How to Read Your Meta Ads Report Without Getting Lost in Metrics.

There's no reliable published ROAS benchmark for home service ads, and Meta's reported ROAS depends on attribution settings. Set your own floor from margin: the ROAS at which gross profit covers ad spend and fees. If you're below it, look at offer, lead response and the landing page before scaling.

Usually attribution settings. Meta's default since March 2026 is 7-day click, 1-day engage-through and 1-day view; some agencies report 7-day click only, which counts fewer leads. Accounts sending events through the Conversions API as well as the Pixel also record more of their real conversions. Ask your agency which setting they report, and compare both numbers with your CRM.

Cost per booked job exceeds your target for 14+ days AND creative has been rotated without improvement. One bad week is noise; two consistent weeks is signal. Pause, audit, relaunch rather than keep burning spend.

Team-facing: booked jobs this week, lead-to-appointment rate (affects their follow-up speed), appointment-to-close rate (affects their sales process). Owner/operator-facing: ROAS, cost per booked job, ad spend vs. target, creative fatigue signals. Mixing these creates confusion — your techs don't need CPM data; you don't need to see every CTR split.

Ads Manager → Reports → Export → 'Standard' format → CSV. Pulls campaign-level metrics for any date range. For daily automated tracking, use Meta's Marketing API or a connector tool (Supermetrics, Stitch, Funnel.io). Small accounts usually don't need the real-time API; a weekly manual CSV export + paste into your CRM-tracking sheet is sufficient.

Default Ads Manager view buries the metrics that matter. Customize columns to show, in order: (1) Cost per Result (your CPL); (2) Frequency; (3) CTR (link click-through rate, NOT all-click); (4) ROAS (Purchase ROAS or your custom event ROAS); (5) Reach (so you can see audience saturation). Hide the noise: CPM, CPC, post engagement, page likes — they're vanity for contractor accounts. Save this as a custom column preset called 'Contractor Weekly Review' and apply it every time you open Ads Manager. Reviews get faster once you know what 'normal' looks like for your account.

Three breakdowns that often reveal savings: (1) AGE — pull last 30 days, group by age bracket. If one age bracket produces leads at a much lower cost per booked job than others, shift creative toward that group; you can only restrict the audience by age if the campaign isn't in the Housing or Financial products special ad category. (2) PLACEMENT — compare Reels, Feed and Audience Network on cost per booked job, not just CPL; Audience Network is a common place for low-quality leads. If its CPL is far above your blended CPL, or its leads don't book, exclude it. (3) GENDER — check your own data; if it shows an extreme skew (e.g., 80% female engagement on a roofing campaign), the targeting may be off. Run breakdowns monthly, not weekly (you need volume to see real signal). Apply changes one breakdown at a time so you can isolate the impact.

It's a column-naming trap. 'Cost per Result' is generic — it equals whatever your campaign objective is optimized for (Reach, Traffic, Engagement, etc.). 'Cost per Lead' specifically counts the standard Lead event from the Pixel. Divergence usually means: (1) your campaign objective isn't 'Leads/Conversions' — check the Campaign tab and confirm; (2) your Lead event isn't firing correctly — Cost per Result might count clicks while Cost per Lead counts actual form submits, so the gap reveals a Pixel issue; (3) you're looking at different attribution windows for each column (the columns can have different default windows). Standardize: force both columns to the same attribution setting (the default since March 2026 is 7-day click, 1-day engage-through and 1-day view) and confirm Lead event is firing. Many 'mystery CPL spikes' turn out to be column-comparison errors rather than real performance issues.

Use multiple comparison frames, not just MoM. Three views to look at simultaneously: (1) MoM (this month vs last month) — useful for spotting acute changes (creative just launched, audience just expanded); volatile, prone to noise; (2) ROLLING 30-DAY (last 30 days vs prior 30 days, updated daily) — smoother trend, catches gradual drift better than calendar months; (3) YEAR-OVER-YEAR (this month vs same month last year) — controls for seasonality, the most useful comparison if your trade has strong seasonal patterns. Looking at MoM alone is misleading: a roofing contractor comparing March 2026 to February 2026 may see CPL drop as spring demand arrives — but that's seasonal, not a campaign improvement. Compare to March 2025 instead. Build all three frames into your dashboard. The right frame depends on what question you're asking; using only one frame leads to wrong conclusions.

The default views show a lot of columns; many are noise for contractor accounts. Remove these to declutter: (1) CPM (cost per 1000 impressions) — vanity metric for contractors, not actionable for lead-gen optimization; (2) CPC (cost per click) — use CTR instead, more meaningful; (3) Post Engagement — engagement metric, not conversion-relevant; (4) Page Likes — vanity, not aligned with Lead objective; (5) Video Plays — overlaps ThruPlays; (6) Cost per 1,000 Accounts Center accounts reached — too granular for contractor decisions. Keep Reach next to Frequency: together they show audience saturation. KEEP these instead: Cost per Result (the standard Lead event), Frequency, CTR (link), CPL, ROAS, Result Rate. Keep Reach next to Frequency, and save the lean view as a 'Contractor Weekly' preset so each review is quicker once you know what 'normal' looks like.

A fixed early-week slot, such as Tuesday morning. Three reasons it matters: (1) DATA FRESHNESS — by Tuesday you have the weekend and Monday in the numbers, so the review covers a full week rather than a flattering slice of it; (2) CALMER DECISIONS — a scheduled review is less likely to turn into a panic pause than a late-night check after a slow day; (3) ACTION TIMING — if you spot issues that need creative refresh, agency follow-up, or budget changes, Tuesday gives you the rest of the week to execute before the weekend. Friday-afternoon reviews leave changes pending over the weekend. Lock the same time slot weekly; treat it as a non-negotiable meeting with yourself. Reviewing ad hoc, when you 'have time', misses patterns that consistent review catches. Same time, every week, fewer surprises.

Three weekly cross-checks: (1) META REPORTED LEADS vs CRM RECEIVED LEADS — for example, Meta says you got 47 leads this week and your CRM received 42. The gap (5 leads, 11%) is your attribution noise. A small gap is normal; a large or growing one means something is broken (Pixel double-firing, CAPI miswired, or bots inflating Meta's count); (2) METHOD-OF-CONTACT survey — every closed customer answers 'how did you find us?' Tally the 'I saw your Facebook ad' answers vs Meta's reported attribution. Meta can overcount because of view-through and modelled conversions, so the survey answers are the conservative view; (3) INCREMENTALITY pause-test once per quarter — pause Meta entirely for 7 days, see what happens to TOTAL business volume. The drop is a rough read on your Meta contribution. All three together give you a 'truth ratio' for your own account; use it to discount Meta-reported ROAS before reporting to lenders or partners.

Frequency = how many times the average user has seen your ad. Four bands: (rules of thumb for 7-day frequency on prospecting audiences) (1) under 2.5 — fresh audience, ad is still finding new people; (2) 2.5-3.5 — fatigue approaching; have a replacement creative ready; (3) above 3.5 = REPLACE. Audience saturation is building and CTR usually starts dropping; (4) well above that, with CTR falling = the audience is burned through. Either pause the ad, refresh creative, or expand the audience. Industry myth: 'Higher frequency means more impressions, that's good.' Wrong — high frequency without commensurate creative refresh = same audience seeing the same ad repeatedly = scroll-past + creative blindness. Track frequency weekly so you catch the early warning. Add it to your column view in Ads Manager + check weekly. The 7-day rolling frequency tells you when fatigue is building before CPL actually spikes.

Three rankings Meta shows on each ad: Quality, Engagement, Conversion. Each is a relative scale (Above Average / Average / Below Average) compared to OTHER ads competing for similar audiences. Quick interpretation: (1) QUALITY RANKING — Meta's assessment of your ad's perceived quality (post-feedback, hide-rates, reports). 'Below Average' often means audiences are signaling negativity even if your CTR is fine; ad can get more expensive over time; (2) ENGAGEMENT RATE RANKING — your ad's expected engagement vs competitors. 'Below Average' here usually means your hook isn't strong enough; (3) CONVERSION RATE RANKING — your ad's expected conversion rate vs competitors. 'Below Average' means downstream funnel issues (LP, offer). Meta says these rankings are diagnostics and aren't used in the auction, but a 'Below Average' ranking usually goes with poorer delivery and higher costs. Action: replace any creative with a 'Below Average' ranking before scaling spend on it. These rankings are free diagnostic signals Meta gives you.