Cost Per Acquisition by Funnel Stage: How Top-of-Funnel vs Bottom-of-Funnel Ad Spend Actually Compares for Local Businesses

The Single-CPA Trap

Here's a scenario that plays out constantly: a local HVAC company spends $4,000/month on ads across Google Search, YouTube pre-rolls, and Facebook awareness campaigns. At month-end, they divide total spend by total new customers and get a CPA of $267. That number feels manageable, so the budget rolls over unchanged.

The problem? That $267 is a blended fiction. It averages together spend that's closing customers today with spend that won't convert anyone for 60–90 days—if ever. Worse, it makes it impossible to see which layer of the funnel is efficient and which is quietly incinerating cash.

Collapsing all spend into one CPA number is the most common budget mistake we see in local ad accounts. This framework shows you how to break it apart.

The Three-Layer Funnel CAC Framework

Think of your funnel in three distinct cost centers, each with a different job, a different conversion timeline, and a meaningfully different cost-per-acquisition:

Layer 1 — Awareness (Top of Funnel / TOF) Channels: YouTube, Facebook/Instagram reach campaigns, Display. Job: Get in front of people who don't know you exist yet. Conversion timeline: 30–120 days, sometimes never directly. Right metric: Cost per 1,000 impressions (CPM) or cost per new audience member reached — not CPA, because asking TOF spend to produce direct acquisitions is the wrong question entirely.

Layer 2 — Consideration (Middle of Funnel / MOF) Channels: Facebook retargeting, Google Display remarketing, YouTube retargeting, local service category keywords. Job: Move warm prospects toward a decision — site visits, phone calls, form fills. Conversion timeline: 7–30 days. Right metric: Cost per qualified lead (CPL) with lead quality scoring. This is where CPL benchmarks become meaningful.

Layer 3 — Intent (Bottom of Funnel / BOF) Channels: Google Search (branded + high-intent keywords), Google Local Services Ads, direct retargeting to cart/quote-page visitors. Job: Close people who are actively shopping right now. Conversion timeline: 0–7 days. Right metric: Cost per acquisition (CPA) and, critically, revenue per acquisition (ROAS). For a deeper look at how keyword strategy affects this layer, see our teardown DKI vs Static Google Ads: Which Cuts CPA for Local Services?

A Labeled Attribution Model: The $4,000 Budget Teardown

Let's rebuild that HVAC example properly. Same $4,000 budget, same month, same 15 new customers — but now we split the spend by funnel layer.

(All figures below are illustrative models, not measured benchmarks.)

| Funnel Layer | Spend | Attributed Customers | Layer CPA | |---|---|---|---| | TOF (Facebook reach + YouTube) | $1,200 | 1–2 (long-lag) | $600–$1,200 | | MOF (retargeting + category keywords) | $1,000 | 3–4 | $250–$333 | | BOF (branded search + high-intent) | $1,800 | 9–11 | $164–$200 |

Blended CPA: ~$267. Looks fine.

BOF CPA: ~$164–$200. Efficient — likely worth scaling.

TOF CPA (direct): $600–$1,200. Alarming if you're judging it by direct CPA — but potentially justified if it's feeding the MOF/BOF pipeline.

The question the blended number can never answer: Is the TOF spend building a pipeline that makes BOF more efficient over time, or is it pure waste? You cannot know without separating the layers first.

Why TOF Spend Has a Different 'Cost' Than It Appears

Top-of-funnel spend should never be judged on direct CPA. Its job is pipeline-loading — it's inventory, not immediate revenue. The right way to evaluate it:

• Assisted conversion rate: What % of your BOF converters touched a TOF ad in the prior 30–90 days? (Check this in Google Analytics attribution reports or Meta's attribution window settings.)
• New audience growth rate: Are you reaching net-new, in-market locals — or re-serving the same people who've already decided against you?
• CPM efficiency: A rough rule of thumb for local campaigns — if your TOF CPM is climbing but your BOF volume is flat, the TOF investment is not loading the pipeline.

If TOF spend has low assisted conversion rates and BOF volume isn't growing, that's the clearest signal to reallocate to MOF/BOF. For a related audit methodology, see our post Google Ads Audience Observation: Budget Waste Audit.

The MOF Problem Most Local Businesses Ignore

Middle-of-funnel is the most neglected layer in local ad accounts — and it's where the biggest efficiency gains typically hide.

Here's the common failure pattern: a local business runs TOF awareness ads that generate site visitors, then sends those warm visitors straight back into the general ad pool with no dedicated retargeting sequence. The warm prospect sees the same awareness ad again, goes cold, and eventually converts through a competitor's BOF search ad.

A basic MOF retargeting setup for local businesses:

• Audience: Website visitors (7–30 day window) who did NOT convert
• Creative: Specific offer, social proof (reviews, before/after), or objection-handling content
• Bid strategy: Lean toward Maximize Conversions within a controlled budget cap — you want aggression here because these people already know you
• Timing: This is also where ad scheduling earns its money. See Meta Ad Scheduling vs Always-On: Which Lowers CPL? for how scheduling changes CPL at this layer specifically.

In our experience, local businesses that build even a simple MOF retargeting layer typically see BOF CPA improve — because the BOF closer doesn't have to do as much persuasion work.

How to Calculate Your Own Funnel-Stage CPA

You don't need enterprise attribution software. You need discipline in how you tag and report. Here's a simple three-step process:

Step 1: Separate campaigns by funnel intent, not by channel. Don't lump 'Facebook' into one bucket. Separate 'Facebook — awareness reach' from 'Facebook — retargeting warm visitors.' Same rule applies in Google Ads.

Step 2: Assign conversion windows by layer.

• TOF: 60–90 day view-through window (or longer)
• MOF: 14–30 day click window
• BOF: 7-day click window

Any platform's default 'last-click' attribution will over-credit BOF and starve TOF of any measured value. Adjust attribution models deliberately.

Step 3: Build a simple monthly layer report. For each layer: total spend, leads generated (with quality flag if possible), attributed customers (using your assigned window), layer CPA, and — at BOF — average revenue per acquired customer so you can calculate ROAS, not just CPA.

A local service business with an average job value of $800 (illustrative) and a BOF CPA of $180 (illustrative) is running at a ~4.4x ROAS on that layer. That's the number worth optimizing — not the blended $267.

The Takeaway: One Number Is Always a Lie

A single blended CPA tells you nothing about where your budget is working. It smooths over inefficiency, masks scaling opportunities at BOF, and gives you false confidence about TOF spend that may or may not be earning its place.

The fix isn't a bigger budget. It's better segmentation and the discipline to judge each funnel layer by the right metric for its job.

To summarize the framework:

• TOF: Judge by pipeline contribution and CPM efficiency, not direct CPA
• MOF: Judge by CPL with lead quality scoring — this is where scheduling and audience precision matter most
• BOF: Judge by CPA and ROAS — revenue per customer, not just cost per lead

If you want a second set of eyes on how your current ad budget is split across funnel layers — and whether your blended CPA is hiding a BOF opportunity or a TOF leak — book a call with the Nika Spark team. We'll show you exactly where the number is lying.