CPM vs CPC vs CPA vs CPVD: Which Ad Pricing Model Is Best for Local Businesses?
Tunnel = a 1-mile stretch of road you own. Zone = a neighborhood-sized area of about 0.28 square miles that you own. Background = the whole member network, no geography limit. Full glossary
The four pricing models, defined
Every digital ad buy resolves to one of four pricing triggers. The trigger decides who carries the risk, what gets measured, and where money leaks.
CPM (Cost Per Mille). You pay per 1,000 impressions served. The advertiser carries impression-quality risk: viewability, fraud, off-target audience.
CPC (Cost Per Click). You pay only when someone clicks. The platform carries the impression risk; you carry click-quality risk and downstream conversion risk.
CPA (Cost Per Acquisition). You pay only when a defined conversion fires (form, call, sale). The platform or affiliate carries everything upstream; you carry attribution-fidelity risk.
CPVD (Cost Per Verified Delivery). You pay a fixed rate, set by your monthly deposit tier rather than by an auction, for each GPS-verified delivery to an opted-in phone moving through a tunnel you own. A delivery is not a lead and not an impression. See What is CPVD? for the full definition.
Head-to-head: pricing model comparison
Each model optimizes for a different stage of the funnel. The cheap-per-unit models (CPM) carry the most hidden cost. The expensive-per-unit models (CPA, CPVD) carry the most signal density.
How CPM, CPC, CPA, and CPVD compare on the metrics local advertisers actually care about.
Metric
CPM
CPC
CPA
CPVD
Pricing trigger
Per 1,000 impressions
Per click
Per defined conversion
Per GPS-verified delivery
What advertiser pays for
An ad served (maybe seen)
A click (maybe a human)
A conversion event (maybe attributed correctly)
Delivery to a real person's phone while they're driving in a tunnel you own
Fraud exposure (measured, not feared)
Small in measured terms: 1.0% fraud and SIVT in North America (DoubleVerify 2025), 0.3% invalid traffic (ANA Q3 2025). The larger loss is unmeasurable and non-viewable inventory, not bots
Same measured base, plus click fraud and accidental taps
Depends on attribution fidelity rather than traffic quality
No exchange surface: delivery is confirmed at the device
Attribution clarity
Weakest: impressions do not equal outcomes
Medium: click is a real action
Strong on paper, fragile in practice (multi-touch decay)
Strong: tunnel + device + time are observable
Scale / reach
Massive (open exchanges, MFA inventory)
Large but capped by query volume
Capped by conversion volume
Bounded by tunnel inventory and traffic flow
Published cost
No source at our sourcing bar publishes a current cross-network CPM, so we quote none
$5.42 all industries, $8.33 Home and Home Improvement (WordStream 2026); $7.85 home services, $9.68 HVAC (LocalIQ, Apr 2024 to Mar 2025)
No qualifying source publishes a current cost per conversion. On a cost-per-lead basis: $66.69 all industries, $90.92 Home and Home Improvement (WordStream 2026)
From $0.25 per background delivery at Enterprise; $0.32 Pro, $0.40 Local, $0.50 Starter. Fixed, no auction. Tunnels and zones cost more
Best fit
National brand awareness at CPG-style budgets
High-intent keyword search (emergency, branded)
Mature funnels with clean conversion tracking
Local service operators on a fixed tunnel they own
Where each model leaks money
The cheap-per-unit advantage of CPM dissolves once you account for the supply chain and for inventory nobody can measure. The ANA publishes a quarterly benchmark built from impression-level logs matched against sell-side logs by TAG TrustNet and Fiducia. In its Q3 2025 findings, on web and mobile excluding connected TV, transaction costs took 26.9 percent of the ad dollar, media-productivity loss took 26.1 percent, and TrueAdSpend, the working media, was 47.1 percent. These are medians across 21 actively contributing marketers and $142 million of spend, and they exclude agency, ad-serving and brand-safety fees, so the waterfall is not all-in.
The composition matters more than the headline. Inside that 26.1 percent productivity loss, non-measurable impressions are 15.2 percent and non-viewable 10.1 percent, against made-for-advertising sites at 0.5 percent and invalid traffic at 0.3 percent. The money is not mostly going to bots. It is going to inventory nobody could measure or nobody could see. DoubleVerify's 2025 measurement agrees on the order of magnitude: fraud and sophisticated invalid traffic at 1.0 percent in North America, 0.8 percent in the US.
For the structure of the waterfall, the ANA's December 2023 study is still the only fully itemized version it has published: on $1,000 of open-web spend, $130 SSP platform, $80 DSP platform, $60 DSP data and $20 DSP additional, then $150 non-measurable, $100 made-for-advertising, $95 non-viewable and $5 invalid traffic, leaving $360. That 36-cent figure is 2023 structure, never today's level. And be careful with the cohort number: the ANA's Q4 2025 findings split participants at the median of TrueAdSpend, with the higher-performing half at 56.7 percent and the lower-performing half at 37.5 percent. That is a median split on the outcome, not a split between advertisers who do and do not run a transparency program.
CPC mostly sidesteps impression quality but inherits click fraud and accidental taps. CPA narrows that further but introduces attribution decay across multi-touch journeys. CPVD removes the auction entirely: you pay a fixed rate per delivery, from $0.25 at the Enterprise tier, and the delivery is GPS-verified by infrastructure WilDi controls. No DSP fee, no SSP rake, no unmeasurable inventory, no bid-stream proxy. What it is not is a cheaper impression; it is a different unit. See What is the Middleman Tax? for the full breakdown.
Working media on web and mobile
47.1%
TrueAdSpend, Q3 2025, web and mobile excluding connected TV. Last quarter the ANA itemized that cut; current all-environments headline 45.1% (Q2 2026)
What the math looks like for a local service operator
LocalIQ's home-services search benchmark, covering April 2024 to March 2025 across 3,211 US campaigns, puts the median home-services cost per click at $7.85, with HVAC at $9.68, plumbing at $10.49, roofing at $10.70 and electricians at $12.18. Cost per lead in the same dataset runs $90.92 for home services overall, $127.74 for HVAC, $129.02 for plumbing and $228.15 for roofing and gutters. WordStream's newer 2026 edition does not break out those trades; it puts Home and Home Improvement as one row at $8.33 per click and $90.92 per lead, against $5.42 and $66.69 across all industries.
On the CPM side there is nothing honest to quote. No source at our sourcing bar publishes a current cross-network display or social CPM, so we do not print one. The structural point survives without a number: a local HVAC operator buying open-exchange impressions is paying for a population that is overwhelmingly outside their service area, not in-market for HVAC, or never going to convert, and the ANA's Q3 2025 waterfall says 15.2 percent of the spend goes to impressions nobody could even measure.
CPVD is priced as fixed ownership of a tunnel, from $0.25 per background delivery at the Enterprise tier. The rate follows the monthly deposit: $500 a month is the Local tier at $0.40, so $500 buys 1,250 verified deliveries, not 2,000. At $1,000 a month the Pro rate of $0.32 buys 3,125; at $3,000 the Enterprise rate of $0.25 buys 12,000. Tunnels and zones cost more than background. No auction, no fee stack, no unmeasurable inventory. Two honest trade-offs: scale is bounded by tunnel traffic rather than exchange inventory, and a delivery is not a lead and not an impression, so you cannot lay it on a CPM chart.
Which model to use, by scenario
No single model wins every scenario. The honest call is to match the model to what you're actually buying.
National brand awareness, large budget: CPM. The fraud and waste are real, but reach economics still favor CPM when the goal is mass exposure and you have CPG-style measurement to absorb the noise.
High-intent keyword search: CPC. When someone types "AC repair near me" at 11pm, paying LocalIQ's $9.68 median HVAC click is rational. The intent is unambiguous.
Mature funnels with clean tracking: CPA. If you can attribute conversions reliably and the platform offers CPA pricing on inventory you trust, this is the cleanest economic alignment.
Local service operators wanting predictable spend: CPVD. Fixed price, GPS-verified delivery, no auction. See HVAC advertising in Jacksonville for the operator math.
Performance-only stack: CPC + CPA in combination. Use CPC to test creatives and CPA-trigger campaigns to scale winners.
Avoid: CPM-only campaigns on open exchanges without third-party verification. Not because of bots, which measure at about one percent, but because the ANA's Q3 2025 waterfall puts 15.2 percent of spend into impressions nobody can measure and 10.1 percent into impressions nobody could see, on top of 26.9 percent in fees.
Why most platforms charge CPM (even when they say CPC or CPA)
Buried in the platform settings, almost every major ad network (Meta, Google Display, programmatic DSPs) runs auctions on an effective-CPM basis under the hood. CPC and CPA are wrappers: the platform predicts your CTR or conversion rate, multiplies by your bid, and pays publishers on a CPM basis.
That's why pure-CPC campaigns on cheap inventory often deliver lousy clicks: the platform is optimizing the eCPM math, not your customer acquisition. The more sophisticated the targeting layer, the more aggressively the platform will route your budget to wherever its algorithm predicts the highest eCPM: frequently MFA-style inventory, low-quality publishers, or out-of-market audiences.
CPVD is a different architecture. WilDi doesn't run an auction. The tunnel is yours, the price is fixed, and the delivery is GPS-verified at the device. There is no eCPM optimization happening underneath, because there is no exchange.
The product
Three ways to deliver: tunnels, zones, background
WilDi Maps is not a single flat-rate product. You pick the tier that matches how local you need to be. All three are GPS-verified per claim, with no auction, no exchange rake, no Middleman Tax.
Tunnel
1-mile road strip
Premium
Hyper-local, just-in-time
Claim a one-mile stretch. When a member enters the strip, they get a just-in-time message, perfect for emergency services, on-route specials, and anything where being right there now beats brand awareness later.
Best for
· HVAC, plumbing, water restoration
· On-route specials (food, fuel, retail)
· Garage door, locksmith, urgent service
Zone
0.28-square-mile area
Premium
Hyper-local, area-based
Claim a one-square-mile block, not tied to a single road. Catches the residential cluster, retail district, or industrial park where your work actually lives. Same just-in-time delivery as tunnels; different geometry.
Best for
· Lawn care, pest control, pool services
· Tree services, landscaping
· Neighborhood-targeted retail
Background
City-wide rotation
From $0.25
per claim, tier-based
City-wide brand presence on rotation. Highest reach for the budget; best when familiarity beats precision. Per-delivery rate drops by tier (Enterprise: $0.25 / Pro: $0.32 / Local: $0.40 / Starter: $0.50). See /pricing for the live rate card.
Best for
· Restaurant brands, retail specials
· Veteran-owned trust signals
· Cross-vertical brand awareness
What the member gets when an ad is claimed
Direct-drive turn-by-turn
If the member wants to act on the ad, the app navigates them straight to the advertiser's location.
Website link
Click-through to any URL: ordering page, brand site, blog post, lead form.
App page
Open a specific page inside the WilDi app: promo details, daily specials, claim instructions.
See the full pricing breakdown on the pricing page.
Frequently asked questions
What is CPM in advertising?
CPM (Cost Per Mille) is a pricing model where advertisers pay per 1,000 ad impressions. No source at our sourcing bar publishes a current cross-network display or social CPM, so we do not quote one; the figures that circulate come from agency roundups. What is published is what happens to the money: the ANA's Q3 2025 benchmark, on web and mobile excluding CTV, puts 26.9 percent into transaction costs and 26.1 percent into productivity loss, of which 15.2 points is impressions nobody could measure and 10.1 is impressions nobody could see. CPM is cheap per unit and carries the most audience-quality risk, because you pay regardless of whether anyone in your target market saw or engaged with the ad.
Is CPC better than CPM for local businesses?
For most local service businesses, yes. CPC is generally better than CPM because you only pay when someone clicks, which filters out unmeasurable and non-viewable impressions along with irrelevant audiences. The trade-off is higher per-unit cost: LocalIQ's home-services benchmark for April 2024 to March 2025 puts the median home-services cost per click at $7.85, HVAC at $9.68, plumbing at $10.49 and electricians at $12.18. WordStream's 2026 edition puts Home and Home Improvement at $8.33 and all industries at $5.42. CPC works best on high-intent search queries; CPC on display inventory can still leak budget to low-quality clicks.
What is CPA, and when does it make sense?
CPA (Cost Per Acquisition) is a pricing model where advertisers pay only when a defined conversion fires: a form fill, phone call, or sale. No source at our sourcing bar publishes a current cross-industry cost per conversion, so we do not quote one. The closest published figure is cost per lead: WordStream's 2026 benchmark puts all industries at $66.69 and Home and Home Improvement at $90.92, and LocalIQ's per-trade medians run from $93.69 for electricians to $228.15 for roofing and gutters. CPA makes sense when your conversion tracking is reliable and the platform offers CPA pricing on inventory you trust. It fails when multi-touch attribution is fragile or when affiliates can game the conversion event.
Why is CPVD different from CPM, CPC, and CPA?
CPVD (Cost Per Verified Delivery) replaces the auction-based programmatic model with a tunnel you own: from $0.25 per GPS-verified delivery on background at the Enterprise tier ($0.32 Pro, $0.40 Local, $0.50 Starter; tunnels and zones cost more), to one opted-in person's phone as they drive through a route you chose. There is no exchange, no DSP or SSP fee stack, no made-for-advertising inventory, and no bid-stream proxy guessing whether a device was nearby. The delivery is verified at the device by infrastructure WilDi controls, not inferred from a third-party SDK. It is a different unit from an impression or a lead, not a cheaper version of either.
Which pricing model is best for HVAC contractors?
HVAC operators typically run a hybrid: CPC on Google Search for high-intent emergency queries (LocalIQ's median HVAC click is $9.68 and its median HVAC lead $127.74, which is rational when the buyer is in distress and a full system install averages $7,500 in HomeAdvisor's cost guides), and CPVD for steady-state coverage of the service area, from $0.25 per verified delivery at Enterprise and $0.40 at the Local tier a $250-a-month deposit qualifies for. CPM rarely pencils for HVAC because the share of any open-exchange impression that is a homeowner with a failing system in your DMA is too small. CPA can work if your call tracking is mature.
Why do most platforms run on CPM under the hood?
Even when ad platforms expose CPC or CPA pricing to the advertiser, the auction underneath almost always settles on an effective-CPM basis: the platform predicts your CTR or conversion rate, multiplies by your bid, and pays publishers per impression. That's why low-quality inventory still gets fed budget: the algorithm is optimizing eCPM, not your customer acquisition. CPVD avoids this entirely because there is no auction to optimize.
How much of a programmatic ad dollar becomes a benchmark-qualified impression?
45.1 cents on the ANA benchmark's current basis, which covers all environments including connected TV (Q2 2026, reported by the ANA's measurement partner Fiducia), with transaction costs at 27.2 percent. The last quarter the ANA itemized web and mobile excluding connected TV was Q3 2025, and on that cut the figure was 47.1 cents: transaction costs, chiefly SSP and DSP fees, took 26.9 percent, and media-productivity loss took a further 26.1 percent, of which 15.2 points is impressions nobody could measure and 10.1 is impressions nobody could see. The two figures sit on different bases and do not belong on one trend line. Read the number for what it measures: TrueAdSpend is spend that produced a benchmark-qualified impression, meaning measurable, viewable, not invalid traffic and not on a made-for-advertising site. Whether a person then looked at it is a separate question the benchmark does not answer. These are medians across participating marketers and they exclude agency, ad-serving and brand-safety fees. Two caveats people get wrong. First, the older 36-cent figure is from the ANA's December 2023 study and describes the structure of the waterfall in that period; it is not today's level. Second, the ANA's Q4 2025 split of 56.7 percent versus 37.5 percent is a median split on TrueAdSpend itself, dividing participants into a higher-performing and a lower-performing half. It is not a comparison of advertisers with and without a transparency program.
How much of a CPM campaign is actually fraud?
Much less than the industry's own marketing suggests, which is why this page does not lead with bots. DoubleVerify measured fraud and sophisticated invalid traffic at 1.0 percent in North America in its 2025 Global Insights report, 0.8 percent in the US. The ANA measured invalid traffic at 0.3 percent of benchmarked spend in Q3 2025. The far larger losses in the same waterfall are the 15.2 percent of spend on non-measurable impressions and the 10.1 percent on non-viewable ones, plus 26.9 percent in fees. CPC and CPA reduce impression-quality exposure; CPVD avoids the open exchange entirely, which removes the surface rather than shrinking it.
What exactly counts as a verified delivery?
One verified delivery of your offer to the phone of one real person who was physically inside your chosen geography at that moment, confirmed by GPS on the device itself. That person also taps to acknowledge the offer, so a delivery is never an invisible impression. Bots, background tabs, and off-screen impressions cannot generate one. You are billed only when a verified delivery happens.
How much does it cost to start advertising on WilDi Maps?
The Starter tier opens with a $50 deposit, and that deposit becomes your ad budget. Background deliveries on Starter run $0.50 per verified delivery, so the first deposit buys 100 GPS-verified deliveries to the phones of real local people out on the road. There is no auction and no platform fee stacked on top.
What are the WilDi Maps plan tiers?
Four public tiers: Starter ($50 minimum deposit, background only), Local ($250, up to 2 tunnels and 1 zone), Pro ($1,000, up to 8 tunnels and 5 zones), and Enterprise ($3,000, up to 25 tunnels and 15 zones). Per-delivery background rates step down by tier, from $0.50 on Starter to $0.25 on Enterprise. An Agency tier is available through sales.
What happens when someone taps my ad?
They are routed wherever you choose: turn-by-turn directions to your door, your website, or your app page. Every claim is logged with time and location, so you can attribute real walk-ins instead of guessing from impression counts.
About this analysis
Written by Timm Ross, founder of WilDi Maps · Jacksonville-based · Veteran-owned. Sources are cited inline; we update the numbers when the underlying research updates.