THINXSTER
Blog/Meta Ads
Meta Ads9 min readJuly 5, 2026

How to Increase ROAS in Meta Ads: A Leverage-Ordered Playbook

A tactical, prioritized playbook for increasing Meta Ads ROAS — because the number is won as much after the click as before it.

RK
Ryan Korsz
Founder & CEO, Thinxster

TL;DR

A tactical, prioritized playbook for increasing Meta Ads ROAS — because the number is won as much after the click as before it.

→ See how this applies to your business (free 30-min call)

Most people trying to increase ROAS in Meta ads spend their entire week inside Ads Manager, tweaking bids and pausing audiences, when the biggest lever is sitting somewhere they never look: what happens in the ninety seconds after someone clicks. ROAS is a fraction. Revenue over spend. You can grind the denominator down forever, or you can attack the numerator where the real money hides. This playbook is ordered by leverage, not by convenience. Start at the top. Don't jump to the fun stuff before you fix the boring stuff, because the boring stuff is where the multiples live.

Fix the Offer Before You Touch a Single Setting

No amount of campaign structure saves a weak offer. If your ad says "Call today for a free quote" and so does every competitor in a fifteen-mile radius, you are asking the algorithm to sell something the market has already priced at zero. Optimization is multiplication, and you cannot multiply a bad offer into a good result.

The offer is the highest-leverage thing you control because it moves conversion rate at every downstream stage at once — click-through, landing page, and close. Rework it before you rework anything else.

  • Add a real reason to act now. "89-dollar diagnostic, this week only" beats "free estimate" because it is specific, it is scarce, and it lowers the risk of the first step.
  • Lead with the outcome, not the service. Homeowners don't want "drain cleaning." They want a working kitchen before guests arrive Saturday.
  • Name the objection in the ad. "No overtime charges, ever" or "We show up in the two-hour window or the visit's free" removes the friction that kills conversion silently.
  • Match the offer to intent. A cold prospect scrolling Reels needs a lower-commitment ask than someone who already searched your name.
  • If you change nothing else this quarter and just sharpen the offer, you will move ROAS more than any bid strategy will.

    Treat Creative as the New Targeting

    In the Advantage+ era, Meta's machine learning does the audience work. Your creative is now the targeting. The image, the hook, the first three seconds of the video — that's what tells the algorithm who to find. Weak creative means the system has nothing to optimize toward, so it spends your budget finding the cheapest impressions, not the most valuable buyers.

    Creative is also the fastest-decaying asset you own. A winning ad fatigues in two to four weeks at meaningful spend. Frequency climbs, CPMs drift up, and the same ad that printed money starts bleeding. If you are not shipping new creative on a schedule, your ROAS decays on a schedule.

    Volume beats brilliance

    Stop hunting for the one perfect ad. Ship volume and let the auction vote. A workable cadence for a local service business:

    1.

    Launch 5 to 8 new concepts every two weeks — not variations of the same idea, genuinely different angles.

    2.

    Test at the concept level, not the color-of-the-button level. Different hooks, different formats, different pain points. Small tweaks produce small, statistically meaningless differences.

    3.

    Give each ad enough budget to exit the learning phase — roughly 50 conversions per ad set per week — before you judge it. Killing an ad at 8 conversions is reading tea leaves.

    4.

    Keep a swipe file of your winners and reshoot them. Your best-performing angle from March is your template for a fresh batch in May.

    The formats that consistently pull for local service work: raw customer testimonials shot on a phone, honest before-and-after footage, a founder talking straight to camera, and problem-agitation clips that name the exact frustration a homeowner feels at 9pm with water on the floor.

    You are not competing for the perfect ad. You are competing for enough shots on goal that the algorithm can find your buyers before your budget runs out.

    Consolidate Campaigns So the Algorithm Can Learn

    The single most common structural mistake that suppresses ROAS: too many campaigns and ad sets, each starved of data. When you split a 3,000-dollar budget across nine ad sets, none of them ever collect enough conversions to exit the learning phase. The algorithm stays permanently stupid because you never let it get smart.

    Meta needs roughly 50 conversions per ad set per week to optimize with confidence. Do the math on your budget and your cost per conversion, and you'll usually find you can support far fewer ad sets than you're running.

  • Collapse your structure. Move toward a small number of campaigns with consolidated budgets. For most local service accounts, that means one or two prospecting campaigns, not twelve.
  • Let Advantage+ pool the budget. Campaign-level budget with broad targeting lets the system move money to what's working in real time instead of trapping it in a dying ad set.
  • Stop the daily fiddling. Every edit resets learning. If you're pausing and un-pausing ad sets each morning, you're holding the algorithm in a permanent cold start. Set it, feed it good creative, and give it a week.
  • Consolidation feels like giving up control. It's actually giving the machine enough signal to do the job you're paying it to do.

    Get Honest About Targeting in the Advantage+ Era

    Here's the uncomfortable truth: for most local service advertisers, hand-built interest stacks and lookalike ladders no longer beat broad targeting plus strong creative. The algorithm has more signal than you do. Fighting it with granular audiences usually just raises your CPMs and shrinks your pool.

    What actually still matters on the targeting side:

  • Geography and radius. This is the one lever where you genuinely know more than Meta. Tighten to the areas you actually service and can reach profitably. Don't pay to advertise a plumbing service to people 40 miles outside your route.
  • Exclusions. Exclude recent converters and existing customers from prospecting so you're not paying to reach people who already booked.
  • Retargeting, but honestly. Warm retargeting works, but the audiences are small for local businesses. Don't build ten retargeting ad sets off a trickle of traffic. One clean retargeting campaign is plenty.
  • Feed the machine, don't cage it. Broad plus great creative plus clean conversion data beats clever audience segmentation almost every time now.
  • If your instinct is to add more layers of targeting, resist it. The leverage moved to creative and data years ago.

    Win the Click You Already Paid For: Landing Pages

    You paid for the click. Now half of you throw that click at a slow, generic homepage and wonder why ROAS is soft. The landing page is a pure ROAS multiplier — improving conversion rate from 4 percent to 8 percent literally doubles your return on identical spend. Nothing in Ads Manager gives you that cleanly.

  • Message match. The headline on the page must echo the promise in the ad. If the ad said "89-dollar diagnostic," those words appear above the fold. Mismatch spikes bounce and torches your conversion rate.
  • Speed. A page that loads in over three seconds loses a large share of mobile visitors before they see anything. Nearly all your Meta traffic is mobile. Compress the images, cut the bloat.
  • One job per page. No navigation bar inviting people to wander. One offer, one form or one call button, repeated. Every extra choice leaks conversions.
  • Proof near the ask. Reviews, photos of real work, a recognizable service-area map, and trust badges sitting right next to the form — not buried at the bottom.
  • Make the form short. Name, phone, zip, problem. Every additional field costs you completions. You can qualify the rest on the call.
  • Run the math before you touch anything else in the account. A landing page lift compounds against your entire ad spend, not just one campaign.

    Feed the Pixel Better Data With CAPI

    Since iOS privacy changes, browser-based pixel tracking loses a meaningful chunk of conversion events. When Meta can't see conversions, it can't optimize toward them, and it can't attribute them — so your reported ROAS looks worse and your actual ROAS gets worse because the algorithm is optimizing half-blind.

    The Conversions API (CAPI) sends conversion data server-to-server, bypassing ad blockers and browser restrictions. It's not optional anymore for anyone serious about performance.

  • Install CAPI alongside the pixel, with proper event deduplication so you're not double-counting.
  • Send back-end events, not just page views. A "Lead" event is fine, but a "Qualified Lead" or "Booked Job" event tells Meta which clicks turn into actual money — and the algorithm starts hunting for more of those people.
  • Pass conversion values. If you feed Meta the dollar value of jobs, it can optimize for revenue instead of raw lead count. That's the difference between more leads and more profit.
  • This is the quiet lever that separates accounts that scale from accounts that plateau. Better data in, better optimization out.

    9.2×
    peak ROAS achieved across Thinxster client accounts

    The Underrated Lever: Speed-to-Lead and Follow-Up

    Here's what almost nobody optimizing Meta ads wants to hear: the same clicks can produce twice the booked jobs, and Ads Manager has nothing to do with it. ROAS is revenue over spend. If two businesses run the identical campaign and one books 30 percent of leads while the other books 15 percent, the first has double the ROAS on identical ad spend. The back end is where clicks become revenue, and most businesses leak the majority of theirs.

    The killer is response time. Lead conversion rates fall off a cliff after the first five minutes. A lead contacted in the first minute converts at a dramatically higher rate than one you call back two hours later — by which point they've already filled out three competitors' forms and booked the one who answered.

    Most local businesses respond to a web lead in hours, not minutes, because a human has to be free to make the call. That's the leak. The fix is systematizing the follow-up so speed doesn't depend on someone being available.

  • Respond in seconds, not hours. Thinxster's AI callers respond to inbound leads within 90 seconds, day or night, so no lead sits cold while a competitor picks up the phone.
  • Build the follow-up sequence, not a single attempt. Most jobs get booked on the second through fifth touch. A structured cadence of calls and texts across the first week — the kind we run inside GoHighLevel pipelines — recovers leads a single voicemail would lose.
  • Qualify before you route. Not every lead deserves your closer's time. Screening for real intent and service-area fit lifts close rates on the leads that matter — our qualification process runs at a 62 percent qualification rate, so human reps spend their hours on people ready to buy.
  • Track lead to booked job, not lead to form fill. If you don't measure the back end, you can't improve it, and you'll keep blaming the ads for a follow-up problem.
  • Fix the follow-up and you increase ROAS without changing a single thing about your ads. Same spend, same clicks, more revenue. That's why the operators generating serious numbers — Thinxster has driven over 102 million dollars for clients — obsess over the back end as much as the campaigns.

    Put It in Order

    The mistake is starting at the bottom of this list because bid tweaks feel like work. Do it in leverage order:

    1.

    Sharpen the offer — multiplies everything downstream.

    2.

    Ship creative volume on a cadence — your real targeting.

    3.

    Consolidate campaigns — let the algorithm learn.

    4.

    Go broad, tighten geography, exclude the right people.

    5.

    Rebuild the landing page — a pure spend multiplier.

    6.

    Install CAPI and send back-end events — optimize on truth.

    7.

    Systematize speed-to-lead and follow-up — turn the same clicks into more jobs.

    ROAS is won as much after the click as before it. The businesses stuck at 2x are usually great at none of these and busy optimizing bids. The ones hitting 9x are great at the whole chain — from the offer all the way through to the ninety-second callback.

    [Book a free strategy call](/book)

    Free Weekly Briefing

    One AI Marketing Tactic.
    Every Tuesday. Free.

    What's actually working across our client accounts right now — ROAS moves, follow-up sequences, creative angles. The stuff that isn't in any blog post yet.

    No spam. Unsubscribe anytime. 1,200+ business owners already in.

    Ready to Deploy

    SEE THIS IN
    YOUR BUSINESS.

    30 minutes. We scope the exact systems that apply to your situation and give you a plan.

    ★★★★★ Trusted by 47+ local service businesses

    BOOK A STRATEGY CALL →