The CRM booked 105 sales appointments this month from 419 leads. 56 of them trace to paid media on $42,523 of matched Meta and Google spend. Google books more than half the leads it produces; Meta books one in seven.
Platform figures cover Aug 29 to Sep 27. CRM figures cover Sep 1 to 27, which is how the MarketSharp pull was run, and the spend used for cost per appointment is restricted to those same 27 days ($29,343 Meta, $13,180 Google) so the comparison is fair. Meta's platform lead count for that window is 280 against 281 Meta leads in the CRM, so the two systems agree to within one record.
The story
The window in three moves.
One channel books, one channel fills, and the account is paying more for both.
01
53.8%
Google books, Meta fills
Google turned 26 leads into 14 appointments. Meta turned 281 into 42. Fewer, better leads against more, weaker ones, and it decides where the next dollar goes.
02
$101
Meta's cost per lead rose 35%
Spend up 20% to $32,564, leads down 11% to 321. Click-through held at exactly 2.22%, so attention is fine. Conversion is what slipped.
03
$254
Search alone beats the old mix
Google's cost per conversion is now better than the Performance Max era managed, while paying $19.73 a click against $9.07. The August pause is still paying off.
Meta Ads
Where the leads came from.
Daily volume and cost, then the campaigns and creative behind it.
Leads & spend per day
Leads Daily spend
Window total 321 leadsHover any day for its leads and spend
→
The last nine days are a different account. Daily spend sat between $655 and $965 through the middle of the window, then stepped up to $1,276 to $1,717 from Sep 19 onward. Leads followed, but not proportionally: the week of Sep 19 cost $10,664 for 90 leads at $118 each, against $91 in the opening week. The final two days are the cheapest in the window at $78, which is worth watching rather than banking.
Campaign
Spend
Leads
Cost / lead
CTR
CPC
CPM
NZ 01 - Austin Best Performers Form
$13,136
95
$138
1.84%
$3.67
$67.56
NZ 01 - DFW Form Best cost / lead
$6,528
116
$56
2.57%
$1.51
$38.84
NZ 01 - Waco/Temple Form
$6,531
32
$204
2.34%
$2.66
$62.15
NZ 01 - DFW Best Performers Form
$6,370
78
$82
2.25%
$3.68
$82.83
Blended total
$32,564
321
$101
2.22%
$2.70
$59.81
i
Why there is no cost per appointment by campaign this window. The CRM records Meta as a single declared source, so its 42 appointments cannot be split back to the four campaigns. Read this table for which campaign produces leads most cheaply, and the pipeline section for what became of them. DFW Form is again the engine: more leads than any other campaign on half of Austin's budget, at $56 each on a $1.51 click.
Top-performing creative
Ranked by leads produced.
Hold rate is the share of people who started the video and stayed to ThruPlay, 15 seconds or the end. Finished is the share who watched all of it. Hook rate is not shown this window because the 3 second metric was not in this pull. 94 creatives carried spend.
The two best ads in the account are both from April. "Before and After" holds 18% of its viewers to ThruPlay and finishes with 6.1%, and the ASMR cut from the same batch holds 16.6% ($1,004, 18 leads at $56). Everything built since holds between 5.5% and 12%. The September re-cut of Ad Creative_v3 did well on volume, 29 leads at $65, and still holds only 8%. The Austin all-in-price ad took the most spend of any single creative, $7,102, at double the cost per lead of April's work. The brief for the next batch is to beat an 18% hold, which nothing in five months has done.
Who the leads come from
Leads by age & gender
Where the leads concentrate.
MenWomen
Leads by placement
Bar length is leads; the label shows cost per lead.
→
Instagram is the cheapest inventory in the account and holds 6% of the budget. Instagram Stories produced leads at $51 and Reels at $59, against $99 on Facebook Reels, which carries 55% of the spend. Facebook in-stream video and notifications together spent $1,481 for six leads and should come off. On audience: men 65+ produced 61 leads at $93 and women 55 to 64 produced 49 at $88, while women 65+ took the largest single slice of spend, $7,412, at $190 a lead.
i
A second account is also buying Austin. The EcoView owned Meta account spent $3,868 in the same window for 34 leads, including $1,785 on an Austin instant form at $78 a lead and $1,763 on an Austin landing page at $441. Two accounts bidding for the same Austin homeowners raises the price for both. Worth agreeing who owns Austin on Meta.
Google Ads
Half the clicks, the same conversions.
Daily conversions against daily spend, Sep 1 to 27.
Conversions & spend per day
Conversions Daily spend
Window total 52 conv.Hover any day for its conversions and spend
i
This is a Search only account now, and it is performing better than the mix it replaced. On the matched window, Search alone spent $13,180 for 52 conversions at $254. The four weeks before, with Performance Max and Demand Gen still running, spent $12,200 for 46 at $265. Click-through went from 1.75% to 3.60% even as the cost of a click rose from $9.07 to $19.73. Fewer clicks, better clicks.
Campaign split
Where the budget worked.
Spend and conversions by campaign. Brand search is separated so the true acquisition cost stays visible.
The one move to make this month is free. Core DFW shows for 11% of its searches and loses 54% of its impressions to budget. In the same account, Waco/Temple loses 3%: its budget physically cannot be spent, because the ceiling there is rank and geography rather than money. Waco and Window Types Waco took $1,167 this window for zero conversions. Moving that budget to DFW costs nothing and buys demand that is currently being turned away.
i
How to read the conversion count. Google counts 54 conversions, but most of the recorded actions are not leads: 159 Google Business Profile engagements, 51 direction requests and 32 profile website visits. The actions that represent someone asking EcoView for something are the 16 landing page leads, the 16 GA4 form submissions (the same people, counted by a second system) and the 52 phone actions. That is why this report leans on the CRM below.
i
Where the money leaks. Of 278 search terms that spent anything, 246 produced no conversion, accounting for $5,483 of $8,235 in term-level spend. The pattern is competitor and adjacent terms: "texas energy rebate" $230, "saginaw surplus" $171, "burris windows" $151, "don young windows dallas tx" $142, "glass guru frisco" $119. EcoView's own brand terms do convert, 4 conversions on $809. Mobile took $12,421 of the spend and produced 48 of the 54 conversions.
Call tracking
What happened on the phone.
Every tracked call in the window and what became of it, measured on the call itself rather than on the ad platform.
73
Tracked calls
Across 19 trackers
100%
Reached someone
73 of 73, no voicemail
51
Ran past a minute
151 minutes on the phone
How long the answered calls ran
Under 30 seconds
5
30 to 60 seconds
17
1 to 3 minutes
35
Over 3 minutes
16
Calls by day of the week
Monday 16 of 16 reached someone
16
Tuesday 10 of 10 reached someone
10
Wednesday 12 of 12 reached someone
12
Thursday 13 of 13 reached someone
13
Friday 17 of 17 reached someone
17
Saturday 4 of 4 reached someone
4
Sunday 1 of 1 reached someone
1
i
59 of the 64 unique callers had never called you before. Volume peaks at midday, 12 calls in the noon hour and 10 each at 9am, 1pm and 3pm, with only two calls after 4pm. The website pool number took 28 of the 73 calls, more than three times any other number.
→
Whisper messages are now set on every active tracker. All nine active trackers, including the 18 number website pool, announce the source before the call connects, so whoever answers knows what the caller responded to before they speak. The four trackers without a whisper are all disabled and carry no numbers. This closes the request from the sales meeting: it looked missing because the Google trackers already had whispers while the website pool, which takes the most calls, did not.
The pipeline
From leads to booked appointments.
What happened next in MarketSharp over Sep 1 to 27. Paid media is not one line in this CRM: it arrives under four separate declared sources, so looking for a single name shows only a fraction of it. Of 419 leads, 329 are paid.
Meta · Facebook / Instagram forms
Leads in CRM
281
Registered inquiries
Meta lead forms that landed in MarketSharp. The platform counted 280 for the same days.
Booked
42 14.9% set
Sales appointments set
One in seven Meta leads booked a sales presentation.
Cost / appt
$699
On $29,343 of spend
Cheaper per appointment than Google, on far more leads.
Google · books by phone
Leads in CRM
26
Forms and calls
15 landing page forms after removing four test entries, plus the calls filed under the Google source.
Booked
14 53.8% set
Sales appointments set
More than half of every lead Google produced booked a presentation.
Cost / appt
$941
On $13,180 of spend
More per appointment than Meta, on a fraction of the lead volume.
Channel
Spend
Leads in CRM
Appts set
Set rate
Cost / lead
Cost / appt
Google
$13,180
26
14
53.8%
$507
$941
Meta · Facebook / Instagram
$29,343
281
42
14.9%
$104
$699
Paid total
$42,523
307
56
18.2%
$139
$759
All sources in the CRM
-
419
105
25.1%
-
-
i
Nativz LP is Google, confirmed three ways. The v6 landing page form files in the CRM as "Nativz LP". GA4 shows all 17 submissions on that form came from google / cpc, none from Meta or organic. Google Ads counted 16 conversions on the same form. The CRM holds 15 real leads once four test entries are removed. Three independent systems, the same number.
i
Google's zero sales does not mean what it looks like. Google shows no closed sales this month. It also delivered nine sales presentations, none disqualified, and four homeowners are in Getting Other Estimates, meaning they saw the pitch and went to compare pricing. At the house close rate of 19%, nine presentations would be expected to produce fewer than two sales. The honest read is nine presentations with four still in play, not a channel that failed to sell. Twenty sales closed in the window across all sources.
i
Meta's lead count is not comparable to Google's without this. Most of Meta's 281 leads carry no city or zip at all. Among those that do, there are leads from Corpus Christi, El Paso, Lubbock, Odessa, Wichita Falls and San Antonio, and some from Kansas and Colorado, all outside the service area. They inflate the lead count and they are part of why the set rate reads 14.9% against Google's 53.8%. Tightening geography should raise the set rate without costing a single bookable lead.
i
There is no click level attribution in this CRM. The Ad Click ID field is empty on all 419 records, the same as in July, so every line above rests on what a person declared or what a rep selected at intake. That is why the whisper work matters, and it is the open question behind Google's cost per appointment: if a rep tagging a lead as "Google" is capturing the phone calls, it is $941; if only the landing page forms count, it is $2,635 on 15 leads and 5 appointments. Confirming what reps actually select is the cheapest measurement fix available.
What's next
Where we take it from here.
Each move ties back to what this window proved.
Operating plan
The next window.
Move the Waco budget to DFW Waco cannot spend what it has, 3% lost to budget, and produced no conversions on $1,167. Core DFW is turning away 54% of its demand for lack of budget. This costs nothing.
Rebuild creative against April, not last month "Before and After" and the ASMR cut still hold twice the attention of anything made since. The brief is to beat an 18% hold rate.
Shift placement money toward Instagram Reels and Stories produce leads at $51 to $59 on 6% of the budget. Facebook in-stream video and notifications spent $1,481 for six leads and come off.
Tighten Meta geography to the service area The fastest available improvement to Meta's set rate, and it costs nothing but a targeting edit.
Settle what a rep picks at intake One conversation decides whether Google's cost per appointment is $941 or $2,635, which changes how both channels get funded next quarter.
Negatives on the competitor and rebate terms $5,483 of term-level spend converted nothing this window.
The next chapter
Now we book more of them.
Google proves the demand is there and books more than half of what it finds. Meta proves the volume is there. The next window is about making Meta's leads look more like Google's, and letting DFW spend what it can clearly use.
