Best ROI Online Marketing: Ranked by Return Per Dollar

For most US service businesses, the highest-ROI online marketing isn't a channel — it's a sequence. Ranked by 12-month return per dollar: (1) responding to the leads you already get in under 5 minutes (near-zero media cost, and per the Oldroyd/InsideSales study published in *Harvard Business Review*, contact within 5 minutes makes qualification up to 21x more likely than at 30 minutes); (2) Google Business Profile and local SEO (typically 4–10x, but 60–120 days to move); (3) email and SMS to your existing list (the widely cited DMA figure is $36 back per $1); (4) Google Search ads on high-intent keywords (2–5x, revenue in week one); (5) paid social (1–3x); (6) display and programmatic (usually below 1x for local service). Here's the math, the payback windows, and when none of it is worth buying.

The ranking, with the arithmetic shown

Channel ROI arguments are usually vibes. Run the numbers instead. A model for a residential HVAC company with a $1,400 average ticket and a 25% close rate on qualified leads:

The pattern: the cheapest ROI gains are almost never in buying more traffic. They're in the conversion math applied to traffic you're already paying for. Our ROI calculator runs these same inputs against your actual ticket size and close rate.

The number your agency quoted you is probably the wrong number

This is the part most "best ROI channel" articles skip, and it's the single most expensive misunderstanding in local marketing.

Blended ROAS is an average. Spending decisions are made at the margin. If you spend $10,000/month on Google Ads and get $40,000 back, that's 4x blended. It does not mean the next $5,000 returns $20,000. Search demand for "emergency plumber near me" in your ZIP code is a finite pool. The first $3,000 buys your highest-intent, lowest-CPC clicks. The next $3,000 buys broader match types, worse hours, and colder keywords.

A realistic marginal curve on a $10,000 local search budget looks like: first $2,500 at 7x, second $2,500 at 4x, third $2,500 at 2.5x, fourth $2,500 at 1.1x. Blended: 3.6x. Marginal on that last tranche: barely break-even before labor cost.

If your agency reports a single blended ROAS number and never shows you the return on your last $1,000, they are structurally incapable of telling you when to stop spending. That is not a reporting gap. That's the entire decision.

The practical test: cut your worst-performing 20% of spend for 30 days. If total revenue drops less than 20%, that spend was subsidizing the average. Most local accounts we audit have 15–30% of spend sitting below a 1.5x marginal return, usually in broad match, Display expansion, and out-of-service-area geo targeting.

Second trap: ROI is a ratio, and you can maximize it by shrinking. A business spending $500/month at 12x ROI ($6,000 back) is doing worse in dollars than one spending $8,000 at 3.5x ($28,000 back). Optimize for total gross profit contributed, not the prettiest multiple.

The three inputs that outrank channel choice

Channel selection explains maybe 30% of outcome variance. These explain most of the rest:

What each channel realistically costs in 2026

Full retainer ranges are on our pricing page, and case studies show the before/after inputs on real accounts.

When this is a bad investment and you should not buy

Plainly, because it costs us deals to say it:

Failure modes that kill ROI in the right channel

The sequence that actually maximizes return

Weeks 1–2: fix answering coverage and speed-to-lead, install call tracking, claim and fully populate Google Business Profile. Weeks 3–6: email and text your existing customer list; launch Local Services Ads if you're licensed and eligible. Weeks 7–12: tightly geo-fenced Google Search ads on high-intent terms only, with a hard marginal-ROAS floor. Month 4 onward: local SEO and content, funded by the paid channels already returning cash.

That order exists for one reason — each step raises the conversion rate that every later step's ROI depends on. Run it backwards and you'll pay premium CPCs to fill a leaky bucket.

If you want the honest version of these numbers against your own close rate, ticket size, and capacity, the free marketing audit will tell you which of the six channels above you should be in — including, sometimes, none of them yet.