What To Do If You Only Have $100 For Marketing (Facebook Ad Budget)

[Operator’s read] This is from inside an actual P&L I’m responsible for — not theory. If something doesn’t earn its keep on a live revenue line, it’s not in here.

Table of contents

Open Table of contents

• The post-iOS 14 reality every small advertiser needs to accept
• What Advantage+ campaigns mean for small budgets
• The Conversions API (CAPI): non-optional in 2026
• A real local-business framework: the Pilates studio example (updated)
  • Budget allocation across $100
  • Targeting setup
  • The ad itself
  • What to expect
• What changed vs. the original post
• Facebook Ad Budget — 2026 FAQ
  • Is $100 still worth running as a Facebook test in 2026?
  • Should I still use detailed interest targeting or switch to broad?
  • What is CAPI and do I really need it for a tiny budget?
  • How long should I run a $100 campaign before evaluating it?
• The shorter version
• Updated for May 2026

The post-iOS 14 reality every small advertiser needs to accept

Apple’s App Tracking Transparency (ATT) prompt, rolled out in 2021, broke most pixel-based retargeting and interest-targeting signal. Meta’s reported conversions dropped sharply for advertisers who relied on third-party cookies and detailed behavioral data.

What replaced it: creative-led, broad targeting backed by the Meta Advantage+ suite. Instead of telling Meta exactly which 43-year-old Pilates enthusiast within seven miles to show your ad to, you now tell Meta your general goal and radius, write a strong creative, and let the ML figure out who converts. That sounds terrifying on a $100 budget — but it actually works because Meta has years of on-platform behavior to infer intent from, and the auction rewards relevance.

The short version: stop obsessing over the audience panel; obsess over the ad itself.

---

What Advantage+ campaigns mean for small budgets

Meta now routes most new campaign creation through Advantage+ campaigns (formerly “Advantage+ Shopping” for e-commerce; expanded to lead gen and local awareness). The key behaviors to know:

• Advantage+ audience: you can set an age range and location, add an optional “audience suggestion” (interests, lookalikes), but Meta is free to go beyond it if it sees conversion signals elsewhere. On a $100 budget the algorithm needs every signal it can get.
• Advantage+ creative: Meta will automatically test headline, text, and creative variations. Give it at least 2–3 images or short videos to pick from. Do not rely on a single static image.
• Advantage+ placements: leave this on. Manually restricting to Feed only on a micro-budget starves the auction.

You are not bypassing the audience — you are giving Meta guardrails (city + radius + reasonable age band) and strong creative, then letting the system optimize.

---

The Conversions API (CAPI): non-optional in 2026

If you are running leads to a landing page or collecting contact info, you need Meta’s Conversions API set up server-side before you spend a dollar. Without it, iOS-privacy limits mean Meta may see a fraction of your actual leads, which trains the algorithm on incomplete signal.

Most CRM and website platforms (Shopify, WordPress with a plugin, GoHighLevel, etc.) have built-in CAPI connectors. It takes an afternoon to set up. It is not optional anymore — it is table stakes. With CAPI in place, Meta sees the lead even when the browser-side pixel fires blind.

---

A real local-business framework: the Pilates studio example (updated)

The original post used a local Pilates studio in Scottsdale as a case study. The core offer logic still holds perfectly: free trial classes in exchange for contact info, targeting local residents. Here’s how I’d run the same campaign in 2026.

Budget allocation across $100

This is qualitative — exact costs-per-lead vary heavily by market, creative quality, offer strength, and season. Use this as a mental model, not a spreadsheet:

• ~$20 learning phase: your first few days will have higher CPL while the algorithm collects signal. Expect to overpay per lead initially. This is normal. Do not pause the campaign in the first 48 hours.
• ~$60–70 actual spend: once the algorithm stabilizes (usually after a handful of leads), CPL should drop. This is where you get most of your leads.
• ~$10–20 holdback for a creative test: if your first ad gets leads but you’re curious whether a video hook would outperform the static image, drop a second creative mid-flight. On a small budget, keep it to one variable.

Targeting setup

• Location: city + radius around the business (use city-level or a tight radius, not county/DMA)
• Age: set a realistic band based on your customer profile
• Advantage+ audience suggestion: add “Fitness” and “Pilates” as suggestions, but leave Advantage+ audience expansion on
• CAPI: connected before the campaign goes live
• Objective: Leads (not Traffic, not Awareness)

The ad itself

This is where the work is. Three things that still hold from Anthony’s original framework:

1. Open by naming your audience in the first line. “Hey Scottsdale Pilates lovers” still works. It creates a scroll-stop because it’s specific.
2. Make the offer impossible to ignore. Three free classes is a strong offer. Whatever your equivalent is, lead with it.
3. Reduce friction on the lead form. Use a native Facebook Instant Form (formerly Lead Ads). It auto-fills name and email. Every extra field you add cuts completion rate. Name + email is enough to follow up; phone is optional.

The creative above (from the original Pilates campaign) still illustrates the formula: specific audience callout, compelling free offer, low-friction action.

What to expect

At a well-run micro-campaign for a local service business, cost-per-lead should be in the single-to-low-double digits (in USD) if your offer is strong. On $100 that means somewhere between several leads and a couple dozen, depending on your market. The original case study got around 14 qualified leads — that’s a realistic outcome when everything is working.

Fourteen contacts who raised their hand for your service, at $100, is a real business outcome. Even if only a fraction convert, if your average customer value is meaningful, the math works.

---

What changed vs. the original post

The original piece was solid but reflected 2021-era Meta:

• Detailed interest targeting as the primary lever → replaced by Advantage+ audience + strong creative as the primary lever.
• No mention of iOS/ATT or signal loss → now the foundational context for any Meta campaign.
• No mention of CAPI → now non-optional.
• “Boost Post” as the cautionary tale → still true, but the alternative is now Advantage+ campaigns, not just “proper Ads Manager setup.”
• Dead external link to futuresharks.com → removed.

The instinct — a specific offer, a named audience, a low-friction lead form — is as valid as ever.

---

Facebook Ad Budget — 2026 FAQ

Is $100 still worth running as a Facebook test in 2026?

Yes, for a local service business with a strong offer and CAPI set up. It will not give you statistical certainty, but it will tell you whether your offer and creative resonate. Treat it as a learning run, not a growth channel.

Should I still use detailed interest targeting or switch to broad?

Broad with Advantage+ audience is now the Meta-recommended approach for most advertisers. Detailed interest stacks lost much of their precision after ATT. On a $100 budget, give Meta your location, a reasonable age band, and let the algorithm do the rest — your creative carries the targeting work.

What is CAPI and do I really need it for a tiny budget?

CAPI (Conversions API) sends event data from your server directly to Meta, bypassing the browser-level signal loss caused by iOS privacy prompts. Without it, Meta may be training on incomplete data. For a lead-gen campaign, set it up before you spend anything — most major CRMs and website platforms have native connectors now.

How long should I run a $100 campaign before evaluating it?

Let it run for at least five to seven days without pausing or making changes. The learning phase needs signal. If you kill the campaign after 48 hours because CPL looks high, you’re reading noise, not signal. Evaluate after the learning phase stabilizes.

Related reading:

• How to sell anything to anyone
• SEO tips to get free traffic
• Marketing and growth strategies

---

The shorter version

If you’re reading this because the workflow it describes is eating your week, that’s the kind of loop I build AI agents for. Two build slots open at a time.

Updated for May 2026

A few things have shifted since this post first went up. Meta dropped the legacy “Page” verification track in 2024 and folded it into Meta Verified ($14.99–$19.99/mo depending on tier and country) — the blue check is now a subscription, not a one-time review. Friend-request flows still work as described, though Meta moved the bulk-cancel UI deeper into mobile settings; the desktop m.facebook.com/friends/center/requests/outgoing route still works (2026-04 spot check).

Worth knowing in 2026: ~3.07B Facebook MAU (Meta Q4 2025 earnings), but the share of time-on-platform relative to Reels and WhatsApp has continued sliding. If this post is part of an outreach strategy, weight WhatsApp and Threads (yes — Threads survived the 2024 pivot speculation and crossed 200M MAU) accordingly.