Across 20 months of ads, 1,538 people messaged us asking about the P.O Box. What we don't know is how many of them actually signed up. That's the only problem left to solve.
Ten P.O Box campaigns across three offices, always the same format: a short vertical video, and people tap through to message us on Instagram or Facebook. Here is every one of them, newest first, with what it cost and what it brought in.
| Month | Office | Offer | Invested | People who messaged | Cost each |
|---|---|---|---|---|---|
| Dec 2025 | El Paso | $55 | $195 | 106 | $1.84 |
| Nov 2025 | El Paso | $55 | $310 | 168 | $1.85 |
| Nov 2025 | Calexico | $30 | $250 | 130 | $1.92 |
| Oct 2025 | El Paso | $65 | $192 | 114 | $1.68 |
| Jul 2025 | El Paso | $65 | $197 | 148 | $1.33 |
| May 2025 | El Paso | $65 | $408 | 201 | $2.03 |
| May 2025 | San Ysidro | $45 | $260 | 136 | $1.91 |
| May 2025 | Calexico | $30 | $220 | 124 | $1.77 |
| Dec 2024 | El Paso | $40 | $330 | 192 | $1.72 |
| Dec 2024 | Calexico | $20 | $330 | 219 | $1.51 |
| Total | 3 offices | — | $2,692 | 1,538 | $1.75 |
The only campaigns in 2026 were the three Tax Season ones in February. That's nine months with no P.O Box advertising — which also means the audience is rested and hasn't been seeing these videos on repeat.
Over those 20 months, across three offices and six different offer prices, the cost of getting someone to message us never left the $1.33–$2.03 range. The advertising is doing its job, and we can plan how many people will come in.
February 2026 is the last thing that ran in the account, and it proves something important: San Ysidro responds very well when we invest in it.
Every campaign landed between $1.33 and $2.03 per person who messaged, and the cost never drifted up over time. We can say $300 will bring around 160 people, approximately, according to the historical data of ten campaigns.
El Paso at $65 produced both the cheapest result of all time ($1.33 in July) and the most expensive ($2.03 in May). At $55 it landed at $1.84 and $1.85. Discounting the P.O Box doesn't make the advertising work better; a stronger video does.
In San Ysidro a static image cost $6.07 per person against $1.60 for the reel — almost four times as much. And in 2026 Facebook barely showed the image posts at all (151 and 391 views). We're putting nothing into static images this month.
One P.O Box campaign in 20 months, against six in El Paso. When we did invest there for Tax Season in February, San Ysidro brought 213 people at $1.28 each — the second-best volume in the whole account. With the P.O Box at $65 there, that's where the margin is.
Every reel we've run, ranked by cost. The "requirements" video alone brought 470 of El Paso's 929 conversations.
This month we're investing $300 per office. Based on the historical data, that should bring around 160 people messaging each one. To cover the investment, five of them need to sign up.
If the team closes 1 in 10 conversations — about 16 customers per office — every $300 invested returns roughly $1,000.
Facebook tells us how many people wrote in, and that's where the information stops. There's no record of how many of those conversations turned into a P.O Box sign-up. That's 1,538 conversations without an answer.
On top of that, many conversations sit waiting. Someone asking about a P.O Box is usually comparing two or three options the same day — if nobody replies within the hour, they go with whoever did.
This isn't an advertising problem, it's a follow-up problem. We could push the cost per interested person from $1.75 down to $1.50 and it would barely move the needle. Going from closing 3 out of every 100 to closing 10 out of every 100 triples the result of the month.
We keep running ads that get people to message us on Instagram and Facebook, because that's what has worked. We reuse the videos with the strongest track record and test new reels per office. Every ad name below links to the actual creative in Ads Manager.
The El Paso test matters beyond this month: if the form works, we'll know whether it's worth building a landing page to capture customers.
Neither one costs money. Without them, September repeats what came before: plenty of conversations, no visibility on what they were worth.
Before we turn the ads on we'll set up an automatic reply that sends the price, the requirements, the address and the hours the second someone writes, with three quick-reply buttons. That covers the first few seconds. A real person should follow within 15 minutes.
Facebook also gives more reach to accounts that answer quickly, so this lowers the cost of the ads on top of improving the close rate.
A simple sheet with four columns: date, office, conversations handled, P.O Box sign-ups. Nothing else.
This is the one number that doesn't exist today. Without it we can't say whether the advertising is profitable, which office deserves more budget, or whether a landing page is worth building.