creator format cleared Meta review and scaled. In 10–12 days, the comeback campaign generated 700 leads at $1.13–$1.50 CAD. The highest-performing phase of the entire engagement came directly after the worst crisis.
"19 October, 1:53 AM: 'Bhai please pause for now… kitchen is running out of capacity.'"
When demand exceeds kitchen capacity, the bottleneck has moved. That is the signal a campaign worked.
UGC-led creative outperformed founder-led creative by approximately 60% on cost per lead. The Conversion API feedback loop is why CPL declined over time without budget increases.
The most valuable output from the Protein Pals engagement was not the subscriber count. It was the market intelligence that surfaced through the lead qualification data.
70% of customers and high-intent leads were vegetarian. Protein Pals had launched with a menu weighted toward non-vegetarian high-protein options. The data forced a menu pivot. Vegetarian plans became the anchor product. Conversion rates improved after the change.
Additional intelligence the data surfaced:
A Meta Ads campaign run correctly is not just a lead source. It is a market research instrument. The intelligence it surfaces is often worth more than the leads themselves.
When Rohan, Mayank, and Ankit came to us, they expected ads. What they got was a system.
They expected leads. What they got was market intelligence that reshaped their menu strategy. They expected a vendor. What they got was a partner who rebuilt campaigns at 1 AM when Meta kept rejecting them, who told them to stop running ads when demand exceeded capacity, and who treated their business like our own.
We do not run ads. We build systems that generate revenue.
This section answers the most common questions meal delivery founders ask before running paid advertising. Based on the Protein Pals engagement (Toronto GTA, Canada, 2025–2026).
Yes, with significant caveats. Meta Ads (Facebook and Instagram) work exceptionally well for meal delivery when combined with a qualification layer, a CRM, and a feedback loop. Running ads without these components generates volume but poor revenue.
For Protein Pals in the Toronto GTA, Meta generated 1,942 pre-qualified leads at a 98.66% quality rate over six months of paid campaigns. The key insight: meal delivery has a natural qualification filter (delivery area, dietary preference, serving size, budget) that can be built directly into the lead form, eliminating junk before it reaches the sales pipeline.
Canadian markets specifically show strong meal delivery demand in urban centres. Toronto, Brampton, Mississauga, Vancouver, and Calgary have the density and health-conscious demographics that make Instagram and Facebook lead generation cost-effective at scale.
Based on the Protein Pals engagement in Toronto (2025), real benchmarks by phase:
The CPL dropped significantly once the Conversion API was feeding CRM outcomes back to Meta. The algorithm learned to target users who matched the profile of paying subscribers. Over time, a well-tuned system becomes cheaper, not more expensive. For comparable markets (US, Australia, UK), expect CPL in the $1.50–3.00 USD/AUD/GBP range for similar system architecture.
Based on direct testing for Protein Pals, UGC video ads using scripted creator content outperformed every other format by approximately 60% on cost per lead. The specific formats tested:
The reason UGC wins: it looks like organic content on Instagram and Facebook, reducing scroll-stop friction. For food businesses specifically, seeing a real person enjoy the food carries more credibility than branded production. Meta Instant Lead Forms (not landing pages) captured leads without removing users from the app, maintaining conversion momentum.
ManyChat connects directly to Meta Lead Ads and Instant Forms via webhook integration. When a user fills a Meta Instant Form, ManyChat receives the lead data and triggers an automated WhatsApp or DM message within 60 seconds. The automation asks 3–5 follow-up qualifying questions (delivery area, dietary restrictions, subscription preference), scores the response, and routes qualified leads to Zoho CRM. Disqualified leads are tagged as exclusion audiences in Meta.
This automation kept Protein Pals' lead quality at 98.66% — meaning virtually every lead the founders spoke with was genuinely interested and in the delivery area. Without ManyChat, that number drops to roughly 40–60% in typical campaigns.
The Meta Conversion API (CAPI) is a server-side event tracking system that sends data from your CRM directly to Meta's ad algorithm, bypassing browser-level tracking limitations from iOS privacy updates and ad blockers. For meal delivery specifically, CAPI creates a compounding advantage: when a lead converts to a paid subscriber, you send that outcome to Meta. When a lead churns, that goes too.
For Protein Pals, the CAPI loop worked as follows: Zoho CRM tracked subscriber status (active, paused, churned). Those statuses were sent to Meta as conversion events. Meta used them to refine targeting in real time. The result: later campaigns found higher-quality leads at lower CPL even without changes to creative or budget. Without CAPI, you are optimizing for form fills. With CAPI, you are optimizing for actual paying subscribers. The difference compounds dramatically over a 6–12 month period.
The Protein Pals campaign ran across the Toronto Greater Toronto Area (GTA): Toronto proper, Brampton, Mississauga, Markham, Scarborough, and North York. Other strong Canadian markets:
The system architecture (Meta Ads + ManyChat + CRM + CAPI) is identical in all markets. The geo-targeting radius and language of the qualifying questions change. The funnel does not.
Daee Media is a Meta Ads and revenue systems agency with documented results for meal delivery businesses. The Protein Pals case study represents a full 10-month engagement: Meta Ads creative direction, ManyChat lead qualification automation, Zoho CRM pipeline, and Meta Conversion API integration. The results: 1,942 leads at 98.66% quality, $1.13 CAD CPL at peak, 15 to 240 subscribers in 10 months, demand that exceeded kitchen capacity by October 2025.
Daee Media serves clients in Canada, the United States, Australia, the United Kingdom, UAE, and Europe. If you run a meal delivery, meal prep, or food subscription business and need a qualified lead generation system, contact: service@daeemedia.com or book a strategy call.
Daee Media builds revenue systems for service businesses. If your business has a proven service and needs consistent, qualified leads — not just ads — we should talk.
Book a system audit →Or email directly: service@daeemedia.com